Neuromodulation Therapies for Headache

Neuromodulation is the functional modification of neural circuits with electromagnetic current, delivered noninvasively or via implanted devices. Considerable progress has been made in using neuromodulation as a treatment option for certain headache disorders, particularly cluster headache (CH) and migraine. Although neuromodulation is typically not considered first-line treatment for headache disorders yet, some forms of noninvasive neuromodulation are approved by the Food and Drug Administration (FDA) for migraine and/or cluster headache.

Noninvasive Neuromodulation

Noninvasive neuromodulation techniques with potential benefit for treating headaches include noninvasive vagus nerve stimulation (nVNS), transcranial magnetic stimulation (TMS), external trigeminal nerve stimulation (eTNS), and peripheral electrical stimulation (PES) (Table 1).

Vagus Nerve Stimulation

Although the mechanism of action for VNS to treat headache disorders is unclear, VNS is thought to act on central pain centers via the nucleus tractus solitarius and input from the vagus nerve.¹ As adjunctive treatment, 20 minutes of nVNS stimulation daily results in a significant reduction in weekly CH attack frequency after 4 weeks.² Benefits of treatment may be seen as early as 2 weeks and have been seen to continue for several weeks in open-label extension studies.^2,3

Treatment with nVNS is FDA-approved for acute treatment of CH. In randomized sham-controlled double-blind studies, participants with episodic CH had higher response rates with active vs sham stimulation. There was, however, no significant difference in the primary endpoint overall responder rate for pooled results from participants with episodic and chronic CH treated with active stimulation vs sham stimulation.^4,5

In an open-label study of nVNS for acute self-treatment of 3 migraine attacks in a 3-week period, 64.6% of people with chronic migraine (CM) or high-frequency episodic migraine (HFEM) had pain relief within 2 hours., and 39.6% of responders had freedom from pain. Those with HFEM had more response than those with CM.⁶ In a randomized sham-controlled study of 248 people with episodic migraine (EM) with or without aura, nVNS was superior to sham for acute treatment as early as 30 and 60 minutes after onset, although no there was no difference 120 minutes after pain onset.⁷

In nVNS studies, safety events were minimal and interestingly, occurred more often in sham groups than nVNS groups. Typical stimulation-related adverse events included lip or face pulling/twitching, dizziness, neck pain, burning, and soreness.

Given its better safety profile, nVNS has more therapeutic potential than invasive stimulation (see below). An nVNS device, gammaCore is available from electroCore and is FDA-approved for acute treatment of CH and migraine in adults as well as adjunctive preventive treatment of CH in adults.

Transcranial Magnetic Stimulation

The mechanism of action of TMS for headache is thought to be disruption of cortical spreading depression (CSD) and modulation of thalamic nociceptive circuit activity via downregulation of thalamocortical neurons.⁸ Although repetitive TMS (rTMS) has been hypothesized as an effective treatment for migraine, optimal stimulation parameters have yet to be determined and results for treating headache disorders with rTMS have thus been mixed. In contrast, single pulse TMS (sTMS) has been shown effective for both acute and preventive treatment of migraine. Active sTMS provided freedom from pain after 2 hours for almost twice as many participants vs sham stimulation (39% vs 22%). No serious adverse events were observed. Active vs sham stimulation provided more freedom from pain 24 and 48 hours after treatment (24 hours, 29% vs 16%; 48 hours, 27% vs 13%).

In a survey of people with migraine who used sTMS over 3 months, there was efficacy for participants with EM or CM with and without aura.⁹ A preventive protocol (4 pulses twice daily plus 3 pulses repeated up to 3 times during any attack that occurred) reduced headache days per month by 2.75, and 46% of participants had reduced headache intensity and reduced acute headache medication use.¹⁰

Typical side effects associated with sTMS include tingling, tinnitus, and lightheadedness without any serious adverse events reported in clinical trials. As with nVNS, compared to invasive neuromodulation, sTMS has a much better side effect profile. Handheld devices for acute, preventive, or combined sTMS protocols are available from eNeura and FDA-approved for acute and preventive treatment of migraine.

