Summary of key evidence on transcutaneous electrical stimulation of the supraorbital nerve for treating and preventing migraine

Study 1 Kuruvilla DE (2022)

Study details

Study type	Double-blind randomised sham-controlled trial (ACME study NCT03465904)
Country	USA (multi-centre)
Recruitment period	April 2018-January 2019
Study population and number	n=538 patients with episodic migraine with and without aura (n=299 verum eTNS versus n=279 sham eTNS)
Age and sex	Mean age 41.14 ± 12 years; 82% female (443/538)
Patient selection criteria	Inclusion criteria: patients aged 18 to 65 years; more than 1-year history of migraine with or without aura according to ICHD, hemiplegic and brainstem migraine; onset before 50 years; having between 2 to 8 moderate or severe migraine attacks (grade 2 or 3) per month or 2 months before screening. Exclusion criteria: difficulty distinguishing migraine attacks from tension-type headaches; more than 15 headache days per month; supraorbital nerve blocks or Botox treatment in the prior 4 months; other primary headaches, rare tension-type headaches; secondary headache disorders including medication overuse headache; use of recreational or illicit drugs; drug/opioid or alcohol abuse or dependence; device in the head; cardiac pacemaker or defibrillator; previous experience with the Cefaly® device; migraine aura without headache; recent participation in a study, inability to use device.
Technique	Neurostimulation was applied with an e-TNS device (CEFALY® Technology) for a 2-hour continuous session. Verum and sham devices used identical rectangular biphasic symmetrical pulses of 250 microseconds, with a width that induced paraesthesia. The sham device provided low frequency pulses of 3 Hz, while the verum device produced high frequency pulses of 100 Hz. During stimulation, the intensity increases linearly to reach a maximum of 16 mA after 14 min and remains consistent for the remainder of the treatment (106 min). Total dose of current delivered during a 2-hour session was 2.728 C.
Follow up	2 months
Conflict of interest/source of funding	Some authors reported receiving travel, research grants, consulting fees honoraria, and personal fees from different companies, and 1 author is the Global Director of Medical Affairs for CEFALY Technology. The study was funded and partly designed by Cefaly Technology.

Follow-up issues: 607 patients were randomised (n=299 in the verum group and n=308 in the sham group) and 11.2% (68/607) were excluded from the ITT analysis. 1.9% (12/607) did not return the migraine diary, 2.3% (14/607) withdrew prematurely, 3.7% (23/607) did not experience a qualifying migraine during the study and 2.9% (18/607) were lost to follow up.

Study design issues: Prospective RCT was done across 10 centres. Patients were randomised 1:1 to either the verum or sham group. They were provided with a device and a diary to keep for 2 months to be used on a qualifying migraine attack (defined as migraine with moderate or severe headache intensity, and with at least 1 MAS [photophobia, phonophobia, nausea, and/or vomiting]). Patients were instructed to self-administer the e-TNS treatment within 4 hours of migraine onset or awakening with a migraine headache. Migraines occurring within 48 hours of a prior migraine did not qualify for e-TNS treatment. Patients were also advised not to start e-TNS treatment for a migraine spontaneously resolving, and not to take any acute migraine medications before or during e-TNS therapy. Patients could use acute migraine medication after a 2-hour e-TNS treatment.

During the acute migraine phase, patients reported data in their headache diary before e-TNS session, 2 hours and at 24 hours after e-TNS treatment. Data recorded were baseline migraine pain severity (no pain in 1, mild in 1, moderate in 59.7% [321/538], or severe pain in 40% [215/538]); presence of MAS (photophobia in 94.1%, phonophobia in 80.9%, nausea in 63.4%, and/or vomiting in 8.7%); MBS; adverse effects and intake of acute antimigraine medication. 41.6% (224/538) of patients reported migraine with aura. Data were analysed on an ITT and per-protocol basis.

Primary outcomes included pain freedom at 2h and resolution of MBS at 2 hours after e-TNS for 1 migraine attack. Secondary outcomes were (1) pain relief at 2 h (defined as a reduction of a moderate or severe migraine headache to mild or no headache), (2) resolution of any MAS at 2 hours after e-TNS, (3) sustained pain freedom (defined as pain freedom at 2 hours and 24 hours without the use of antimigraine medication), (4) sustained pain relief at 24 hours (defined as mild or no headache at 2 hours and 24 hours after eTNS without the use of antimigraine medication), (5) use of a rescue medication between 2–24 hours after e-TNS.

The study was adequately powered based on calculated assumptions.

Study population issues: There was no statistically significant difference in baseline characteristics between patients in the verum and sham groups.

Other issues: 6 patients over the age of 65 (1 in the verum group and 5 in the sham group) were mistakenly included in the ITT analysis.

