Long-term effect of vagus nerve stimulation on interictal epileptiform discharges in refractory epilepsy

Journal of the Neurological Sciences 284 (2009) 96–102

Contents lists available at ScienceDirect

Journal of the Neurological Sciences j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n s

Long-term effect of vagus nerve stimulation on interictal epileptiform discharges in refractory epilepsy Haiyang Wang a,1, Xiaoguang Chen a,1, Zhiguo Lin a,⁎, Zhengbo Shao b, Bomin Sun c, Hong Shen a, Li Liu a a b c

Department of Neurosurgery, The First Affiliated Hospital of Harbin Medical University, Harbin, 150001, China The Second Affiliated Hospital of Harbin Medical University, Harbin, 150001, China Clinical Center of Functional Neurosurgery, Ruijin hospital, Shanghai, 200025, China

a r t i c l e

i n f o

Article history: Received 16 December 2008 Revised 19 March 2009 Accepted 8 April 2009 Available online 8 May 2009 Keywords: Vagus nerve stimulation EEG Refractory epilepsy Interictal epileptiform discharges Seizure frequency Epileptiform activity

a b s t r a c t Background: Vagus nerve stimulation (VNS) therapy has been widely recognized as an effective alternative for the treatment of refractory epilepsy. However, the precise mechanism of VNS is poorly understood. The purpose of this study was to observe the long-term interictal EEG changes induced by VNS, and to investigate the probable mechanism of action of VNS in achieving seizure control. Methods: Eight patients with VNS were selected from two epilepsy centers in China (Harbin and Shanghai) between 2001 and 2004. We studied the clinical efficacy by long-term follow-up, ranging from 37 to 81 months (mean 55.8 months). Moreover, serial EEGs were performed at the different time (preoperative baseline, 3, 6, 12, and 24 months after VNS initiation) and the different states of VNS stimulator (“activation”, “deactivation” and “reactivation”). Results: A ≥ 50% seizure reduction was achieved in 12.5%, 62.5%, 75%, 62.5% and 75% of the total patients (n = 8) at 6, 12, 18, 24 and 36 months of post-VNS, respectively. The results revealed a statistically significant progressive decrease in the number of IEDs (interictal epileptiform discharges) on EEG with time (P b 0.01). Significant correlation had been highlighted after 6 months of VNS stimulation, between the reduction of seizure frequency and the decreasing of IEDs (P b 0.01). Furthermore, statistically significant difference of IEDs was seen when comparing the state of “deactivation” with the states of “activation” and “reactivation”, respectively (P b 0.01). However, there was no significant difference in IEDs between “activation” and “reactivation” (P N 0.05). Conclusions: VNS is an efficient, well-tolerated therapy for refractory epilepsy. It can induce progressive electrophysiological effect on epileptiform activity over time. This may reflect the mechanism of chronic action of VNS with desynchronization of EEG in achieving seizure control. © 2009 Elsevier B.V. All rights reserved.

1. Introduction The epilepsies are among the most common serious brain disorders, can occur at all ages, and are characterized by a variety of presentations and causes. There were about 50 million epileptics in the world, with 40 million of them in developing countries. In China, there are 400,000 new cases of epilepsy in every year. And approximately 9 million Chinese people have been suffering from this disease till now. Clinical researches show that with proper medical treatment, approximately 70% of the patients can avoid epilepsy seizures [1]. Moreover, despite the correct use of antiepileptic drugs (AEDs), in close to 30% of epilepsy patients, the seizure control is either not

⁎ Corresponding author. Tel.: +86 451 53672890. E-mail address: [email protected] (Z. Lin). 1 Both authors contribute to this work equally. 0022-510X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2009.04.012

satisfactory or it is intractable to pharmacotherapy. Among the nonpharmacological treatment options for refractory epilepsy, vagus nerve stimulation (VNS) occupies a unique position as an adjunctive treatment in the prevention and control of seizures in these cases. Since 1988, 65,000 patients with refractory epilepsy throughout the world have been treated by VNS, and it has proven to be effective and well-tolerated [2,3]. As an adjuvant therapy, VNS has comparable efficacy to most of the newer generation AEDs, with a better side effect profile [4]. Furthermore, there have been many case-reports of VNS in different types of epilepsy such as tuberous sclerosis, status epilepticus, hamartoma syndrome, infantile spasms, and progressive myoclonic epilepsy of Unverricht-Lundborg type. All reported good efficacy and low frequency of side effects [5]. Therefore, it has proven to be an efficacious broad-spectrum treatment for epilepsy that fully exerts its antiepileptic effect after several months of treatment [5]. Up to now, it has been extensively used in the Western countries. However, VNS is still a novel therapeutic approach in China, which has not been widely accepted by Chinese patients due to the high prices.

H. Wang et al. / Journal of the Neurological Sciences 284 (2009) 96–102

97

Table 1 Characteristics of epilepsy patients and effect of VNS on seizure frequency. Patient

Age (year)

Sex

Seizure types

Age at first seizure (year)

Epilepsy year (year)

Seizure frequency at baseline

Duration of follow-up (month)

Reduction in seizures (%)

1 2 3 4 5 6 7 8

17 31 34 41 33 35 25 28

F M F F F M M M

ABS CPS + SGTC CPS + SGTC CPS CPS + SGTC ABS GTCS CPS + SGTC

0.5 11 15 14 5 2 3 12

16.5 20 9 27 28 33 22 16

506 452 163 407 221 184 109 93

72 36 60 36 48 36 72 48

89.5 69.5 64.4 57.7 48.0 56.0 70.6 71.0

ABS: absence; CPS: complex partial seizures; SGTC: secondary generalized tonic clonic; GTCS: generalized tonic-clonic seizures.