External Trigeminal Nerve Stimulation

Transcutaneous eTNS is delivered with a device worn over the forehead, designed to stimulate the bilateral supratrochlear and supraorbital nerves. Some of the first evidence for treating headache with stimulation to this pathway came from a study showing the combination of occipital nerve stimulation and supraorbital nerve stimulation was more efficacious for treating migraine than occipital nerve stimulation alone.¹¹

For migraine preventive treatment, 20 minutes of active stimulation daily for 3 months resulted in reduced mean number of headache days with a 50% responder rate of 38.1% vs 12.1% for active vs sham stimulation.^12,13 For acute migraine treatment, 1 hour of eTNS given in an outpatient clinic within 3 hours of onset resulted in 57.1% pain reduction after 1 hour and 52.8% after 2 hours. In addition, in the 24 hours after intervention, only 34.6% of patients had used rescue medication.¹⁴

The most frequently reported adverse effects of eTNS include intolerance due to paresthesia, and sleep-wake disturbances, but no serious adverse events have been reported. An eTNS device produced by Cefaly is FDA approved for acute and preventive treatment of migraine.

Peripheral Electric Stimulation

In PES, electrodes placed on the upper arm generate a nonpainful electric stimulus thought to exert a generalized analgesic effect through conditioned pain modulation via activation of inhibitory pain pathways. In a randomized sham-controlled trial (n = 71), 64% of participants who had 20 minutes of PES soon after migraine attack onset had at least 50% pain reduction in more than half of their treated attacks, compared with 26% of the participants who received sham stimulation.¹⁵ A prospective randomized sham-controlled multicenter trial of this device has also shown early promising results. Discomfort is the most commonly reported adverse effect. No serious adverse events were reported in these trials. A noninvasive PES device (Nerivio Migra), modeled on this principle, has electrodes placed on the upper arm that generate a nonpainful electrical stimulus that can activate the inhibitory pain circuit. An application for FDA approval for migraine acute treatment has been submitted for this investigational device.

Caloric Vestibular Stimulation

Caloric vestibular stimulation (CVS), used for evaluation of balance disorders and brainstem function, is being studied for migraine preventive treatment. Thought to act through modulation of brainstem circuits, CVS is delivered via a headset that inserts a stimulator into the external ear canal. In a multicenter sham-controlled study using a CVS device (TNM device; Thermoneuromodulation, Scion NeuroStim), participants treated with CVS for 3 months had 2.8 fewer migraine days per month compared with participants given sham stimulation.¹⁶

Percutaneous Mastoid Electric Stimulation

In percutaneous mastoid electrical stimulation (PMES), electrodes placed over the mastoid area behind each ear are thought to inhibit nociceptive circuits in migraine by modulating activity of the trigeminocervical complex and/or fastigial nucleus. In a randomized sham-controlled trial in persons with EM, 82.5% of participants had a 50% or more reduction in migraine days compared to 17.5% of participants who received sham stimulation.¹⁷

Invasive Neuromodulation

Deep Brain Stimulation

Studies using functional neuroimaging showed hypothalamic activation in cluster headache (CH)¹⁸ and other trigeminal autonomic cephalalgias (TACs), paving the way for the deep brain stimulation (DBS) in the hypothalamus,^19-21 especially for the 15% to 20% of people who have chronic CH refractory to medical treatment.²²

There have been promising results reported as cases and case series, altogether suggesting that 50% to 60% of participants receiving DBS had at least a 50% reduction in headache intensity or frequency.^22-24 A single crossover double-blind sham-controlled study of 11 patients did not show a statistically significant benefit of DBS after 1 month; however, during the 1-year open-label extension study, 6 of 11 participants had more than 50% reduction in pain and 3 of those individuals achieved freedom from pain.²⁵ Although these studies have primarily targeted the posterior hypothalamus with DBS, other targets have been suggested and tested including the junction between the diencephalic and mesencephalic tegmentum.^26-28

The main side effects of DBS include transient visual disturbances, diplopia, vertigo, nausea, euphoria, bradycardia, and appetite changes; most improve once stimulation parameters are adjusted. There is a small risk of more serious morbidity and mortality including intracerebral hemorrhage, hardware infection, transient loss of consciousness, and skin erosion or hardware malfunction requiring reimplantation.²⁹

Overall, DBS is a promising treatment modality for refractory chronic CH, although larger randomized trials are needed to better determine efficacy and target location (Table 2).

Occipital Nerve Stimulation

Another neuromodulation approach is occipital nerve stimulation (ONS), thought to work via peripheral and central mechanisms, including:

• Peripheral stimulation of large sensory afferents that likely inhibit nociceptive activity in small fibers
• Centrally normalization of pain processing centers that were hypermetabolic before stimulation.³⁰

After promising results in case series of ONS for occipital neuralgia and chronic cluster headache, larger studies were conducted that, together, included 105 participants treated with ONS for chronic CH. Participants had overall pain improvement of 52.9% to 70% and significant improvements in attack frequency and functional/emotional scale scores. Use of preventive medications was reduced.^31-33 Some suggest that these results and the better side effect profile of ONS make it a more appropriate first choice for neuromodulation for CH.