Key efficacy findings

Number of patients analysed: 538 (259 verum eTNS versus 279 sham eTNS)

Outcome measures (ITT analysis)

Variable	Verum % of patients (n=259)	Sham % of patients (n=279)	Total % of patients (n=538)	X^2-statistic	Effect size	P value between two groups
Pain freedom at 2 hours	25.5 (66/259)	18.3 (51/279)	21.7 (117/538)	4.10	0.18	0.043
Absence of MBS at 2 hours	56.4 (146/259)	42.3 (118/279)	49.1 (264/538)	10.65	0.28	0.001
Pain relief at 2 hours	69.5 (180/259)	55.2 (154/279)	62.1 (334/538)	11.67	0.30	0.001
Absence of all migraine-associated symptoms (MAS) at 2 hours	42.5 (110/259)	34.1 (95/279)	38.1 (205/538)	4.04	0.18	0.044
Use of rescue meds at 2 to 24 hours	31.7 (82/259)	37.6 (105/279)	34.8 (187/538)	2.11	-0.12	0.146
Pain freedom at 24 hours	22.8 (59/259)	15.8 (44/279)	19.1 (103/538)	4.26	0.18	0.039
Pain relief at 24 hours	45.9 (119/259)	34.4 (96/279)	40.0 (215/538)	7.45	0.24	0.006

Key safety findings

Rate of adverse events (ITT group)

Adverse event	Verum % of patients (n=259)	Sham % of patients (n=279)	Total % of patients (n=538)	P value between groups
Forehead paraesthesia, discomfort or burning	3.5 (9/259)	0.4 (1/279)	1.9 (10/538)	0.009
Nausea/vomiting	1.5 (4/259)	0	0.7 (4/538)	0.053
Dizziness	0.4 (1/259)	0.7 (2/279)	0.6 (3/538)	1.000
Neck stiffness/muscle tension	0.4 (1/259)	0.7 (2/279)	0.6 (3/538)	1.000
Worsened headache	0.8 (2/259)	0	0.4 (2/538)	0.231
Orodynia—tooth or jaw pain	0.4 (1/259)	0.4 (1/279)	0.4 (2/538)	1.000
Restlessness	0.4 (1/259)	0.4 (1/279)	0.4 (2/538)	1.000
Abdominal discomfort or cramping	0	0.7 (2/279)	0.4 (2/538)	0.500
Dry mouth	0.4 (1/259)	0	0.2 (1/538)	0.481
Excessive sweating	0.8 (2/259)	0	0.4 (2/538)	0.231
Sedation/sleepiness	0.4 (1/259)	0	0.2 (1/538)	0.481
All	8.5 (22/259)	2.9 (8/279)	5.6 (30/538)	0.004

Study 2 Chou DE (2019)

Study details

Study type	Double-blind randomised sham-controlled trial (AMCE study NCT02590939)
Country	USA (multi-centre)
Recruitment period	February 2016- March 2017
Study population and number	n=106 (52 verum eTNS versus 54 sham eTNS) Indication: acute treatment in patients with episodic or chronic migraine with or without aura Migraine with aura: eTNS=12, sham=5; migraine without aura: eTNS=40, sham n=49 Migraine attack duration: 6.00 hours [range 4.00–24.00] No of patients on medication use: eTNS=17 versus sham =14
Age and sex	Mean 39.90 ± 13.14 years; female 87% (92/106)
Patient selection criteria	Inclusion criteria: patients 18-65 years old, episodic or chronic migraine attack with or without aura (ICHD-3 criteria), with headache lasting for at least 3 hours and pain intensity stable for at least 1 hour; and may have used any acute medications 3 hours before enrolment. Exclusion criteria: use of botulinum toxin, supraorbital nerve blocks, opioids medication in past 4 months, diagnosis of other primary or secondary headache disorders; headache location not involving the frontal, retro- or peri-orbital regions; forehead skin allodynia; intake of acute migraine medication within the 3 hours before enrolment; implanted metal or electrical devices in the head; cardiac pacemaker or implanted or wearable defibrillator; or previous experience with e-TNS.
Technique	Sham and eTNS (with a Cefaly device) treatments were delivered for 1 hour session. eTNS group: the electrical pulses are transmitted transcutaneously via a supraorbital bipolar self-adhesive electrode placed on the forehead covering the supratrochlear and supraorbital nerves bilaterally. Impulses were delivered at pulse frequency 100 Hz, pulse width 250µs, maximum current dose 1.284 C and maximum intensity of 16 mA. Sham stimulation group: the sham device was identical in shape, pulses were delivered at low-frequency of 3 Hz, width 250µs and duration of 1 hour.
Follow up	24 hours
Conflict of interest/source of funding	Some authors received travel, research grants, honoraria, personal fees from different companies. The study was funded by Cefaly Technology.

Analysis

Follow-up issues: 9.6% (3/52) in the eTNS group and 3.7% (2/54) in the sham group were lost to follow up. Three patients (2 in active eTNS treatment group and 1 in sham group) could not bear the paraesthesia and stopped treatment within 5 minutes, 4 patients (3 in the eTNS group and 1 in sham group) withdrew from the study and stopped the stimulation before 1 hour. 99 patients completed the study and provided pain scores at 1 hour.

Study design issues: study was done in 3 headache clinics; sample size was small, adequately powered based on calculated assumptions. During recruitment phase, patients meeting inclusion criteria were randomised 1:1 to receive eTNS or sham stimulation. Those who could not tolerate the stimulation during the first 4 minutes stopped taking part in the trial. Treatment allocation was concealed; patients and staff were blinded from randomisation. Devices and electrodes were identical, programmed by manufacturers, coded and randomly distributed to 2 treatment groups in blocks of 4.

Primary outcome measure was the mean change in pain intensity at 1 hour compared with baseline. Pain intensity was scored using a VAS (0 to 10, with higher scores indicating more pain) at baseline, 1 hour, 2 hours and 24 hours after stimulation. Migraine rescue medications were not allowed before 1 hour and during the 1 hour treatment period but were allowed after this period and medication use was recorded after 2 and 24 hours. Data were analysed on an ITT and per-protocol basis.

Study population issues: patients in both groups had unbalanced baseline demographic characteristics. Baseline migraine symptoms were not reported.

