Commentary: Physical Approaches for the Treatment of
Epilepsy: Electrical and Magnetic Stimulation and Cooling
Wolfgang Löscher,*†Andrew J. Cole,‡and Michael J. McLean?
*Department of Pharmacology, Toxicology, and Pharmacy, University of Veterinary Medicine Hannover, Hannover D-30559,
Germany,†Center for Systems Neuroscience, Hannover, D-30559 Germany,‡Epilepsy Service, Massachusetts General Hospital,
Boston, Massachusetts 02114,?Vanderbilt University Medical Center, Department of Neurology, Nashville, Tennessee 37212
Summary: Physical approaches for the treatment of epilepsy
currently under study or development include electrical or
magnetic brain stimulators and cooling devices, each of
which may be implanted or applied externally. Some devices
may stimulate peripheral structures, whereas others may be
implanted directly into the brain. Stimulation may be deliv-
ered chronically, intermittently, or in response to either man-
ual activation or computer-based detection of events of
interest. Physical approaches may therefore ultimately be
appropriate for seizure prophylaxis by causing a modifica-
tion of the underlying substrate, presumably with a reduction
in the intrinsic excitability of cerebral structures, or for
seizure termination, by interfering with the spontaneous dis-
charge of pathological neuronal networks. Clinical trials of
device-based therapies are difficult due to ethical issues sur-
rounding device implantation, problems with blinding, poten-
tial carryover effects that may occur in crossover designs if
substrate modification occurs, and subject heterogeneity. Un-
resolved issues in the development of physical treatments in-
clude optimization of stimulation parameters, identification of
the optimal volume of brain to be stimulated, development of
adequate power supplies to stimulate the necessary areas, and a
determination that stimulation itself does not promote epilep-
togenesis or adverse long-term effects on normal brain func-
tion. Key Words: Brain stimulation, vagus nerve, hypother-
mia, neuromodulation, seizures.
At least 30% of patients with epilepsy have uncon-
trolled seizures, despite treatment with anti-epileptic
drugs (AEDs). In addition, adverse effects of AEDs com-
promise the quality of life of many patients whether or
not seizures are controlled adequately or may even pre-
vent titration to effective doses. Carefully selected AED-
resistant patients may benefit from surgical resection, but
epilepsy surgery has risks and costs, and many forms of
AED-resistant epilepsy are not surgically amenable. The
adverse seizure and neurobehavioral prognosis in patients
with AED-resistant epilepsy provides ample justification to
search for alternative treatments. Physical approaches, in-
cluding brain stimulation and cooling, are among the most
promising experimental alternative therapies.
During the Workshop on “New Horizons in the
Development of Antiepileptic Drugs: Non-Traditional
Approaches to Treat Epilepsy” (March 5–7, 2008,
Clontarf Castle, Dublin, Ireland), one session focused
on electrical and magnetic stimulation and cooling as
experimental alternative therapies for AED-resistant
epilepsy. There were five talks within this topic, fol-
lowed by “pro” and “con” summaries by M. J. McLean
and W. Löscher, and general discussion. The session
was chaired by N. Delanty and A. Cole. In the follow-
ing, we will briefly discuss the alternative therapies
reviewed in this session.
ELECTRICAL BRAIN STIMULATION
Neurostimulation-based treatments such as vagus nerve
stimulation (VNS), deep brain stimulation (DBS), and re-
sponsive neurostimulation (RNS), have gained increasing
attention in recent years.1,2Underlying all these tech-
niques is the idea that extrinsic stimulation can reduce
hyperexcitability in seizure-generating networks, per-
haps by interrupting the build-up to the ictal transition.
VNS, the only approved stimulation therapy for epilepsy,
has been licensed for nearly a decade in several countries
Address correspondence and reprint requests to: Wolfgang Löscher,
M.D., Department of Pharmacology, Toxicology and Pharmacy, Uni-
versity of Veterinary Medicine, Bünteweg 17, Hannover, D-30559
Germany. E-mail: email@example.com.