Although the evidence base supporting the efficacy of VNS is strong, its exact mechanism of action remains unknown. In the past few years, some progress has been made through neurophysiological, neuroanatomical, neurochemical, and cerebral blood flow studies in patients and animals undergoing VNS [6–11]. Interesting results have been found in VNS-treated patients that underwent evoked potential measurements, cerebrospinal fluid investigation and functional imaging testing, such as PET, SPECT and fMRI and so on [10–16]. Current information suggests that VNS activates neuronal networks in the thalamus and other limbic structures and that norepinephrine may mediate the anti-seizure activity of VNS [5]. Desynchronization of abnormal synchronous epileptic activity is one of the hypotheses on the mode of action that might primarily be responsible for an antiseizure effect [17–20]. And following up on this idea, over the past years, numerous studies either on animals or on humans have been done to investigate the effect of VNS on EEG. According to these researches, VNS could suppress interictal spikes and seizures during stimulation [17,20,21]. The observation that VNS desynchronized the EEG activity suggested that this mechanism could be involved in VNS antiepileptic effects. However, other studies evaluating the effects of VNS on interictal EEG have produced controversial results. They did not find that VNS reduced interictal spikes and provided negative results [22–24]. Moreover, it is still on argument that whether this effect of VNS on interictal EEG is immediate/acute or chronic [12,22,25,26]. This is a pilot study that describes the long-term interictal EEG changes through the first group of VNS patients with refractory epilepsy in two departments (in Harbin and Shanghai, China). According to a long-term follow-up, we have studied serial EEGs on pre-operation, different period of post-VNS and different states (“activation”, “deactivation” and “reactivation”) of VNS stimulator. And the purpose of this study was: (a) to observe and verify the effect of suppression on ictal and interictal epileptiform activity (IEA) by VNS; (b) to determine this effect of VNS on interictal EEG is immediate/acute or chronic; (c) to evaluate whether the decreasing of interictal epileptiform discharges (IEDs) is correlated to the reduction of seizure frequency; and (d) to investigate the probable mechanism of action of VNS on the EEG in achieving seizure control.

2. Methods 2.1. Patients' characteristics Eight epilepsy patients with VNS implantation were received between 2001 and 2004. The group of patients was composed of seven adults and one adolescent, with the mean age of 30.5 years (mean 30.5 ± 7.3 years) at implant. Average age of the patients at implantation was 36.0 ± 7.3 years (ranging from 17 to 41 years). Mean age at onset of epilepsy was 7.8 ± 5.8 years (ranging from 6 months to 15 years of age) and mean duration of epilepsy (at the time of implantation) was 21.4 ± 7.7 years (ranging from 9 to 13 years). All

eight patients visited regularly on each follow-up time for at least 3 years (36 months). Duration of the follow-up (FU) time had a mean of 55.8 months (ranging from 37 to 81 month), and the baseline characteristics and FU details are summarized in Table 1. All of the patients had failed at least three antiepileptic medications and were not eligible for surgical resection of an epileptic focus. And a detail history, routine EEG, neuropsychological examination, CT, SPECT and structural magnetic resonance imaging were performed as a part of preoperative evaluation in these patients. In particular, seizure frequency and long-term video-EEG monitoring (Grass Technologies, Astro-Med, Inc. U.S.A.) recordings were recorded as preoperative baseline data. Informed consent was obtained from all patients or their guardians before the study. 2.2. VNS implantation and stimulation paradigms The Neuro Cybernetic Prosthesis vagal nerve stimulator is a battery-powered device. The body of the device is implanted in the upper chest; two connecting wires with electrodes are placed subcutaneously and attached to the left vagus nerve. The electrodes are connected to an implantable pulse generator (IPG), which is programmed by computer via a wand placed on the skin over it. VNS therapy pulse model 101 or 102 generator (Cyberonics Inc., Houston, TX, U.S.A.) was implanted for the patients, strictly followed the standard surgical technique [27] (the detailed information on implantation procedure of VNS therapy system can be found at http://www.vnstherapy.com/epilepsy/hcp/manuals/default.aspx). The stimulator began to switch on 2 weeks after the implantation with the initial parameters as 0.25 mA current, frequency 20 Hz, pulse width of 250 µs, a 30 s signal on time/5 min signal off time cycle. Visits for interrogation and adjustment of the device occurred at least every 1–2 months for the first 6 months and every 6–12 months thereafter. The stimulation intensity was gradually increased by 0.25 mA increments over weeks or months to individual appropriate parameters, which would be dependant on individual patient's response and tolerability (see Table 2). The magnetic parameters were adjusted

Table 2 Programming parameters of vagus nerve stimulation. No. Output Frequency current (Hz) (mA)

Pulse width (μs)

On/off cycles (s/min)

Magnet output current (mA)

Response to magnet use

1 2 3 4 5 6 7 8

250 500 250 250 500 250 500 250

30/5.0 30/5.0 30/5.0 30/5.0 30/5.0 30/5.0 30/5.0 30/5.0

1.25 1.50 1.00 2.00 1.75 1.50 1.25 1.50

P P P N N N P P

1.00 1.25 0.75 1.75 1.50 1.25 1.00 1.25

30 30 20 30 20 20 30 30

No., number; P, positive; N, negative.