Other issues: authors state that patients were not given any advice after eTNS session about acute migraine medication use.

Key efficacy findings

Number of patients analysed: 106 (52 eTNS versus 54 sham) in the ITT analysis.

Outcome measures (ITT analyses)

	eTNS (mean ± SD) n=52	Sham (mean ± SD) n=54	Comparison between the 2 groups
VAS scores
Baseline	5.92 ± 1.68	6.17 ± 1.81
1 hour	2.46 ± 2.23	4.39 ± 2.44
2 hour	3.06 ± 2.30	4.31 ± 2.51
24 hours	2.46 ± 2.27	3.79 ± 2.74
Mean change in pain intensity compared with baseline
1 hour (absolute value)	-3.46 ± 2.32 (p<0.0001)	-1.78 ± 1.89 (p<0.0001)	p<0.0001
1 hour (relative %)	-59% ± 35%	-30% ± 31%	p<0.0001
2 hours (absolute value)	-2.87 ± 2.24 (p<0.0001)	-1.85 ± 1.96 (p<0.0001)	p=0.028
2 hours (relative %)	-50% ± 36%	-32% ± 37%	p=0.026
24 hours (absolute value)	-3.46 ± 2.65 (p<0.0001)	-2.38 ± 2.27 (p<0.0001)	p=0.062
24 hours (relative %)	-57% ± 37%	-40% ± 40%	p=0.037
Rescue medication use % (n)
No of patients used medication at 2 hours	6 (3/52)	4 (2/54)	p=0.66
No of patients used medication within 24 hours	35 (18/52)	39 (21/54)	NS
Pain freedom % (n)
At 1 hour	29 (15/52)	6 (3/54)	p=0.0016
At 2 hours	17 (9/52)	7 (4/54)	p=0.15
At 24 hours	29 (15/52)	13 (7/54)	p=0.056
Sustained pain freedom for 24 hours	6 (3/52)	0	p=0.11
≥30% pain relief % (n)
At 1 hour	79 (41/52)	39 (21/54)	p<0.0001
At 2 hours	65 (34/52)	52 (28/54)	p=0.17
At 24 hours	73 (38/52)	61 (33/54)	p=0.22
Sustained ≥30% pain relief for 24 hours	38 (20/52)	20 (11/54)	p=0.055
≥50% pain relief % (n)
At 1 hour	63 (33/54)	31 (17/54)	p=0.0017
At 2 hours	54 (28/52)	41 (22/54)	p=0.24
At 24 hours	60 (31/52)	48 (26/54)	p=0.25
Sustained ≥50% pain relief for 24 hours	29 (15/52)	15 (8/54)	p=0.10

Outcome measures (modified ITT analysis -patients who completed the 1-hour treatment phase [n=99, eTNS 47 versus 52 sham])

	eTNS n=47	Sham n=52	Comparison between the 2 groups
Mean change in pain intensity compared with baseline	-65%	-32%	p<0.0001
Pain freedom at 24 hours %	32%	13%	0.032
≥30% sustained pain relief %	43%	21%	0.030

Sub-group analysis

Pain reduction at 1 hour (VAS score)	eTNS (mean± SD) n=47	Sham (mean± SD) n=52	Comparison between the 2 groups
Migraine without aura attacks	-3.3 ± 2.4	-1.7 ± 1.9	p=0.0006
Migraine with aura attacks	-4.3 ± 1.8	-2.6 ± 1.9	0.060
Patients who did not use any acute medications	-3.6 ± 1.7	-1.8±1.9	p<0.0001

Key safety findings

Adverse events	eTNS group n=52	Sham group n=54
Unable to tolerate paraesthesia sensation (treatment stopped within 5 minutes), n	2	1
Nausea during stimulation (resolved without treatment after 20 minutes), n	1	0

Adverse events

eTNS group

n=52

Sham group

n=54

Unable to tolerate paraesthesia sensation (treatment stopped within 5 minutes), n

Nausea during stimulation (resolved without treatment after 20 minutes), n

Study 3 Chou DE (2017)

Study details

Study type	Prospective case series (NCT02411513)
Country	USA
Recruitment period	April 2015 to October 2015
Study population and number	n=35 Indication: acute treatment in patients with episodic or chronic migraine with or without aura
Age and sex	Mean 39.4 ± 12.5 years; female 80% (24/35)
Patient selection criteria	Inclusion criteria: patients 18-65 years old, episodic or chronic migraine attack with or without aura (ICHD-3 criteria), with headache lasting for at least 3 hours and pain intensity stable for at least 1 hour, frontal-retro-peri-orbital headache and may have used any acute medications 3 hours before enrolment. Exclusion criteria: use of botulinum toxin, supraorbital nerve blocks, opioids medication in past 4 months, diagnosis of other primary or secondary headache disorders (except medication overuse), headache location not involving the frontal, retro or peri-orbital regions, forehead skin allodynia, intake of acute migraine medication within the 3 hours before enrolment, implanted metal or electrical devices in the head, cardiac pacemaker or implanted or wearable defibrillator, previous experience with e-TNS or those with complicated migraine.
Technique	eTNS (with a Cefaly device) was delivered for a 1 hour session. The electrical pulses are transmitted transcutaneously via a supraorbital bipolar self-adhesive electrode placed on the forehead covering the supratrochlear and supraorbital nerves bilaterally. Impulses were delivered at a pulse frequency 100 Hz, pulse width 250µs, maximum current dose 1.284 C and maximum intensity of 16 mA. Full stimulation intensity (16 mA) in 17 patients, limited current output (average of 9.51 mA) in 13 patients.
Follow up	24 hours
Conflict of interest/source of funding	First author received travel, research grants, consultation fees from different companies. The study was funded by Cefaly Technology.