Neurotherapeutics: The Journal of the American Society for Experimental NeuroTherapeutics
Vol. 6, 258–262, April 2009 © The American Society for Experimental NeuroTherapeutics, Inc.
as an adjunctive therapy for refractory epilepsy,3,4and
for several years as a treatment for medication-resistant
depression. Depression is a frequent comorbid condition
with epilepsy,5and its presence before seizure onset
predicts refractoriness of partial seizures.6Some AEDs
may exacerbate depression and increase the risk of sui-
cide.7Thus, VNS, alone or in combination with AEDs,
offers the possibility of improving the quality of patients’
lives by improving seizure control, minimizing the sys-
temic load of AEDs, and improving mood. Coexistence
of depressive symptoms may be a factor in selecting
VNS treatment.8Despite the hopes and expectations
these prospects raise, efficacy of VNS has been limited.
VNS uses a pulse generator that transmits impulses to
the left vagus nerve via an implantable electrode and can
be implanted by surgeons familiar with the anatomy of
the cervical vagus nerve. Afferent fibers of the vagus
nerve project to several brain regions, including the thal-
amus, amygdala, and forebrain, through the nucleus trac-
tus solitarius and the medullary reticular formation. Elec-
trical stimulation of the vagus nerve is thought to affect
pathogenetic mechanisms, which are thought to be in-
volved in the generation and propagation of seizures3and
depression.6,7However, the exact mechanisms by which
VNS exerts its effects are unknown. As an adjunctive
therapy in AED-resistant patients, VNS typically reduces
seizure frequency in “VNS responders,” although few
patients become seizure-free.3The efficacy of VNS in
reducing the frequency or severity of seizures is highly
variable, and no clear predictive factors for responders
have been identified. Even though the pulse parameters
can be varied widely, protocols that treat seizure types
selectively have not been established. Early VNS treat-
ment in the course of pharmacoresistant epilepsy may
improve the benefit.9Several controlled trials have as-
sessed the efficacy of VNS in patients with epilepsy.2
Because patients can sense VNS stimulation, traditional
blinded, placebo-controlled trials have not been practi-
cal. Instead, to obtain “controlled” data, a dose-response
design with “high” and “low” stimulation was adopted,
showing 25 to 30% mean reduction in seizure frequency
for the high-stimulation group versus 6 to 15% for the
low-stimulation group.2Thus, the clinical efficacy of
VNS is limited, and at best comparable with adjunctive
therapy with novel AEDs. However, individual patients
may benefit from VNS, particularly because the thera-
peutic efficacy is sustained and VNS is generally well-
tolerated. There appears to be a delay before the maximal
anticonvulsant effect of VNS develops, a phenomenon
that is not understood. If a reduction in seizure frequency
is not observed after a prolonged period (e.g., 18 months
as recommended by Cyberonics, Inc.), the device should
be deactivated and removed.
DBS is an investigational approach that requires in-
tracranial (stereotaxic) surgery, and thus is more invasive
than VNS.1,2DBS has been used extensively as a treat-
ment for movement disorders, such as Parkinson’s dis-
ease.10Various brain targets for DBS in epilepsy have
been investigated, including thalamus, hippocampus, and
subthalamic nucleus11,12with the idea of stimulating
more discreet networks that may be more closely linked
to the seizure generator than the more widespread effects
of VNS.13Several results have been encouraging, but not
conclusive. As with other investigational therapies, the
efficacy of DBS was higher in open trials than in the few
available controlled trials.2The best structure to stimu-
late and the most effective stimuli to use are unknown as
yet. The risks of DBS demands careful consideration of
its value on an individual basis. Electrode implantation in
the brain has been associated with a 5% risk for intra-
cranial hemorrhage and 5% for infection.1
RNS is a more recent concept based on the idea that
spontaneous epileptic activity can be terminated by in-
hibitory polarization produced by slow intermittent elec-
trical stimulation applied to the region of onset.14,15Po-
tential advantages of RNS include the ability to be
applied to bilateral or multiple foci, eloquent cortex that
can not be excised, pathways of seizure propagation or
pathways with global modulatory effects à la VNS, with
or without tailored AED therapy.14Seizure-prediction
algorithms could be used, even though only seizure de-
tection algorithms have been studied so far.14Combining
other physical approaches, such as cooling, with RNS is
another prospect that comes to mind.16In contrast to
DBS, the electrical stimulation is not continuously ap-
plied, but only acutely after detection of seizure onset.