For migraine, ONS has shown positive results in observational studies and trends toward reducing the number of headache days per month, but these have not been statistically significant compared with sham stimulation.³⁴ Adjustable ONS seems to be more effective than preset ONS (39% vs 6% response).³⁵ Reduction in pain, headache days, and migraine related disability have also been seen with ONS.³⁶ Long-term effectiveness for migraine is still being investigated; meta-analysis suggests that the the continuation rate drops to 50% to 60% after 1 year.³⁷ Limitations of ONS studies include the difficulty blinding participants to treatment because paresthesia is part of the mechanism of neuromodulation (Table 2).

Although potential complications associated with ONS are generally less severe than with DBS, there are still significant side effects, including lead migration, battery replacement, site infection, local pain and numbness, and rare intraabdominal bleed at the site of the battery.³³

Sphenopalatine Ganglion Stimulation

Under investigation for CH, sphenopalatine ganglion (SPG) stimulation is thought to act through inhibition of postganglionic parasympathetic outflow and modulation of the trigeminal nucleus caudalis.³⁸

In a randomized sham-controlled trial, participants who received full active stimulation had more pain relief at 15 minutes and reduced cluster attacks per week by 88% on average compared with shamstimulation.³⁹ After 24 months, SPG stimulation was an effective acute therapy for 45% of participants, and 33% had decreased frequency of attacks, leading investigators to conclude that benefits of SPG stimulation were due to the actual stimulation and not the implantation procedure itself.⁴⁰ In a prospective open-label study 68% of patients had improvement, with 55% having a 50% reduction in attack frequency at 1 year and 32% of patients experiencing acute pain relief in at least 50% of their cluster attacks (Table 2).⁴¹

Similar to ONS and DBS, the potential complications with SPG stimulation include lead migration, lead misplacement, and infection, with 18.8% of these adverse events requiring surgical revision. In addition, up to 81% of patients experience sensory disturbances that typically resolve around 3 months after implantation, and 33% of patients developed contralateral attacks during the 24 months following implantation.⁴⁰

Clinical Considerations

Rapid development of noninvasive neuromodulation devices has been occurring and presents the potential for paradigm shift in headache medicine. There are 3 FDA-approved noninvasive neuromodulation devices (nVNS, TMS, e-TNS) for preventive and/or acute treatment of primary headache disorders. Several others have shown promise in early clinical trials . Wider use of neuromodulation devices has been limited primarily due to lack of coverage by majority of the third-party payers. During the past year, however, some commercial payers have started covering some of these devices, albeit in a restrictive way (eg, requiring failure of 2 to 3 categories of oral migraine medications first).

Patient Selection

Because of the coverage restrictions in place, a typical person to consider for neuromodulation therapy is someone who is intolerant to oral medications due to side effects, has headache refractory to oral treatment, or has contraindications to oral medications. With the evolving payer/coverage landscape, this may soon change as neuromodulation therapy becomes considered as a first- or second-line therapy option.

Device Selection

For acute treatment of migraine with aura, we typically prefer sTMS, which has the best data for this patient subgroup. For acute treatment of migraine without aura, any of the 3 FDA approved devices (Table 1) is a reasonable option. For migraine preventive therapy (with or without aura), either sTMS or eTNS is a reasonable first-line option, and other devices may soon be approved for this indication. Patient preference plays a role in selecting a particular device for a therapeutic trial. Some devices are smaller and easier to carry and some are more prominent during use than others. At this time, invasive neuromodulation has limited role in the treatment of migraine.

For people with CH, nVNS is the only noninvasive neuromodulation option currently available. For those with CH refractory to treatment with oral medications and/or noninvasive neuromodulation, invasive options such as SPG stimulation, ONS, or DBS can be consideration on a case-by-case basis.

Conclusion

Neuromodulation is an exciting and promising development in the field of headache medicine. Invasive neuromodulation options have limitations due to safety, tolerability, and lack of generalizable data. The noninvasive devices, in contrast, have the potential to fundamentally change the future therapeutic conversions. Although, payer coverage remains a concern, ongoing innovation for cost reduction and payer coverage efforts hold promise to make these devices mainstream.