98

H. Wang et al. / Journal of the Neurological Sciences 284 (2009) 96–102

Fig. 1. Changes in total seizure frequency of all eight patients at baseline vs. follow-up.

in all subjects for extrastimulation. The values of this output current were programmed at 0.25 mA higher than the automatically delivered stimulation in each subject and the other parameters for the extrastimulation remained unchanged. Once a patient responded to VNS therapy, further increases in output current were not necessary. 2.3. Follow-up All patients had at least one baseline assessment within 4–8 weeks before implantation. Thereafter, seizure-frequency data were prospectively obtained from patients' regular follow-up visits at 1, 3, 6, 12, 18, 24 months and then yearly interview after initial parameters were programmed, as well as seizure type, severity and all adverse events. AEDs were typically left unchanged during the first 12 months unless adjustments were necessitated by increasing seizures or drug toxicity. During the total period of follow-up, stimulator parameters could be changed in order to reach the best appropriate therapy for each individual patient. For evaluation of VNS-inducing EEG changes, serial EEGs were performed at the different time (preoperative baseline, 3, 6, 12, and 24 months after VNS initiation) and three different states of VNS stimulator (“activation”, “deactivation” and “reactivation”). Nineteen electrodes were placed according to the 10/20 system with 32–64 channel video-EEG monitoring. After at least 30 min of EEG was recorded at each follow-up visit, the pulse generator of VNS was deactivated or inhibited with the Cyberonics magnet or using the programming wand, in order to record the EEG under the “deactivation” state of VNS stimulator. After that, the stimulator would be restarted following 30 min of EEG recording in the same way. And then, EEG data were recorded again under the “reactivation” state for 30 min. In other words, at each follow-up visit, continuous EEG recordings were performed during at least 90 min to obtain the EEG data under three different states of VNS, i.e. “activation”, “deactivation” and “reactivation”. 2.4. Data analysis We analyzed all the seizure-frequency data, comparing the data at each follow-up time with baseline. Patients were identified as responders by having ≥50% reduction in seizure frequency from the baseline before the VNS implantation. In this study, interictal epileptiform activity (IEA) was defined by the number of interictal epileptiform discharges (IEDs), including isolated spikes, spikes and slow waves, spikes–waves, and polyspikes– waves during 30 min EEG recording at each follow-up. These EEG data were interpreted and analyzed by a neurophysiologist blinded of the experimental design. And the number of IEDs was counted by visual detection. In addition, the IEDs change was compared with seizure

reduction to analyze the correlation between them whether it existed or not. 2.5. Statistical analysis Software SPSS version 13.0 was used for all statistical analysis. Two-way ANOVA was performed to examine changes in the number of IEDs at baseline pre-VNS implantation and each follow-up with VNS stimulator at different conditions. LSD-t-test was used to multiple comparison of the number of IEDs recorded at different time. Paired ttest was applied to analyze the number of IEDs after VNS implantation with the stimulator at three conditions, “activation”, “deactivation” and “reactivation”, respectively. Spearman rank correlation coefficient (r) was used to calculate the correlation between extent of reduction in IEDs and seizure frequency. The level of significance was set at P-value b 0.05. 3. Results 3.1. Clinical efficacy As compared to baseline, data showed variable degrees reduction of seizure frequency on each follow-up time after VNS initiation. The mean seizure reduction after 1–6 years was, respectively, 50.9% (n = 8), 58.1% (n = 8), 59.9% (n = 8), 73.0% (n = 5), 66.3% (n = 3) and 80.1% (n = 2) (see Fig. 1). No patient was observed responding to VNS until 6 months after stimulation. A ≥50% seizure reduction was achieved in 12.5%, 62.5%, 75%, 62.5% and 75% of the total patients (n = 8) at 6, 12, 18, 24 and 36 months of post-VNS, respectively. At their last follow-up visit none of the eight patients was free of seizures. With a standard duty cycle (30 s on time/5 min off time), similar VNS stimulation parameters were used in eight patients. The different parameters were signal frequency (20 Hz or 30 Hz), output current (range from 0.75 to 1.75 mA) and pulse width (250 µs or 500 µs) (see Table 2). No one required replacement of the generator's battery after VNS initiation. Eight patients experienced VNS-inducing symptoms (hoarseness, sore throat, coughing, and shortness of breath) which were usually mild and transient. There were no cases of lead fracture and battery depletion. 3.2. EEG changes In Fig. 2, the black columns represent the number of IEDs of all patients at baseline and each follow-up time after VNS initiation. When comparing the post-VNS EEGs, including 6-month, 12-month and 24month, with baseline EEGs, respectively, there was a statistically

H. Wang et al. / Journal of the Neurological Sciences 284 (2009) 96–102

99

nificant difference in IEDs between “activation” and “reactivation” (P N 0.05, See Fig. 2). In follow-ups, six auras and three seizure attacks happened to be recorded during video-EEG monitoring in three of eight patients. And then, these attacks were terminated by the extrastimulation of VNS with a handheld magnet. Correspondingly, EEG data showed that the ictal synchronized paroxysmal epileptic discharges were also terminated by VNS (see Fig. 3). 4. Discussion