The ultimate goal of such a closed-loop feedback system
for seizure detection and brain stimulation is the creation
of a totally implantable device. The optimal location
(deep brain or cortical), characteristics of the stimulation
(frequency, current, duration), and threshold for event
detection and stimulation are still to be determined. In a
feasibility study of 65 patients, the responder rate was
43% for complex partial seizures and 35% for partial
(simple and complex) and secondarily generalized tonic-
clonic seizures.15A large, controlled clinical trial is cur-
rently underway. Controls will receive sham stimulation
via implanted electrodes. Risks of RNS are comparable
to those of DBS. Furthermore, both DBS and RNS may
be associated with micro-lesions and mechanical effects
due to electrode implantation, which by themselves may
affect seizure activity.
Transcranial magnetic and direct current stimulation
In comparison to VNS, DBS, and RNS, transcranial
magnetic stimulation and transcranial direct current stim-
ulation (tDCS) are the simplest and least invasive ap-
proaches to brain stimulation for epilepsy.17–19Transcra-
nial magnetic stimulation with either a hand-held magnet
or a frame that can be aligned to predetermined coordi-
PHYSICAL APPROACHES FOR EPILEPSY259
Neurotherapeutics, Vol. 6, No. 2, 2009
nates is used as a diagnostic tool in patients, particularly
to measure parameters related to cortical excitatory and
inhibitory function.19Low-frequency, repetitive trans-
cranial magnetic stimulation (rTMS) reduces motor cor-
tex excitability in humans, and a number of studies have
explored this technique as a therapeutic approach in pa-
tients with AED-resistant epilepsy.2Results have been
equivocal, with small and transient responses. A small,
controlled study with rTMS in patients with mesial tem-
poral or neocortical localization-related epilepsy did not
demonstrate any significant effect on seizure frequency,
although actively treated patients had a 16% (nonsignif-
icant) mean reduction in seizure frequency, but only in
the first two weeks after stimulation.20There was a trend
toward a greater effect in patients with neocortical rather
than mesial temporal foci, suggesting that mesial tempo-
ral structures probably lie too deep beneath the surface
for effective stimulation by rTMS.2In another study with
slow rTMS (0.3 Hz) significant reduction of EEG epi-
leptiform activity was observed in one third of subjects,
supporting a biological effect; however, seizure fre-
quency was not significantly reduced.21A few rTMS
experiments have been performed in animal models of
epilepsy (e.g., Ebert and Ziemann22), but the need to
immobilize the animals during stimulation and the local
heating associated with rTMS have prevented any mean-
More recently, tDCS has been proposed to constitute a
simple and effective means for brain stimulation in pa-
tients with epilepsy.17tDCS involves the application of
low currents to the scalp via cathodal or anodal elec-
trodes and has been shown to affect a range of motor,
somatosensory, visual, affective, and cognitive func-
tions.18A controlled, clinical trial of cathodal tDCS in
patients with malformations of cortical development and
refractory epilepsy showed a nonsignificant trend toward
decrease in seizure frequency after active versus sham
DC treatment.23In an animal model of neocortical epi-
lepsy, cathodal tDCS induced anticonvulsant effects, al-
beit of lower magnitude than those induced by the AED
An attractive alternative physical approach for termi-
nating and perhaps preventing focal seizures is local
(focal) cooling.16The potential clinical utility of thera-
peutic cooling or hypothermia for neurological disorders
associated with neuronal injury or excessive electrical
activity, including traumatic brain injury, stroke, as-
phyxia, and epilepsy, has been discussed for decades, but
there has been renewed interest in this approach in recent
years. Of particular interest has been the potential of
hypothermia to be neuroprotective, which has been dem-
onstrated in a variety of preclinical and clinical stud-
ies.25,26Different techniques for therapeutic cooling
have been used both preclinically and clinically, includ-
ing surface cooling, endovascular cooling, and direct
brain cooling, such as focal cooling using a piezoelectric
device.16,27In vitro and in vivo experimental epilepsy
studies have demonstrated that cooling diminishes par-
oxysmal bursting and can reduce or stop seizure
activity.16External cooling was reported to terminate
