Abstract
Background
Migraine is one of the most disabling neurological disorders. The current pharmacological armamentarium is not satisfying for a large proportion of patients because the responder rate does not exceed 50% on average and the most effective drugs often induce intolerable side effects. During recent years, noninvasive central and peripheral neuromodulation methods have been explored for migraine treatment.
Overview
A review of the available evidence suggests that noninvasive neuromodulation techniques could be beneficial for migraine patients. The transcranial stimulation methods allow modulating selectively cortical activity and can thus be curtailed to the patient’s pathophysiological profile, while transcutaneous stimulation of pericranial nerves likely modulates central pain control centers. Occipital single-pulse transcranial magnetic stimulation and transcutaneous supraorbital stimulation have the strongest evidence respectively for acute and preventive treatment. Transcranial direct current stimulation and repetitive magnetic stimulation are promising in pilot studies, but large sham-controlled trials are not yet available.
Conclusions
The noninvasive neurostimulation methods are promising for migraine treatment and devoid of serious adverse effects allowing their combination with drug therapies. Their application in clinical practice will depend on the industry’s capacity to develop portable and user-friendly devices, and on the scientists’ capacity to prove their efficacy in randomized sham-controlled trials.
Keywords
Background
Noninvasive neurostimulation methods used in migraine target either transcutaneously peripheral nerves or transcranially the brain itself. The rationale for the two is different (see Coppola et al. (1) in this issue). Peripheral nerve stimulation has been used for a long time in chronic pain disorders and, like percutaneous occipital nerve stimulation in refractory chronic cluster headache, it probably acts by modulating central pain control areas (2).
The rationale for using transcranial neurostimulation methods in migraine therapy is based on the fact these methods are susceptible to normalizing the abnormalities of cortical responsivity found in migraine patients between attacks (see Coppola et al. (1) in this issue).
The noninvasive neurostimulation methods are of particular interest in migraine because they are much better tolerated and accepted than invasive techniques and hence must not be restricted like the latter to the most disabled refractory patients. At present, however, few randomized sham-controlled trials are available and the data presented for several neurostimulation methods or devices must therefore be considered with some reservation. In this review we have summarized the advances in the use of noninvasive neurostimulation methods for the acute and prophylactic therapy of migraine.
Transcranial magnetic stimulation (TMS)
Single-pulse TMS (sTMS) (Table 1)
Single-pulse TMS studies as acute treatment.
CM: chronic migraine; MA: migraine with aura patients; MO: migraine without aura patients; MOH: medication-overuse headache; N/A: not applicable; sTMS: single-pulse TMS; TMS: transcranial magnetic stimulation; HIT-6: Headache Impact Test; M: migraine patients.
In a multicenter, randomized, double-blind, parallel-group, sham-controlled study, Lipton et al. (4) assessed the efficacy of a novel portable TMS device (Neuralieve®, eNeura, USA) applied over the visual cortex in 164 MA patients (n = 82 sham), treating up to three attacks with aura over three months within one hour of aura onset. Real sTMS was superior to sham for pain freedom at two hours (39% vs. 22% respectively, i.e. 17% of therapeutic gain), and for sustained pain freedom at 24 hours (29% vs. 16%) and 48 hours (27% vs. 13%). There was no statistical difference between the real sTMS and sham stimulation groups for the combination of no or mild pain at two hours, use of rescue medication, consistency of effect, patients’ global assessment of pain or the Migraine Disability Assessment Score Questionnaire (MIDAS) score (4). Because of the rather small effect size compared to placebo and the lack of significant changes in patients’ satisfaction and disability, one may question the clinical usefulness of sTMS in MA. In a post-market pilot program in the United Kingdom based on telephone interviews, Bhola et al. (5) evaluated sTMS in 190 episodic (EM) or chronic migraine (CM) patients who had previously found acute medications intolerable, ineffective, or contraindicated. Sixty-two percent of patients reported the device was effective in reducing or alleviating their migraine headache, and 59% reported a reduction in the number of headache days after 12 weeks of treatment. No serious adverse events occurred, but the discontinuation rate was higher than expected (55%) probably because of dissatisfaction. Besides the facts that this survey is non-controlled, based on telephone interviews for follow-up and not based on headache diaries, other major weaknesses are inclusion of patients taking preventive antimigraine drugs or overusing acute medications. At this stage, sTMS can thus not be considered to have an established utility for migraine treatment in clinical practice.
Repetitive TMS (Table 2)
Repetitive TMS studies as preventive treatment.