Fig. 2. The comparisons of the number of IEDs in baseline, each follow-up time and three different states of VNS stimulator, i.e. “activation”, “deactivation” and “reactivation”. ⁎P b 0.01 vs. baseline; #P N 0.05 vs. baseline; ΔP b 0.01; ☆P N 0.05.

significant decrease in the number of IEDs (P b 0.01). However, no statistically significant EEG changes were seen when comparing 3month post-VNS with the baseline (P N 0.05). Furthermore, after 6 months of VNS stimulation, significant correlation had been highlighted between the reduction of seizure frequency with the decreasing of IEDs (r = 0.929, P = 0.001 at 12-month; r = 0.766, P = 0.027 at 24month). Fig. 2 showed the number of IEDs in baseline, different time of follow-up and three different states of VNS stimulator, i.e. “activation”, “deactivation” and “reactivation”. When compared with baseline EEGs, the IEDs of the three different states totally decreased significantly (P b 0.01) at 6, 12 and 24 months. At each follow-up time, statistically significant difference of IEDs was seen when comparing the state of “deactivation” with states of “activation” and “reactivation”, respectively (P b 0.01). However, there was no sig-

Currently, VNS is considered as an effective adjunctive treatment for different types of seizures of patients with intractable epilepsy. A number of long-term VNS outcome studies, including E01–E05 studies, reported an improved effect in seizure reduction over a period of time [28–33]. In this study, clinical efficacy of VNS was observed by long-term follow-up in our group. The results illustrated continued seizure reduction after VNS therapy. Moreover, seizure rates of these eight patients declined with increasing duration of VNS therapy. Variable degrees reduction of seizure frequency was showed on each follow-up time after VNS initiation in all subjects (see Fig. 1). Improving seizure control supported the possibility of a sustained VNS effect on seizure reduction over time. Many studies have shown that efficacy in “VNS responders”, those with 50% or greater seizure reduction, increases steadily up to 40% to 50% over 3 years [34,35]. Recent literatures provided higher efficacy on seizure reduction after VNS operation. A Korean bicentric study [36] included 16 patients which could be followed up for at least more than 12 months. They concluded that, VNS resulted in a ≥50% reduction in seizure frequency in 50.0% (8/16) of children with 31.3% (5/16) of patients achieving a N90% reduction. A Belgian multicenter study [37] reviewed 138 patients with a follow-up of at least 12 months. The overall reduction in mean monthly seizure frequency was 51%. Responder rate was 59%. In this study, a ≥50% seizure reduction was achieved in 75% of the total patients at 36 months after VNS initiation. However, none of the patients was free of seizures at their last follow-up visit in our study.

Fig. 3. The EEG change induced by the extrastimulation of VNS with a handheld magnet. The ictal synchronized paroxysmal epileptic discharges were terminated following extrastimulation of VNS.

100

H. Wang et al. / Journal of the Neurological Sciences 284 (2009) 96–102

This result was similar to other studies [38,39]. It indicated that VNS is not curative but essentially a palliative treatment. Although VNS has been in clinical use for over a decade, the precise mechanism of action of VNS remains undetermined. Numerous animal and human studies have given insight into its therapeutic effect. A seizure is characterized by hypersynchronous and rhythmic electrical discharges. Therefore, EEG studies are one of the mainly focused topics on the mechanism of VNS action. However, when comparing animal and human EEG studies, there is some concern about the discrepancies that have been found. In various animal models of seizures and in humans, VNS disrupts this synchronicity and rhythmicity [40,41]. Yet, other studies had different viewpoints on EEG changes by VNS. In Hammond's research, VNS-induced suppression of interictal epileptic activity recorded by scalp EEG could not be demonstrated [23]. Another study assessed the acute effects of VNS on background EEG activity using magnet mode activation [22], and quantitative analysis did not reveal any changes in EEG background activity. In this study, serial EEG data were prospectively recorded at different time after VNS initiation. As shown in Fig. 2, when comparing the post-VNS EEGs, including 6-month, 12-month and 24-month, with baseline EEGs, respectively, there was a statistically significant decrease in the numbers of IEDs (P b 0.01). However, no statistically significant EEG changes were seen when comparing 3-month postVNS with the baseline (P N 0.05). This illustrated that EEG changes could be induced by VNS, and the improvement in interictal EEG of epilepsies was not immediate but tended to increase over months until a treatment plateau of VNS was reached. After VNS was approved by FDA in USA, VNS had been shown in several human studies [12,25,42] to suppress both ictal and interictal epileptiform EEG activity. Koo [25] studied long-term effect of VNS on EEG. In his study, 5 patients were found to have progressive increase in duration of spike-free intervals and decrease in duration and frequency of spikes/spike and wave activity with time. 16 patients showed a progressive decrease in the number of spikes on EEG over time. However, Kuba [12] studied IEDs on EEG and showed the clear effect of acute VNS on reducing IEDs in humans during the stimulation period as compared to the baseline. And he demonstrated the acute effect of VNS on the suppression of the interictal and ictal epileptic EEG finding. In our study, we took the number of IEDs as the observation index which could reflect the changes in interictal epileptiform activity (IEA). And we found that there was a statistically significant progressive decrease in the number of IEDs on EEGs of all patients studied (see Fig. 2). These changes were progressively showed on interictal EEG with time. In other words, it signified that VNS probably played a role in EEG changes with its chronic effect. Although the clinical effectiveness of VNS has been demonstrated, the action mechanism and relation between the seizures and IEDs remain unknown [8]. Previous studies evaluating the correlation between spikes on EEG and seizure frequency have produced controversial results. In two studies [12,25] where changes in EEG due to VNS were found, one of them did not show any correlation between them. Besides, it was found in a recent study [43] that there was no direct correlation between the extent of clinical improvement and the degree of spike reduction. Nevertheless, different viewpoints were concluded in other studies [13,44,45]. In our study, after 6 months of VNS stimulation, significant correlation had been highlighted between the reduction of seizure frequency and the decreasing of IEDs (r = 0.929, P = 0.001 at 12-month; r = 0.766, P = 0.027 at 24month). The result was consistent with that observed in another study [45], where it showed an association between clinical improvement in seizure reduction and decrease in epileptiform activity in the EEG. Clinically, patients with more intractable seizures were often the ones who had more IEDs on EEG and were more likely to benefit from VNS, with a higher seizure reduction [30,46]. This seems to indicate that interictal epileptic activity may reflect a state of neuronal excitability related to seizure frequency. However, there was no correlation