Analysis

Follow-up issues: 5 patients were lost to follow up during test phase (1 at screening due to opioid use in past 3 months, 4 due to paraesthesia and inability to tolerate eTNS within 5 minutes-nociceptive test phase). No loss to follow up was reported during the study.

Study design issues: prospective study with small sample size at a single headache clinic. Primary outcome measure was the mean change in pain intensity at 1 hour compared with baseline. Pain intensity was scored using a VAS (0 to 10, with higher scores indicating more pain) at baseline, 1 hour, 2 hours and 24 hours after stimulation. Migraine rescue medications were not allowed before 1 hour and during the 1 hour treatment period but were allowed after this period and medication use was recorded after 2 and 24 hours. Data were analysed on a modified ITT basis.

Study population issues: baseline migraine symptoms were not reported.

Other issues: authors state that patients were not given any advice after eTNS session about acute migraine medication use.

Key efficacy findings

Number of patients analysed: 30

Outcome measures (ITT analysis)

Mean change in pain intensity compared with baseline	mean± SD
1 hour after treatment (absolute value)	From baseline 5.63 to 2.42; -3.22 ± 2.40 (p<0.001)
Relative %	57.1
2 hours after treatment (absolute value)	From baseline 5.63 to 2.66; -2.98 ± 2.31 (p<0.001)
Relative %	52.8
Pain freedom	% (n)
1 hour after treatment	20 (6/30)
2 hours after treatment	13.3 (4/30)
≥30% pain relief	% (n)
1 hour after treatment	83.3 (25/30)
2 hours after treatment	70 (21/30)
≥50% pain relief	% (n)
1 hour after treatment	76.7 (23/30)
2 hours after treatment	56.7 (17/30)
Not needing rescue medication	% (n)
2 hours after treatment	100
24 hours after treatment	65.4 (17/26)
Needed rescue medication at 24 hours	34.6 (9/26)

50% (5/10) of patients receiving partial current output used rescue medication, compared with 25% (4/16) of patients who received the full current dose, the difference was NS (p=0.23).

Key safety findings

Adverse events	eTNS, n=35
Unable to tolerate paraesthesia sensation (treatment was stopped within 5 minutes)	n=4

Study 4 Kuruvilla D (2019)

Study details

Study type	Prospective case series (NCT03217968)
Country	USA
Recruitment period	August 2017 to January 2018
Study population and number	n=59 Indication: acute treatment in patients with single moderate or severe migraine attack (grade 2 or 3) at home Migraine with aura 25% (15/48); migraine without aura 55% (33/48) Duration of migraine: >1 year
Age and sex	Mean 46.85 ± 10.2 years; female 90% (43/48)
Patient selection criteria	Inclusion criteria: patients 18-65 years old, more than 1 year history of episodic migraine attack with or without aura (ICHD-3 criteria), migraine onset before 50 years, 2 to 8 moderate migraine attacks per month in each of the 2 months before screening. Exclusion criteria: patients' difficulty distinguishing migraine from tension-type headache, more than 15 headaches per month (chronic migraine), migraine aura without headache, hemiplegic migraine and brainstem aura migraine, patients with supraorbital nerve blocks or Botox in the head in the prior 4 months, migraine prophylaxis modification in prior 3 months, other primary or secondary headache disorders (medication overuse), patients with opioid, alcohol or illicit drugs abuse, metallic or electric device in head, cardiac pacemaker, implanted or wearable defibrillator; previous experience with Cefaly, participation in other study in past 30 days, patients unable to self-service or bear the eTNS stimulation.
Technique	eTNS (with a Cefaly device) was delivered for a 2 hour session at home. The electrical pulses are transmitted transcutaneously via a supraorbital bipolar self-adhesive electrode placed on the forehead covering the supratrochlear and supraorbital nerves bilaterally. Impulses were delivered at a pulse frequency 100 Hz, pulse width 250 microseconds, maximum intensity of 16 mA for 120 minutes.
Follow up	24 hours
Conflict of interest/source of funding	Authors received travel, research grants, consultation fees from different companies. The study was funded by Cefaly Technology.

Analysis

Follow-up issues: only 48 patients were included in the analysis. 4 patients withdrew from the study initially (as 2 had rescue medication and 2 did not have migraine), 1 patient was lost to follow up and 6 patients did not report outcomes at 2 hours.

Study design issues: a small prospective study done at 1 research centre. Patients were recruited from a database and screened during a visit. Study patients reported outcomes in a headache dairy. Primary outcome measure was freedom from pain and from MBS (photophobia, phonophobia, nausea, vomiting) at 2 hours after acute treatment with eTNS. Other outcomes assessed include reduction of moderate to severe migraine headache (measured on a scale 0 to 3, with higher scores indicating severe pain), percentage of patients with absence of migraine-associated symptoms at 2, 24 hours and use of rescue medication between 2 and 24 hours. Data were analysed on a modified ITT basis.

Study population issues: patients with a migraine (moderate or severe headache pain severity grade 2 or 3) and with at least 1 migraine associated symptom were included. The headache had to have started less than 4 hours before eTNS, or the patient woke up with a migraine. Patients could also not have had another headache in the previous 48 hours. Patients were also advised not to start using the device on a migraine resolving spontaneously or if acute antimigraine medication had been taken since the beginning of the attack.