BDI: Beck Depression Inventory; CDH: chronic daily headache; CM: chronic migraine; HF: high frequency; LF: low frequency; M: migraineur patient; MA: migraine with aura patient; MIDAS: Migraine Disability Assessment Score Questionnaire; MO: migraine without aura patient; MT: motor threshold; rTMS: repetitive transcranial magnetic stimulation; STAI: state-trait anxiety inventory; DCS: transcranial direct current stimulation.
Repetitive TMS (rTMS) has been tested as a preventive treatment both for EM and for CM with equivocal results. In an open-label, sham-controlled study in which 12 high-frequency (HF) (20 Hz) daily rTMS sessions were applied over the left dorsolateral prefrontal cortex (DLPFC), Brighina et al. (6) demonstrated the potential safety and efficacy of rTMS as headache prophylaxis in 12 CM patients (six real, six sham). They found a significant reduction of attack frequency at two months with rTMS (−53%) compared to sham (−7%, therapeutic gain of 46%) and a parallel reduction of acute medication intake and headache severity, without side effects.
In a recently published study, a group of 14 CM patients (seven real, seven sham) underwent a total of 23 sessions of either HF (10 Hz) rTMS or sham treatment over the left DLPFC within eight weeks. Real rTMS was less effective in reducing headache days than sham (−15% vs. −58.1%), with a negative therapeutic gain (−43.1%), while pain intensity, depression and anxiety indexes, as well as MIDAS scores, decreased equally in both groups (10). Inclusion of patients, most of whom were acute medication overusers and taking prophylactic drugs, is a major weakness of this study. This may account for negative outcome, although the authors argue that activation of the DLPFC might have attenuated the placebo effect.
In a group of EM patients (N = 27), Teepker et al. (7) failed to find a significant reduction of headache days after five sessions of low-frequency rTMS versus sham over the vertex over eight weeks. However, severity of headache, functional disability and number of acute rescue medications were significantly lower in the real rTMS group compared to the sham group.
Another group of researchers administered a total of three sessions of HF rTMS or sham over the left frontal cortex on alternate days first in an open-label study (8) and thereafter in a randomized, placebo-controlled, double-blind trial (9). In the latter trial, they enrolled a heterogeneous group of 100 patients with various diagnoses: 93 MO, 7 MA, among whom 60 were CM and 28 medication-overuse headache (MOH) (sham n = 50 or rTMS n = 50). They observed after one month a significant improvement in headache frequency in the rTMS group (−78.7%) compared to sham (−33.3%) and in headache intensity (−76.6% vs. −27.1%). The clinical improvement was especially pronounced in CM patients. Functional disability and use of acute rescue medication were significantly lower in the rTMS group. Again, a major shortcoming of this study was that it included patients with medication overuse and taking preventive antimigraine drugs.
Finally, in a proof-of concept study based on the neurophysiological evidence that the sensory cortices are hyperexcitable in CM, inhibitory quadripulse rTMS was applied over the visual cortex twice per week for one month in 12 CM patients. The treatment significantly decreased headache and migraine days, total headache duration, headache intensity and acute medication use. Six out of 12 patients (40%) had a ≥50% reduction in migraine days and nine patients (60%) reversed from CM to EM (11).
To summarize, rTMS could thus be effective in migraine prophylaxis but it is difficult to implement in clinical practice because of the non-portability of the devices and their high cost.
Transcranial direct current stimulation (tDCS) (Table 3)
tDCS studies as preventive treatment.
CM: chronic migraine; M: migraine patients; MA: migraine with aura; MO: migraine without aura; N.S.: not specified; PT: phosphine threshold; tDCS: transcranial direct current stimulation; VAS: visual analog scale.
tDCS has over rTMS the advantage that it can be applied with devices that are inexpensive and portable.
In two studies, repeated cathodal tDCS, i.e. an inhibitory current, was applied over the visual cortex without significant change in migraine attack frequency. The first was a randomized, sham-controlled trial in 26 migraine patients (12 MO, 14 MA) who underwent three daily 15-minute sessions/week of active (n = 15) (cathode at Oz, anode at Cz) or sham tDCS (n = 15) for three weeks. Neither tDCS nor sham stimulation significantly reduced migraine attacks or duration at two months (12). In the second randomized, double-blind, parallel-group controlled pilot trial, Rocha et al. (16) administered 12 20-minute sessions (three times a week for four weeks) of cathodal or sham tDCS over the visual cortex in 15 migraine patients (10 real, 5 sham) and likewise found no difference between sham and verum in change of attack frequency, duration or intensity.