between seizure reduction and IEDs decreasing during the first 6 months after VNS implantation (r = 0.286, P = 0.493 at 3-month; r = 0.429, P = 0.289 at 6-month). The long-term changes of interictal EEG during VNS and its delayed correlation with the seizure reduction indicated that VNS often took more time, weeks, months and even years, to achieve optimum seizure control and interictal EEG changes. VNS efficacy may be based on long-term neuronal changes. Zagon [6] suggested that chronic VNS can suppress neuronal excitability and induce a sustained hyperpolarization in cell membranes of cortical neurons. This study focused on the question of EEG changes induced by VNS. We found that interictal EEG activity was influenced by VNS, and IEDs in the EEGs of all patients progressively decreased with time. Moreover, six auras and three seizure attacks in three of the eight patients happened to be recorded during video-EEG monitoring. And then, these attacks were terminated by the extrastimulation of VNS with a handheld magnet. Correspondingly, video-EEG data showed the ictal synchronized paroxysmal epileptic discharges were also terminated by VNS (see Fig. 3). Previous studies expounded it to us that VNS is programmed to produce chronic intermittent electrical pulses. By moving a handheld magnet over the device, additional electrical stimulation trains can be provided by the patient or a companion in case of an aura or seizure [47]. Boon [48] intended to evaluate the clinical efficacy of VNS in both the programmed intermittent stimulation and the magnet stimulation mode in his study. And he found that more than half of the patients who reported a positive effect of magnet stimulation became responders. The magnet feature of the device was a helpful tool to improve seizure control. His study is the first to explore the efficacy of magnet-induced VNS. In our study, 62.5% (5/8) of patients had positive response to magnet use (see Table 2). And this magnet-induced extrastimulation was observed to exert a role in controlling the oncoming aura and seizure. The result was in agreement with the literature, which found that ictal event halted both electrographically and behaviorally [23]. Current studies indicated that the anti-seizure effect of VNS were concordant both in animals and human subjects. Furthermore, the seizure could be interrupted if the stimulation activated early enough before the ictal EEG discharge. Some authors suggested that this anti-seizure effect was strong enough in some patients to control epilepsy [49,50]. Therefore, this in-demand magnet-induced therapy can be useful for epilepsy patients to abort or reduce the severity or duration of seizures. As we know, the VNS therapy system delivers stimulation on a chronic, intermittent basis. The initial typically recommended stimulation parameters are a 30-second period of stimulation, which is referred to as ON time, followed by a 5-minute period without stimulation, which is referred to as OFF time. In addition, epilepsy patients can use a small, handheld magnet provided with the VNS Therapy System to activate or deactivate stimulation manually. In our study, serial EEG data in three different states of VNS stimulator, i.e. “activation”, “deactivation” and “reactivation” were also recorded in addition. The aim was to observe and analyze the VNS-induced interictal EEG changes in different conditions of stimulator, and determine whether the EEG changes were acute or chronic effect of VNS. Firstly, from Fig. 2 we could find that, the IEDs in the three different states totally decreased significantly (P b 0.01) when compared with baseline, no matter if the VNS stimulator was in the state of activation or deactivation. This illustrated that VNS could progressively cause interictal EEG changes in human patients over time. Such changes seemed to be unobvious until 6 months after VNS initiation. This indicated the chronic effect or action of VNS on human EEG. And the inhibition effect of VNS on IEDs became more and more obvious with time. This was in agreement with the literatures where interictal epileptiform discharges were decreased after VNS gradually over time [12,25,43]. Moreover, there was significant difference between the IEDs in the states of activation and deactivation. From Fig. 2, we found