Key efficacy findings

Number of patients analysed: 59

Outcome measures (ITT analysis)

Primary outcomes	Baseline % (n)	2 hours % (n)	24 hours % (n)
Patients with moderate pain intensity (grade 2)	68.8 (33/48)
Patients with severe pain intensity (grade 3)	31.2 (15/48)
Pain freedom		35.4 (17/48)	25 (12/48)
MBS freedom		60.4 (29/48)
Patients with nausea as MBS	22.9 (11/48)
MBS nausea freedom		45.5 (5/11)
Patients with vomiting as MBS	2.1 (1/48)
MBS vomiting freedom		0 (0/1)
Patients with photophobia as MBS	56.2 (27/48)
MBS photophobia freedom		63 (17/27)
Patients with phonophobia as MBS	18.8 (9/48)
MBS phonophobia freedom		77.8 (7/9)
Secondary outcomes
Pain relief*		70.8 (34/48)
Absence of symptoms		45.8 (22/48)
Use of rescue medication at between 2-24 hours		50 (24/48)

*measured on a scale 0 to 3, with higher scores indicating severe pain.

Device compliance: 71.2% (42/59) patients were compliant with the eTNS 2 hours session.

Key safety findings

Adverse events	eTNS, n=59
Total minor adverse events (majority were reversible, uncomfortable paraesthesia: forehead sensations including burning, itching, tingling, stinging, and numbness; 4 were unable to complete the session).	n=15

Study 5 Penning SC (2020)

Study details

Study type	Observational survey (NCT02616978)
Country	Europe (Belgium, Switzerland and France)
Recruitment period	Not reported
Study population and number	n=413 Indication: patients with migraine (acute migraine n=366) Average number of monthly migraine attacks: 9.47
Age and sex	Not reported
Patient selection criteria	Inclusion criteria: patients from the Cefaly customer database who were identified as regular users.
Technique	eTNS with a Cefaly device.
Follow up	Not reported
Conflict of interest/source of funding	One author is a consultant for Cefaly technology.

Analysis

Study design issues: retrospective survey was done online using an 8-item questionnaire to collect data from users. Study participation was on a voluntary basis and responses were anonymised and analysed. Primary outcome was reduction in the mean number of acute antimigraine drug intake per month per patient. Secondary outcome measures were percentage of subjects using the device during an attack, attacks treated with reduction of acute antimigraine drugs.

Key efficacy and safety findings

Number of patients analysed: 413

Outcome measures

Mean number of acute antimigraine drugs avoided per month per patient
In total patients (n=413)	2.93
In patients with migraine attack (n=366)	3.31
Patients using the device to treat attacks, %	88.6%
Attacks treated with the device, %	71.8%
Treated attacks for which acute antimigraine drug intake is reduced, %	42.6%
Reduced intake of drugs
Triptans	54.9%
Analgesics/NSAIDs	64.9%
Other	10.7%
Patients unable to reduce acute medication intake, %	18.3%
Reasons for not using Cefaly to treat migraine attacks
I cannot bear the feeling during an attack (unbearable sensations during stimulation)	14.9%
It does not provide sufficient relief (lack of efficacy)	48.9%
I never tried	10.6%
Others	12.8%

Study 6 Beh SC (2020)

Study details

Study type	Retrospective case series
Country	USA
Recruitment period	May 2018 to June 2019
Study population and number	n=19 Indication: patients with acute VM Mean time from onset to eTNS: 8.2 ± 12.1 hours (range 15 minutes to 2 days).
Age and sex	Mean 48.1 ± 12.2 years; female 89.5% (17/19)
Patient selection criteria	Inclusion criteria: patients diagnosed with VM according to ICHD criteria (at least 5 episodes of vestibular symptoms of moderate or severe intensity lasting between 5 minutes and 72 hours, half of the episodes associated with migraine features), history of migraine with/without aura, use of no other neuromodulator therapies, rescue medications before eTNS and those on pharmacological migraine prevention, those with no Meniere's disease, otological or intracranial surgery or stroke.
Technique	eTNS (with a Cefaly device) was delivered for 20 minutes session in an out-patient setting. The electrical pulses were transmitted transcutaneously via a supraorbital bipolar self-adhesive electrode placed on the forehead covering the supratrochlear and supraorbital nerves bilaterally. Impulses were delivered at pulse frequency 60 Hz, pulse width 250 microseconds, at a maximum intensity of 16 mA.
Follow up	15 minutes after treatment and between 3 to 6 months
Conflict of interest/source of funding	No conflicts of interest or disclosures.

Analysis

Follow-up issues: all patients completed treatment.

Study design issues: a small retrospective study done at 1 Vestibular & Neuro-Visual Disorders centre. Outcomes assessed were severity of vertigo/headache (using a 10 point VAS score, with higher scores representing worst imaginable vertigo/headache) at baseline and 15 minutes after eTNS and neuro-otological examination before and after treatment.

Study population issues: patients with inadequate relief were advised to take rescue medications after eTNS treatment.

Key efficacy findings

Number of patients analysed: 19

Outcome measures

Primary outcomes	Baseline	After eTNS
Mean vertigo severity^	6.6 ± 2.1	2.7 ± 2.6
Mean improvement in vertigo severity* %		61.3 ± 32.6
Headache	73.6% (14/19)^^	57% (8/14)^^^
Mean headache severity	4.8 ± 2.4	1.4 ± 2.4
Mean improvement in headache %		77.2 ± 32.7

*All patients reported improvement of vertigo. 7 patients experienced complete resolution of vertigo, and 3 had 50% improvement in vertigo.