By contrast, in a proof-of-concept study based on the rationale that between attacks the visual cortex of migraine patients has a reduced pre-activation level, but an increased responsivity with repeated stimulation (see Coppola et al. (1) this issue), anodal tDCS, thus an activating stimulation, over the occiput was beneficial in EM patients (15). In 10 MO patients who underwent 10 15-minute sessions (twice a week) of anodal tDCS over the visual cortex for eight weeks, there was a significant reduction in attack frequency (−38%), migraine days (−48%), attack duration (−60%) and acute drug intake (−28%) at month 2 of follow-up, i.e. one month post-treatment. Besides being open and uncontrolled, this study suffers from the inclusion of some patients taking preventive medications and the results need to be confirmed in a randomized sham-controlled trial (15).
Two small studies using anodal tDCS over the primary motor cortex M1 reported a positive outcome. In 13 CM patients treated with 10 20-minute sessions of anodal (N = 8) or sham (N = 5) tDCS over M1 for four weeks, a significant reduction of headache intensity was found after four months in the real tDCS group and a trend for reduced migraine attack duration (13). Auvichayapat et al. treated 37 EM patients with anodal (N = 20) or sham (N = 17) tDCS for 20 minutes over 20 consecutive days. They observed significant reductions favoring tDCS over sham at 4, 8, and 12 weeks for attack frequency, headache intensity and number of abortive medications taken (14).
To sum up, tDCS can thus be considered a more interesting therapeutic option than rTMS because it is portable, more user-friendly and affordable.
Transcutaneous peripheral (cranial) nerve stimulation
Neurostimulation of pericranial nerves with subcutaneous implanted electrodes has been explored as therapy for primary headaches since more than 30 years (see Ambrosini et al. this issue). Percutaneous suboccipital nerve stimulation (ONS) was found useful for CM in several non-controlled studies and case reports, but the global effect in three sham-controlled studies was far from being convincing (17–19), and adverse effects were rather frequent. These methods should thus be restricted to the most-disabled CM patients. Technological advances during the last decade have led to the development of noninvasive, transcutaneous stimulators of peripheral nerves that can be used also in less-disabled patients and are deemed to have comparable efficacy to the minimally invasive methods.
Trigeminal (and suboccipital) stimulation with the Cefaly® device
A novel transcutaneous supraorbital electrostimulation device, the Cefaly® (Cefaly Technology, Liège, Belgium), has recently been approved for the preventive treatment of migraine by the United States Food and Drug Administration. This decision was chiefly based on the results of an investigator-initiated, randomized, double-blind, sham-controlled trial (PREMICE) coordinated by the Belgian Headache Society (20). The study included 67 migraine patients with at least two migraine attacks per month. After a one-month baseline period, patients were randomized to a verum or a sham stimulator for three months and asked to apply one 20-minute stimulation session per day. At the end of this treatment period, there was a greater reduction in the average monthly number of migraine days in the verum group (from 6.9 to 4.8) than in the sham group (from 6.5 to 6.2) and a higher percentage of 50% responders (38.2% vs 12.1%) with a therapeutic gain of 26.1%. This was accompanied by a reduction in the mean number of triptans taken per month in the verum group. The effect size was considered to be comparable to that of several preventive antimigraine drugs (20). It was inferior to that reported in the pooled placebo-controlled topiramate trials, but was not accompanied by any cumbersome side effects nor any drop-outs, while in the topiramate trials 50% of patients had side effects that led to a drop-out rate of one out of four. Additional statistical analyses (21) showed that age and disease duration did not affect outcome, but that the effect size was directly related to the number of migraine days during baseline. This suggests that Cefaly® might be more beneficial for patients with more frequent migraines. Trials on its effect in chronic migraine patients are underway (ClinicalTrials.gov identifier: NCT02342743).
In a post-marketing survey of 2313 triptan participants who rented the Cefaly® over the Internet for six weeks after which they had to buy it or send it back, 53.4% found it effective and purchased it (22). Surprisingly, out of the 46.6% dissatisfied participants who sent it back, only 40% used it for the recommended time, which suggests that the true (“per protocol”) non-responder rate could be as low as 18.6%. In the survey only 4.3% of patients reported at least one adverse event during treatment with the Cefaly®, none of them serious and all reversible. The most frequent was intolerance to the paresthesias induced by the electrical stimulation, which led in some cases to interruption of treatment. Other adverse events included drowsiness during treatment and reversible skin irritation under the forehead electrode. The efficacy (50% responder rate of 75%) and excellent tolerability of Cefaly® was recently confirmed in a small, open-label, 60-day trial on 24 MO patients (23).