H. Wang et al. / Journal of the Neurological Sciences 284 (2009) 96–102

that more numbers of IEDs in the state of deactivation than activation and reactivation at each follow-up. However, although the number of IEDs increased when the pulse generator was deactivated or stopped by magnet, the increased IEDs in the state of deactivation had not yet recovered to baseline, but still significantly less than the one in the same state at previous follow-up (see Fig. 2). It demonstrated that VNS still had inhibition effect on interictal epileptiform activity, although the stimulation was suspended for a period of time. We believe that this inhibition effect is a result of long-term action of VNS, with the gradual improvement on EEG. Furthermore, we found in our study that, there was no difference in the number of IEDs between “activation” and “reactivation” (P N 0.05), even though significant difference of IEDs was seen when comparing the state of “deactivation” with “activation” and “reactivation”, respectively. This illustrated that the changes of IEA induced by VNS was not completely the result of acute or transient effectiveness of VNS. This may be due to a cumulative effect induced by a long-term action of VNS before the pulse generator was deactivated manually. The long-term cumulative effect of VNS was mentioned in many previous studies [25,31,32,51,52]. Improvement effect in seizure reduction was a result of VNS stimulation over a period of time. In Hallböök's study [43], he reported long-term effects of VNS on epileptiform activity, and found the persistence of the anti-seizure effects and the gradual decrease in IEDs with time. Then he believed that VNS could induce long-term neuron-modulating effects. In our study, IEDs increased when VNS was deactivated. It should be noted that, firstly, this increasing of IEDs occurred after a sudden deactivation of VNS stimulator. Secondly, IEDs increased but the number was still less than baseline and the one in the same state at previous follow-up. Lastly, IEDs decreased again when the pulse generator of VNS was reactivated, and no difference in the number of IEDs between the states of activation and reactivation. We could not believe these changes were due to immediate effect of VNS. In order to better explain the reasons for this phenomenon, it was postulated that the EEG change in different states of VNS was a “false appearance” which might be caused by a “rebound effect” of VNS stimulation, just like “drug removal reaction” of antiepileptic drugs. Actually, as a chronic and intermittent electrical stimulation, VNS affects brain regions commonly involved in epileptic networks [53]. And VNS probably induces such changes through modulation or alteration of neuronal synapses, which takes time to develop, accounting for the progressive EEG changes with time [25]. Therefore, it often needs more time for VNS to play its role in controlling seizure. However, there has no any literature where a “rebound effect” of VNS stimulation is mentioned or described. So far we could only deduce in this way, and further study is expected in the future. In this study, we further confirm the VNS-induced chronic and cumulative effect on interictal EEG. Although approved for an exact effect on EEG and seizure control, VNS cannot absolutely reduce seizure frequency and epileptic discharges on EEG. It does not usually result in complete cessation of seizures clinically. The role of VNS on interictal EEG is not believed to be an “all-or-none” effect. VNS plays an important role in inhibiting interictal epileptiform activity, and its effects gradually increase over time. In this sense, VNS is still a palliative adjuvant therapy. Similarly, many studies [30,54–57] have shown that VNS is an efficacious and palliative therapeutic alternative in the treatment of refractory epilepsy. Our results are in agreement with these studies. In summary, although VNS therapy has been shown to be an effective alternative, the precise mechanism of action of VNS and how it suppresses seizures remain to be elucidated. In this pilot study, we observed the first group of VNS patients in two departments of China. According to a long-term follow-up, we have studied clinical efficacy and serial EEGs at different time of follow-up and different states of VNS stimulation. We confirmed that VNS is an efficient, well-tolerated and add-on therapy for drug-resistant epilepsies, with mild stimulation-related side effects. And also VNS can produce a measurable

101

electrophysiological effect on epileptiform activity. Furthermore, these continued seizure reduction and improvement in EEG are both induced by VNS over time. As far as we know, this is also the first report on interictal EEG change at different states of VNS stimulator. Our results indicate that VNS achieve seizure control and effect on EEG in the way of progressive desynchronization of EEG. This clinical efficacy and VNS-induced EEG change is chronic or long-term effect rather than acute or immediate reaction. We can also indicate that the mechanism of VNS is absolutely not fairly simple. However, VNS affects the epileptic pathophysiological process on different functional levels. Of note is that the sample size of our group is small and more research needs to be done to provide evidence of VNS-induced chronic EEG changes. And controlled trials, larger groups of patients and more prolonged follow-ups are also needed to verify our clinical observations. However, this study has shown a Chinese experience on VNS therapy. Moreover, this pilot study has accumulated valuable data for the further research and has also laid a foundation for clinical application of VNS in the future.

Acknowledgements We are grateful to Prof. Fuming Yang for his guidance in this study and to Dr. Xiaohua Hou for her assistance in analyzing the EEG data.