^8 patients reported a reduction in vertigo intensity between 25 and 50%. Of these, 5 patients suffered from constant dizziness with superimposed episodes of VM.

^^13 patients experienced improvement in headache severity, and 1 reported no change at all.

^^^8 patients reported complete resolution of headache, and 4 described partial relief (two patients with more than 50% improvement).

Improvement in other signs and symptoms (neuro-otological assessment)

Upbeat nystagmus during VM attack in 1 patient resolved after eTNS treatment. Head and eye pressure experienced by some patients during VM episodes were also improved. One patient reported improvement of chronic refractory right ear and facial pain. 2 other patients experienced resolution of ictal head and eye pressure.

Compliance to eTNS at 3- 6 months follow up (n=17)

8 patients continued to use the eTNS device for VM, 5 patients continued using pharmacological agents, and 4 were unable to afford the device.

Key safety findings

Adverse events	eTNS, n=19
Recurrence or worsening of vertigo	0

Prevention of migraine

Study 7 Schoenen J (2013, 2016)

Study details

Study type	Double-blind randomised sham-controlled trial
Country	Belgium (multicentre)
Recruitment period	2009-11 (PREMICE trial)
Study population and number	n=67 (34 tSNS versus 33 sham) Indication: migraine prevention Migraine with aura: tSNS=10, sham=10; migraine without aura: tSNS=24, sham n=23
Age and sex	Mean 36.79±10.63 years; female 91% (61/67)
Patient selection criteria	Inclusion criteria: 18-65 years old, with migraine with or without aura (ICHD-2 code 1.2 or 1.1, and at least 2 attacks per month. Exclusion criteria: use of a preventive antimigraine treatment in the previous 3 months, failure on> 3 well-established preventive drug treatments, medication overuse headache (ICHD-2 8.2), frequent chronic tension type headache ( ICHD-2 2.2/2.3) and other severe neurological or psychiatric disorders.
Technique	Sham and tSNS (with a Cefaly device) treatments were delivered with a self-adhesive electrode placed on the forehead covering the supratrochlear and supraorbital nerves bilaterally. Impulses were delivered at a frequency of 1 Hz for sham and 60 Hz for tSNS and intensity 1 mA for sham and 16 mA for tSNS. Daily sessions lasted for 20 minutes for 3 months. An intermediate visit was done after 45 days and a final visit at the end of the study.
Follow up	3 months
Conflict of interest/source of funding	First author is a consultant for ATI redwood California, advisory member for ST Jude, Allergan, and ATI; received research grants from Medtronic and Cyberonics. One author is a scientific advisor for Allergan. The study was funded by Walloon Region DH06, research convention from National Fund for Scientific Research, and a grant from the University of Liege. STX Med provided the devices.

Analysis

Follow-up issues: complete follow up.

Study design issues: small study, done by members of Belgian Headache Society in 5 tertiary headache clinics. After a run-in phase of 1 month (during which no preventive treatment was used and headache data were collected in a diary) those meeting the inclusion criteria were randomised 1:1 to receive sham or tSNS.

Treatment allocation was concealed; patients and neurologists were blinded from randomisation. Devices and electrodes were identical, programmed by manufacturers, coded and randomly distributed in blocks of 2 tSNS and 2 sham stimulators. Everyone had some degree of electrical stimulation but sensory perceptions were different.

Headache diaries recorded headache occurrence and its severity on a 4-point scale (presence of an aura, nausea, phonophobia or photophobia and acute antimigraine drug intake). A database was created from these diaries by 2 independent investigators. Data were analysed on an ITT and per-protocol basis. Migraine days not separated by at least 1 headache-free day were considered to belong to the same migraine attack.

Study population issues: patients in both groups had similar characteristics. Those in the tSNS group were slightly younger (by 4.47 years) than those in the sham group.

Other issues: Authors state that partial unblinding might have occurred in the study. They also state that patients had 4 attacks or 7 migraine days per month and are representative of most patients in the general population who need preventive treatment.

Key efficacy findings

Number of patients analysed: 67 (34 tSNS versus 33 sham)

Outcome measures (ITT analyses)

	tSNS (mean± SD)	Sham (mean± SD)	Comparison between the 2 groups
Change in monthly migraine days^
Baseline	6.94±3.04	6.54±2.61
3 months	4.88±3.46	6.22±2.99
p value	0.023	0.608	0.054~
*50% responder rate % (n)**	38.24 (13)	12.12 (4)
p value	0.023
25% responder rate % (n)**	58.8 (20)	27.3 (9)
p value	0.014
Change in monthly migraine attack frequency
Baseline	4.37±1.87	4.04±1.52
3 months	3.55±2.94	3.89±1.89
p value	0.058	0.516	0.044
Change in monthly frequency of any headache++
Baseline	7.78±4.00	6.72±2.63
3 months	5.27±3.55	6.49±3.20
p value	0.011	0.674	0.041
Severity of migraine days^
Baseline	1.96±0.46	1.78±0.41
3 months	1.80±0.60	1.73±0.53
p value	0.131	0.443	0.301
Change in monthly acute antimigraine drug intake
Baseline	11.45±8.35	9.24±4.75
3 months	7.25±7.31	9.28±5.69
p value	0.0057	0.822	0.0072

* a >50% reduction of monthly migraine days; ** >25% improvement in migraine days

~ difference becomes significant (p=0.044) when baseline migraine days are considered as a covariate

^ defined as a day with headache fulfilling ICHD-2 criteria for migraine, except for duration if treated

^^measured on a 4-point scale (0-3, 3 indicating severe pain prohibiting daily activities)

++ a headache of grade 1 severity without associated symptoms and not treated with an acute medication.