In pilot trials with the Cefaly® both its effect on attack prevention and acute attack treatment was examined (24). In 10 patients a total of 30 attacks were treated with a stimulation protocol that differed from that used for prevention mainly by a higher stimulation frequency (100 Hz). Within 30 minutes the Cefaly® provided total migraine relief without rescue medication in 13% of attacks and partial relief with rescue medication in 45%. Randomized trials of the efficacy of Cefaly® as an acute migraine treatment are in progress (ClinicalTrials.gov identifier NCT02411513).
The Cefaly® device can also be used to deliver transcutaneous suboccipital neurostimulation (tONS). Preliminary pilot studies suggest that tONS with the Cefaly® might be effective in CM patients (ClinicalTrials.gov identifier NCT02307071) (25).
To summarize, supraorbital neurostimulation with the Cefaly® is effective as preventive treatment in EM and has excellent tolerability and a favorable safety profile (see Riederer et al. for a review (26)). It might also be useful for acute attack treatment and for chronic migraine, but these effects remain to be confirmed in sham-controlled trials. As with other transcutaneous neurostimulation devices, adherence to treatment may be problematic in the long term.
Vagus nerve stimulation (VNS) with the gammaCore®
Summary of recommendations and levels.
rTMS: repetitive transcranial magnetic stimulation; sTMS: single-pulse transcranial magnetic stimulation; tDCS: transcranial direct-current stimulation; VNS: vagus nerve stimulation.
Overall, large sham-controlled trials are needed in order to validate transcutaneous VNS as an effective acute and/or preventive treatment of migraine.
Conclusion
Noninvasive neurostimulation methods are promising for the treatment of EM and CM because of the available data on clinical efficacy and their excellent tolerability and safety. For some of them the efficacy-side effect profile may thus be better than for certain pharmacological treatments. There is strong evidence that transcutaneous pericranial nerve stimulation with the Cefaly® is effective in the prevention of EM; the favorable safety profile remains to be confirmed in long-term studies; the results of controlled trials in CM and in acute treatment are eagerly awaited. tVNS with the gammaCore® lacks at present evidence for efficacy in acute or preventive treatment of migraine. Beneficial effects of attack treatment were found in open-label trials.
By contrast, the transcranial magnetic (TMS) and direct current (tDCS) neurostimulation techniques allow us to target specific cortical areas and to influence them in opposite directions. The rationale for their use in migraine is thus of the uttermost importance and conditions treatment outcome. The stimulation protocol and site are crucial, as cortical changes vary not only over the migraine cycle, but also with attack frequency and reversal from EM to CM (32,33). There is at present evidence for an effect on migraine headache of single-pulse TMS over the occipital cortex in MA and possibly in MO, but the overall effect size seems small. Repetitive TMS over the dorsolateral prefrontal cortex, the motor cortex or the visual areas might be effective in EM and CM depending on the stimulation protocol, but large sham-controlled trials are lacking. Although it can induce long-lasting changes of the underlying cortex, rTMS is difficult to implement in clinical practice because of the cost and non-portability of the devices. tDCS does not have these handicaps, since the devices are not expensive and portable. Controlled studies based on the rationale that in migraine the cerebral cortex is hyperexcitable and hence using cathodal tDCS inhibition of the visual areas found no significant therapeutic effect. This contrasts with positive proof-of-concept studies reasoning that the cortex is not hyperexcitable, but hyper-responsive in EM between attacks, and that the visual areas should thus be activated with anodal tDCS. A sham-controlled trial, however, confirming this beneficial effect in EM prevention is not yet available. Finally, anodal tDCS over the primary motor cortex was found effective for the preventive treatment of EM and CM in small, placebo-controlled trials.
There is little doubt that in the near future evidence-based data will be accruing for noninvasive neurostimulation in the various migraine types, but also that novel devices and stimulation protocols will improve their efficacy.
Article highlights
Noninvasive transcutaneous and transcranial neurostimulation methods seem promising treatments for migraine based on available efficacy data but more randomized sham-controlled trials are needed. Thanks to their excellent tolerability and safety, noninvasive neurostimulation methods can be combined with drug treatments and the potential benefit of combining them between each other merits to be explored. Transcranial magnetic and direct-current stimulation can selectively influence brain areas while transcutaneous supraorbital or suboccipital stimulation seems to modulate pain control centers. They allow us therefore to target specific facets of migraine pathophysiology.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: JS and DM are supported by funding from the European Union’s Seventh Framework programme (2007–2013) under grant agreement no. 602633. GC is supported financially by the Italian Ministry of Health and Fondazione Roma.