References [1] Villanueva V, Carreño M, Herranz Fernández JL, Gil-Nagel A. Surgery and electrical stimulation in epilepsy: selection of candidates and results. Neurologist 2007;13(6 Suppl 1):S29–37. [2] Roux FX, Turak B, Landré E. Vagus nerve stimulation for the treatment of refractory epilepsy. Neurochirurgie 2008;54:332–9. [3] Schachter SC. Vagus nerve stimulation therapy summary: five years after FDA approval. Neurology 2002;59(6 Suppl 4):S15–20. [4] Guberman A. Vagus nerve stimulation in the treatment of epilepsy. CMAJ 2004;171:1165–6. [5] Ben-Menachem E. Vagus-nerve stimulation for the treatment of epilepsy—a review. Lancet Neurol 2002;1:477–82. [6] Zagon A, Kemeny AA. Slow hyperpolarization in cortical neurons: a possible mechanism behind vagus nerve simulation therapy for refractory epilepsy? Epilepsia 2000;41:1382–9. [7] Rutecki P. Anatomical, physiological, and theoretical basis for the antiepileptic effect of vagus nerve stimulation. Epilepsia 1990;31(Suppl 2):S1–6. [8] Henry TR. Therapeutic mechanisms of vagus nerve stimulation. Neurology 2002;59:S3–S14. [9] Follesa P, Biggio F, Gorini G, Caria S, Talani G, Dazzi L, et al. Vagus nerve stimulation increases norepinephrine concentration and the gene expression of BDNF and bFGF in the rat brain. Brain Res 2007;1179:28–34. [10] Vonck K, Boon P, Van Laere K, D'Havé M, Vandekerckhove T, O'Connor S, et al. Acute single photon emission computed tomographic study of vagus nerve stimulation in refractory epilepsy. Epilepsia 2000;41:601–9. [11] Henry TR, Votaw JR, Pennell PB, Epstein CM, Bakay RA, Faber TL, et al. Acute blood flow changes and efficacy of vagus nerve stimulation in partial epilepsy. Neurology 1999;52:1166–73. [12] Kuba R, Guzaninova M, Brazdil M, Novák Z, Chrastina J, Rektor I. Effect of vagal nerve stimulation on interictal epileptiform discharges: a scalp EEG study. Epilepsia 2002;43:1181–8. [13] Ebus SC, Majoie HJ, Arends JB, Boon PJ. Can spikes predict seizure frequency? Results of a pilot study in severe childhood epilepsies treated with vagus nerve stimulation. Seizure 2004;13:494–8. [14] Hammond EJ, Uthman BM, Wilder BJ, Ben-Menachem E, Hamberger A, Hedner T, et al. Neurochemical effects of vagus nerve stimulation in humans. Brain Res 1992;583:300–3. [15] Ben-Menachem E, Hamberger A, Hedner T, Hammond EJ, Uthman BM, Slater J, et al. Effects of vagus nerve stimulation on amino acids and other metabolites in the CSF of patients with partial seizures. Epilepsy Res 1995;20:221–7. [16] Lomarev M, Denslow S, Nahas Z, Chae JH, George MS, Bohning DE. Vagus nerve stimulation (VNS) synchronized BOLD fMRI suggests that VNS in depressed adults has frequency/dose dependent effects. J Psychiatr Res 2002;36:219–27. [17] Zabara J. Peripheral control of hypersynchronous discharge in epilepsy. Electroencephalogr Clin Neurophysiol 1985;61:162. [18] Zanchetti A, Wang SC, Mortuzzi G. The effect of vagal stimulation on the EEG pattern of the cat. Electroencephalogr Clin Neurophysiol 1952;4:357–61. [19] Woodbury DM, Woodbury JW. Effects of vagal stimulation on experimentally induced seizures in rats. Epilepsia 1990;31(Suppl 2):S7–S19. [20] McLachlan RS. Suppression of interictal spikes and seizures by stimulation of the vagus nerve. Epilepsia 1993;34:918–23.

102

H. Wang et al. / Journal of the Neurological Sciences 284 (2009) 96–102

[21] Chase M, Nakamura Y, Clemente C, Sterman M. Afferent vagal stimulation: neurographic correlates of induced EEG synchronization and desynchronization. Brain Res 1967;5:236–49. [22] Salinsky MC, Burchiel KJ. Vagus nerve stimulation has no effect on awake EEG rhythms in humans. Epilepsia 1993;34:299–304. [23] Hammond E, Uthman B, Reid S, Wilder B. Electrophysiological studies of cervical vagus nerve stimulation in humans: I. EEG effects. Epilepsia 1992;33:1013–20. [24] Lockard JS, Congdon WC, DuCharme LL. Feasibility and safety of vagal stimulation in monkey model. Epilepsia 1990;31(Suppl 2):S20–6. [25] Koo B. EEG changes with vagus nerve stimulation. J Clin Neurophysiol 2001;18:434–41. [26] Zhang JL, Zhang SP, Zhang HQ. Antiepileptic effects of electroacupuncture vs vagus nerve stimulation on cortical epileptiform activities. J Neurol Sci 2008;270:114–21. [27] Heck C, Helmers SL, DeGiorgio CM. Vagus nerve stimulation therapy, epilepsy, and device parameters: scientific basis and recommendations for use. Neurology 2002;59:S31–7. [28] Spanaki MV, Allen LS, Mueller WM, Morris III GL. Vagus nerve stimulation therapy: 5year or greater outcome at a university-based epilepsy centre. Seizure 2004;13:587–90. [29] Uthman BM, Reichl AM, Dean JC, Eisenschenk S, Gilmore R, Reid S, et al. Effectiveness of vagus nerve stimulation in epilepsy patients, a 12-year observation. Neurology 2004;63:1124–6. [30] Vonck K, Boon B, D'Havé M, Vandekerckhove T, O'Connor S, De Reuck J. Long-term results of vagus nerve stimulation in refractory epilepsy. Seizure 1999;8:328–34. [31] DeGiorgio CM, Schachter SC, Handforth A, Salinsky M, Thompson J, Uthman B, et al. Prospective long-term study of vagus nerve stimulation for the treatment of refractory seizures. Epilepsia Sep 2000;41(9):1195–200. [32] Morris GL, Mueller WM. The Vagus Nerve Stimulation Study Group EO1–EO5. Long-term treatment with vagus nerve stimulation in patients with refractory epilepsy. Neurology 1999;53:1731–5. [33] Labar D. Vagus nerve stimulation for 1 year in 269 patients on unchanged antiepileptic drugs. Seizure 2004;13:392–8. [34] Helmers SL, Wheless JW, Frost M, Gates J, Levisohn P, Tardo C, et al. Vagus nerve stimulation therapy in pediatric patients with refractory epilepsy: retrospective study. J Child Neurol 2001;16:843–8. [35] Murphy JV, Torkelson R, Dowler I, Simon S, Hudson S. Vagal nerve stimulation in refractory epilepsy: the first 100 patients receiving vagal nerve stimulation at a pediatric epilepsy centre. Arch Pediatr Adolesc Med 2003;157:560–4. [36] Kang HC, Hwang YS, Kim DS, Kim HD. Vagus nerve stimulation in pediatric intractable epilepsy: a Korean bicentric study. Acta Neurochir Suppl 2006;99:93–6. [37] De Herdt V, Boon P, Ceulemans B, Hauman H, Lagae L, Legros B, et al. Vagus nerve stimulation for refractory epilepsy: a Belgian multicenter study. Eur J Paediatr Neurol 2007;11:261–9. [38] Montavont A, Demarquay G, Ryvlin P, Rabilloud M, Guénot M, Ostrowsky K, et al. Long-term efficiency of vagus nerve stimulation (VNS) in non-surgical refractory epilepsies in adolescents and adults. Rev Neurol (Paris) 2007;163:1169–77. [39] Abubakr A, Wambacq I. Long-term outcome of vagus nerve stimulation therapy in patients with refractory epilepsy. J Clin Neurosci 2008;15:127–9. [40] Jaseja H. Vagal nerve stimulation technique: enhancing its efficacy and acceptability by augmentation with auto activation and deactivation mode of operation. Med Hypotheses 2004;63:76–9.