AED-refractory status epilepticus in two clinical inves-
tigations and one study in a rat model.28,29In patients,
direct cooling of the brain under anesthesia was de-
scribed to block seizures.30,31Transient head cooling in
infants with hypoxic-ischemic encephalopathy increased
survival and reduced the severity of EEG abnormali-
ties.32However, such techniques of therapeutic cooling
are not suitable for chronic application in patients with
AED-resistant epilepsy. In a series of experiments, Roth-
man et al.16have developed and evaluated thermoelectric
devices that may be applicable for the chronic therapy of
partial epilepsy. This technique suppresses fully kindled
seizures in awake, freely moving rats33without produc-
ing any signs of neuronal injury.34The ultimate expec-
tation of Rothman et al.16is an implantable device that
would allow cooling at a site of seizure initiation under
closed-loop feedback control, similar to the design used
for RNS. Despite the optimism generated by preclinical
and uncontrolled clinical studies in different neurological
diseases, there is currently no evidence from randomized
controlled trials to support the routine use of cooling in
traumatic brain injury, stroke, or epilepsy.26
Physical (nonpharmacological) approaches offer inter-
esting alternatives for the treatment of AED-resistant
epilepsies. Table 1 lists relative advantages and disad-
vantages of each of the approaches discussed based on
currently available data and experience. One of these
approaches, VNS, has been used in tens of thousands of
patients with epilepsy since its approval, whereas other
physical techniques are still at an experimental stage.
Overall, the anticonvulsant efficacy that can be obtained
with physical approaches is relatively low and compara-
ble with the moderate reduction in seizure frequency that
is typically seen with adjunctive therapy with new AEDs
in patients with intractable epilepsy. Furthermore, in
contrast to drug trials, traditional placebo-controlled tri-
als are not practical with devices such as VNS, DBS,
RNS, or cooling, due to the patients’ ability to perceive
when stimulation or cooling is “on,” In some studies,
sham controls with the same invasive procedure, but no
stimulation have been used, but this does not resolve the
problem of sensing the stimulus. Because of the diffi-
culty of blinding the subjects, trials with high- and low-
stimulation parameters may be useful, as performed with
VNS. The phenomenon of “reversion to the mean” may
LÖSCHER ET AL.260
Neurotherapeutics, Vol. 6, No. 2, 2009
Table 1. Potential Advantages and Disadvantages of Proposed Physical Treatment Approaches*
Foci Efficacy Tolerability Potential LimitationsPotential Advantages
Broad range of stimulation parameters possible, but
values for specific seizure types not completely
explored or understood.
– Approved by regulatory agencies in
– Conveniently programmable.
– Seizure freedom possible for some.
– Stimulation of discreet networks close
to seizure generator is possible.
– Multiple possible implantation sites
affords some individualization.
– Configurable detection and
stimulation algorithms allow
individualization of therapy.
– Programmable restimulations for
persistent seizure activity.
– Adjustment of stimulus parameters
minimizes proconvulsant effects and
possibility of microlesions.
– Large multicenter controlled clinical
trial in progress.
– Potential combination with other
physical approaches, e.g. cooling, at
– Intermittent or daily stimulation may
be effective, as for bipolar disorder,
an indication approved by regulatory
agencies in some countries.
– Direct hyperpolarizing current is not
Sterotaxic electrode placement necessary.
Risk of hemorrhage and infection.
Determining the best site for electrode placement is
complex, invasive, and largely empirical.
Efficacy of treatment triggered by seizure initiation
(vs seizure prediction) remains to be determined.
Seizure freedom not observed in proof-of-concept
Electrode placement is critical.
Microlesions, mechanical effects, and electrolysis
near electrodes can affect seizure frequency.
Available technology limits stimulation frequency.
Depth reached by effective field is limited, thereby
limiting efficacy and utility.
The range of useful current strengths is likely to be
Depth of current spread is likely to be limited or
Optimum treatment protocols have not been determined.
Excessive focal cooling could produce tissue damage.
Implantable (e.g., piezoelectric) devices useful for
clinical trials in man are limited.