Patient satisfaction 3 months after treatment

	tSNS % (n)	Sham % (n)
Very satisfied	29.4 (10)	18.2 (6)
Moderately satisfied	41.2 (14)	21.1 (7)
Not satisfied	21.2 (7)	51.5 (17)
Not available	8.8 (3)	9.1 (3)

Device compliance (usage recorded in the device)

The mean numbers of sessions were 61.7% (55.54 sessions, 1,110 minutes) in tSNS group and 54.4% (49 sessions, 980 minutes) in the sham group. The difference between the 2 groups was not statistically significant.

Key safety findings

No adverse events or side effects occurred in either group.

Study 8 Deng Y 2020

Study details

Study type	RCT (ChiCTR-1800018257)
Country	China (multi-centre)
Recruitment period	June 2017- October 2018
Study population and number	n=90 patients with at least 2 episodic migraine attacks per month. STNS group (n=45) versus PMES group (n=45). Disease duration: 5.37 ± 3.68 years Attack type: migraine with aura 16.7% (15/90); migraine without aura 83.3% (75/90)
Age	Mean age 32.53 ± 7.85; 84.4% (76/90) female
Patient selection criteria	Inclusion criteria: patients between 18–65 years; episodic migraine with or without aura (ICHD-3 code 1.1 or 1.2); first migraine attack before 50 years; at least 1-year history of migraine; and 2 migraine attacks per month. Exclusion criteria: use of preventive treatment in the previous 3 months; failure on 3 or more preventive drug treatments; chronic migraine (ICHD-3 code 1.3) and medication overuse headache (ICHD-3 code 8.2); alcohol and/or drug abuse; severe neurological, psychiatric or primary systemic disorders including heart, brain, liver, kidney, and hematopoietic system; pregnant and lactating women; blood pressure <90/50 mmHg or >180/100 mmHg.
Technique	STNS group Stimulation electrodes were placed on the forehead, covering the supratrochlear and supraorbital nerves. The stimulus parameters were set as follows: pulse width 250 microseconds, frequency 60 Hz, peak current 16 mA. Stimulation was daily for 20 minutes for 3 months duration. PMES group: Stimulation electrodes were placed on the bilateral mastoid area behind the ear. The stimulus parameters were set as follows: pulse width 90 microseconds, frequency 1.8 kHz, peak current 10 mA. Stimulation was daily for 45 minutes for 3 months duration.
Follow up	3 months
Conflict of interest/source of funding	The authors declare that they have no conflict of interest. Authors received financial support/grants from Science and Technology Department, Medical Science Research Project and Scientific and Research Project of Health and Family Planning Committee of Sichuan Province of China of Sichuan.

Analysis

Follow-up issues: follow up was done every month until 3 months. Very few patients were lost to follow up in the STNS group (1 due to lack of efficacy and 3 due to discomfort sensations during the stimulation) and PMES group (1 due to lack of efficacy and 1 declined to return).

Study design issues: prospective study in 3 centres; after a 1 month run in phase, patients were randomised equally to 2 treatments. Blinding of treatment was not possible because of different locations of stimulation. Outcomes were assessed using migraine diaries and questionnaires. The primary outcomes were change in monthly migraine days and the 50% response rate. Statistical analysis was done by one investigator.

Study population issues: the patient demographic characteristics were not different between the groups.

Key efficacy findings

Number of patients analysed: 90

supraorbital transcutaneous stimulator (STNS) group (n=45) versus PMES group (n=45).

Outcome measures (data are expressed as mean ± SD or percent [no of patients]).

Clinical outcome	STNS (n=45)	PMES (n=45)	P value
Migraine days
Baseline, days	6.50 ± 2.07	4.71 ± 2.06
3 months, days	3 ± 0.79	1.86 ± 0.34
Change from baseline to 3 months, days	-3.5 (95% CI -4.74 to − 0.74, p=0.012)	-2.85 (95% CI -4.55 to − 0.17, p=0.027)
% decrease from baseline	53.8%	60.5%	0.88
50% responder rate (≥50% reduction in no. of migraine days/month),% (n)	62.2% (28/45)	77.8% (35/45)	0.070
Average migraine days (run-in versus average of the 3 month treatment)
Baseline, days	4.71 ± 2.06	6.50 ±2.07
3 months, days	2.62 ± 1.15	3.94 ± 1.88
P value	0.017	0.033
% decrease compared with ITT baseline	-39.4%	-44.4%	0.45
Migraine attacks frequency
Baseline, days	4.50 ± 2.59	3.29 ± 2.22
3 month, days	2.96 ± 1.67	1.86 ± 1.34
P value	0.0017	0.021
% decrease compared with ITT baseline	-34.2%	-43.5%	0.87
Migraine severity (assessed by VAS)
Baseline, days	6.33 ± 1.63	7.57 ± 0.78
3 month, days	3.50 ± 1.87	4.14 ± 2.04
P value	0.027	0.016
% decrease compared with baseline	-44.7%	-45.3%	0.65
Monthly acute antimigraine drug use
Baseline, days	6.2 ± 2.06	5.5 ± 2.83
3 month, days	2.43 ± 1.95	1.9 ± 1.67
P value	0.004	0.003
% decrease compared with baseline	-65.5%	-60.8%	0.71
Accompanying symptoms
baseline, days	2.17 ± 0.83	2.29 ± 1.13
3 month, days	1.80 ± 0.23	1.93 ± 0.31
P value	0.51	0.50
% decrease compared with baseline	-17.1%	-15.7%	0.99
Headache Impact Test-6 (assess headache related disability)
Baseline, days	63.50 ± 7.48	69.29 ± 12.98
3 month, days	40.33 ± 4.22	51.57 ± 3.55
P value	0.007	0.028
% decrease compared with baseline	-36.5%	-25.6%	0.041