[41] Marrosu F, Santoni F, Puligheddu M, Barberini L, Maleci A, Ennas F, et al. Increase in 20–50 Hz (gamma frequencies) power spectrum and synchronization after chronic vagal nerve stimulation. Clin Neurophysiol 2005;116:2026–36. [42] Olejniczak PW, Fisch BJ, Carey M, Butterbaugh G, Happel L, Tardo C. The effect of vagus nerve stimulation on epileptiform activity recorded from hippocampal depth electrodes. Epilepsia 2001;42:423–9. [43] Hallböök T, Lundgren J, Blennow G, Strömblad LG, Rosén I. Long term effects on epileptiform activity with vagus nerve stimulation in children. Seizure 2005;14:527–33. [44] Janszky J, Hoppe M, Behne F, Tuxhorn I. Vagus nerve stimulation: predictors of seizure freedom. J Neurol Neurosurg Psychiatry 2005;76:384–9. [45] Santiago-Rodríguez E, Alonso-Vanegas M, Cárdenas-Morales L, Harmony T, Bernardino M, Fernández-Bouzas A. Effects of two different cycles of vagus nerve stimulation on interictal epileptiform discharges. Seizure 2006;15:615–20. [46] Labar D, Murphy J, Tecoma E. Vagus nerve stimulation for medication-resistant generalized epilepsy. E04 VNS Study Group. Neurology 1999;52:1510–2. [47] Terry R, Tarver WB, Zabara J. An implantable neurocybernetic prosthesis system. Epilepsia 1990;31(Suppl 2):S33–7. [48] Boon P, Vonck K, Van Walleghem P, D'Havé M, Goossens L, Vandekerckhove T, et al. Programmed and magnet-induced vagus nerve stimulation for refractory epilepsy. J Clin Neurophysiol 2001;18:402–7. [49] Vonck K, Van Laere K, Dedeurwaerdere S, Caemaert J, De Reuck J, Boon P. The mechanism of action of vagus nerve stimulation for refractory epilepsy: the current status. J Clin Neurophysiol 2001;18:394–401. [50] Fanselow EE, Reid AP, Nicolelis MAL. Reduction of pentylenetetrazole-induced seizure activity in awake rats by seizure-triggered trigeminal nerve stimulation. J Neurosci 2000;20:8160–6. [51] George R, Salinsky M, Kuzniecky R, Rosenfeld W, Bergen D, Tarver WB, et al. Vagus nerve stimulation for treatment of partial seizures. 3. Long-term follow-up on first 67 patients exiting a controlled study. Epilepsia 1994;35:637–43. [52] Ardesch JJ, Buschman HP, Wagener-Schimmel LJ, van der Aa HE, Hageman G. Vagus nerve stimulation for medically refractory epilepsy: a long-term follow-up study. Seizure 2007;16:579–85. [53] Li Y, Mogul DJ. Electrical control of epileptic seizures. J Clin Neurophysiol 2007;24:197–204. [54] Ben-Menachem E, Mañon-Espaillat R, Ristanovic R, Wilder BJ, Stefan H, Mirza W, et al. Vagus nerve stimulation for treatment of partial seizures: 1. A controlled study of effect on seizures. First International Vagus Nerve Stimulation Study Group. Epilepsia 1994;35:616–26. [55] George R, Sonnen A, Upton A, Salinsky M, Ristanovic R, Bergen D, et al. A randomized controlled trial of chronic vagus nerve stimulation for treatment of medically intractable seizures. Neurology 1995;45:224–30. [56] Handforth A, DeGiorgio CM, Schachter SC, Uthman BM, Naritoku DK, Tecoma ES, et al. Vagus nerve stimulation therapy for partial-onset seizures: a randomized activecontrol trial. Neurology 1998;51:48–55. [57] Boon P, Vonck K, Vandekerckhove T, D'have M, Nieuwenhuis L, Michielsen G, et al. Vagus nerve stimulation for medically refractory epilepsy; efficacy and costbenefit analysis. Acta Neurochir 1999;141:447–52.