Determination of temperature at cooling site is essential
for optimization and is difficult with extracranial
Systemic effects (e.g., on the heart, limit external
– Chronic focal cooling with
electrodes of the type used in RNS
may be possible.
– Combination with other physical
techniques, e.g. RNS, may be
– Cooling may be neuroprotective.
?, ??, ???, and ???? represents relative value. – indicates noninvasive or not available. ? indicates no data available.
DBS ? deep brain stimulation; RNS ? responsive neurostimulation; rTMS ? repetitive transcranial magnetic stimulation; tDCS ? transcranial direct current stimulation; VNS ? vagus nerve stimulation.
*All physical approaches below can be combined with antiepileptic drug therapy. Stimulation techniques have potential proconvulsant effects that can be addressed by varying stimulation parameters.
PHYSICAL APPROACHES FOR EPILEPSY
Neurotherapeutics, Vol. 6, No. 2, 2009
produce an additional bias.2Patients are more likely to
enter an experimental trial when their seizure frequency
is high, leading to the risk that decline in frequency
during the trial, due to random variation, might be mis-
interpreted as a therapeutic effect. This risk may be
higher for studies involving invasive procedures or sur-
gery than AEDs.2This by no means proves that physical
approaches are not effective, but emphasizes the impor-
tance of randomized clinical trials to establish the effi-
cacy and safety of these methods. Table 2 lists some of
the potential difficulties encountered in conducting trials
of physical approaches to epilepsy treatment.
1. Theodore WH, Fisher RS. Brain stimulation for epilepsy. Lancet
2. Theodore WH, Fisher R. Brain stimulation for epilepsy. Acta
Neurochir Suppl 2007;97:261–272.
3. Groves DA, Brown VJ. Vagal nerve stimulation: a review of its
applications and potential mechanisms that mediate its clinical
effects. Neurosci Biobehav Rev 2005;29:493–500.
4. Boon P, De H, V, Vonck K, Van Roost D. Clinical experience with
vagus nerve stimulation and deep brain stimulation in epilepsy.
Acta Neurochir Suppl 2007;97:273–280.
5. Milby AH, Halpern CH, Baltuch GH. Vagus nerve stimulation for
epilepsy and depression. Neurotherapeutics 2008;5:75–85.
6. Kanner AM. Depression in epilepsy: a complex relation with un-
expected consequences. Curr Opin Neurol 2008;21:190–194.
7. Miller JM, Kustra RP, Vuong A, Hammer AE, Messenheimer JA.
Depressive symptoms in epilepsy: prevalence, impact, aetiology,
biological correlates and effect of treatment with antiepileptic
drugs. Drugs 2008;68:1493–1509.
8. Brodtkorb E, Mula M. Optimizing therapy of seizures in adult
patients with psychiatric comorbidity. Neurology 2006;67(suppl
9. Helmers SL, Griesemer DA, Dean CJ, et al. Observations on the
use of vagus nerve stimulation earlier in the course of pharmacore-
sistant epilepsy: patients with seizures for six years or less. The
10. Breit S, Schulz JB, Benabid AL. Deep brain stimulation. Cell
Tissue Res 2004;318:275–288.
11. Vonck K, Boon P, Van Roost D. Anatomical and physiological
basis and mechanism of action of neurostimulation for epilepsy.
Acta Neurochir Suppl 2007;97:321–328.
12. Halpern CH, Samadani U, Litt B, Jaggi JL, Baltuch GH. Deep
brain stimulation for epilepsy. Neurotherapeutics 2008;5:59–67.
13. Li Y, Mogul DJ. Electrical control of epileptic seizures. J Clin
14. Morrell M. Brain stimulation for epilepsy: can scheduled or re-
sponsive neurostimulation stop seizures? Curr Opin Neurol 2006;
15. Sun FT, Morrell MJ, Wharen Jr. RE. Responsive cortical stimula-
tion for the treatment of epilepsy. Neurotherapeutics 2008;5:
16. Rothman SM, Smyth MD, Yang XF, Peterson GP. Focal cooling
for epilepsy: an alternative therapy that might actually work. Ep-
ilepsy Behav 2005;7:214–221.
17. Paulus W. Transcranial direct current stimulation (tDCS). Suppl
Clin Neurophysiol 2003;56:249–254.
18. Been G, Ngo TT, Miller SM, Fitzgerald PB. The use of tDCS and
CVS as methods of non-invasive brain stimulation. Brain Res Rev
19. Chen R, Cros D, Curra A, Di L, V, Lefaucheur JP, Magistris MR,
et al. The clinical diagnostic utility of transcranial magnetic stim-
ulation: report of an IFCN committee. Clin Neurophysiol 2008;
20. Theodore WH, Hunter K, Chen R, Vega-Bermudez F, Boroojerdi
B, Reeves-Tyer P, et al. Transcranial magnetic stimulation for the
treatment of seizures: a controlled study. Neurology 2002;59:560–
21. Cantello R, Rossi S, Varrase C, et al. Slow repetitive TMS for
drug-resistant epilepsy: clinical and EEG findings of a placebo-
controlled trial. Epilepsia 2007;48:366–374.
22. Ebert U, Ziemann U. Altered seizure susceptibility after high-
frequency transcranial magnetic stimulation in rats. Neurosci Lett
23. Fregni F, Thome-Souza S, Nitsche MA, Freedman SD, Valente
KD, Pascual-Leone A. A controlled clinical trial of cathodal DC
polarization in patients with refractory epilepsy. Epilepsia 2006;
24. Liebetanz D, Klinker F, Hering D, et al. Anticonvulsant effects of
transcranial direct-current stimulation (tDCS) in the rat cortical
ramp model of focal epilepsy. Epilepsia 2006;47:1216–1224.
25. Gunn AJ, Gunn TR. The ‘pharmacology’ of neuronal rescue with
cerebral hypothermia. Early Human Development 1998;53:19–35.
26. Polderman KH. Induced hypothermia and fever control for pre-
vention and treatment of neurological injuries. Lancet 2008;371:
27. Hemmen TM, Lyden PD. Induced hypothermia for acute stroke.
Stroke 2007;38(Suppl 2):794–799.
28. Vastola EF, Homan R, Rosen A. Inhibition of focal seizures by
moderate hypothermia. A clinical and experimental study. Arch
29. Schmitt FC, Buchheim K, Meierkord H, Holtkamp M. Anticon-
vulsant properties of hypothermia in experimental status epilepti-
cus. Neurobiol Dis 2006;23:689–696.
30. Sartorius CJ, Berger MS. Rapid termination of intraoperative stim-
ulation-evoked seizures with application of cold Ringer’s lactate to
the cortex. Technical note. J Neurosurg 1998;88:349–351.
31. Karkar KM, Garcia PA, Bateman LM, Smyth MD, Barbaro NM,
Berger M. Focal cooling suppresses spontaneous epileptiform ac-
tivity without changing the cortical motor threshold. Epilepsia
32. Gluckman PD, Wyatt JS, Azzopardi D, et al. On behalf of the
CoolCap Study Group. Selective head cooling with mild systemic
hypothermia after neonatal encephalopathy: multicentre random-
ized trial. Lancet 2005;365:663–670.
33. Burton JM, Peebles GA, Binder DK, Rothman SM, Smyth MD.
Transcortical cooling inhibits hippocampal-kindled seizures in the
rat. Epilepsia 2005;46:1881–1887.
34. Yang XF, Kennedy BR, Lomber SG, Schmidt RE, Rothman SM.
Cooling produces minimal neuropathology in neocortex and hip-
pocampus. Neurobiol Dis 2006;23:637–643.
Table 2. Common Problems in Designing Trials for
Assessing the Efficacy of Physical Approaches
● Blinding may be difficult or impossible
● Relationship of implant localization and zone of
epileptogenesis may be difficult to establish
● Risks of infection, injury, and potential epileptogenisis
● Availability of well established alternative treatments
for focal epilepsy negatively impact recruitment
● Choice of stimulation parameters is often based on little
● Chronic intermittent stimulation may lead to
tachyphylaxis and/or proconvulsant effects that
confound interpretation of study results
● Potential latency to full effect may lead to false-
LÖSCHER ET AL. 262
Neurotherapeutics, Vol. 6, No. 2, 2009