There were no significant difference in compliance rates between the 2 groups (93.3 versus 88.9%, p=0.29).

Key safety findings

Adverse event	STNS % (n)	PMES % (n)	P value
Discomfort paraesthesia during simulation	13.3 (6/45)*	0	0.026

*2 patients stopped treatment because of intolerance of stimulation sensations.

Study 9 Jiang Y 2019

Study details

Study type	RCT
Country	China
Recruitment period	January 2015 to May 2017
Study population and number	n=165 patients with episodic migraine flunarizine plus e-TNS (n=51) versus e-TNS (n=51) versus flunarizine alone (n=52) Indication: migraine prophylaxis Migraine duration (mean years): tSNS 6.21, flunarizine 5.12, tSNS plus flunarizine 6.77 (p=0.18) Migraine type: migraine with aura: tSNS 16% (8/51), flunarizine 15% (8/52), tSNS plus flunarizine 18% (9/51); p=0.95 migraine without aura: tSNS 84% (43/51), flunarizine 85% (44/52), tSNS plus flunarizine 82% (42/51)
Age and sex	Mean age (years) tSNS 29.67, flunarizine 30.96, tSNS plus flunarizine 30.74 (p=0.70) Female tSNS 74% (38/51), flunarizine 75% (39/52), tSNS plus flunarizine 76% (39/51)
Patient selection criteria	Inclusion criteria: patients between 18–65 years; episodic migraine with or without aura (ICHD-3 beta) and 2 migraine attacks per month. Exclusion criteria: use of preventive anti‐migraine medications before trial entry; depressive illness, Parkinson disease or other severe neurological or psychiatric disorders; implanted with a cardiac pacemaker, a defibrillator, or an implanted metallic or electrical device; pregnancy, lactation, or child‐bearing potential without adequate contraception.
Technique	tSNS alone Stimulation electrodes were placed on the forehead, covering the supratrochlear and supraorbital nerves. The stimulus parameters were set as follows: pulse width 250 microseconds, frequency 60 Hz, maximum intensity of 16 mA. Stimulation was daily for 20 minutes for 3 months duration. Flunarizine alone: 5 mg per day for 3 months according to Chinese guideline of migraine prevention. tSNS plus flunarizine: combination of both tSNS for 20 minutes daily at daytime plus flunarizine 5 mg per day at night for 3 months.
Follow up	3 months
Conflict of interest/source of funding	The authors declare that they have no conflict of interest. Authors received funding from Chongqing municipal commission of health and family planning; Chongqing municipal education commission; Chongqing science and technology commission. Medications used in this study were provided by a pharmaceutical company and medical device of tSNS (Cefaly) was supplied by the manufacturer.

Analysis

Follow-up issues: follow up was done every month until 3 months from a 30 day baseline period. Few patients were lost to follow up: 6 patients (4 in tSNS group, 2 in tSNS plus flunarizine group) were not compliant and 5 patients (3 in flunarizine group, 2 in tSNS plus flunarizine group) dropped out due to adverse effects. Patients not included in the analysis were equally distributed among the 3 groups (p=0.22).

Study design issues: prospective study in 1 headache clinic; after a 1 month run in phase, patients were randomised equally to 3 treatments. Randomisation was done through random number sequencing; patients were not blinded to treatment but outcome assessors were blinded. Allocation and treatment management was done by 3 doctors. The primary outcome measures were changes in migraine days and 50% responder rate of monthly migraine days. Secondary outcome measures were the changes in migraine intensity and intake of rescue medication, patient satisfaction and adverse events. Outcomes were recorded using structured headache diaries and interviews. Headache intensity was assessed using a 10‐points VAS (0: no pain; 10: severe pain prohibiting daily activities).

Study population issues: the patient demographic characteristics and clinical features were not different between the groups. Medication use was not under medical supervision.

Key efficacy findings

Number of patients analysed: 154

flunarizine plus e-TNS (n=51) versus e-TNS (n=51) versus flunarizine alone (n=52)

Outcome measures (data are expressed as mean ± SD or percent [no of patients]).

Clinical outcome	tSNS (n=51)	Flunarizine (n=52)	tSNS plus flunarizine (n=51)	P value between groups
Migraine days
Baseline, days/month	5.92 ± 1.04	5.68 ± 2.51	5.17 ± 2.02	0.27
3 month, days	3.73 ± 2.13	3.43 ± 2.56	2.00 ± 1.94
Change from baseline to 3 months, days	2.20 ± 2.43 (p<0.001)	2.25 ± 2.08 (p<0.001)	3.17 ± 2.44 (p<0.001)	0.039^^ 0.041^
50% responder rate (≥50% reduction in no. of migraine days/month) %	39.22%	46.15%	78.43%	<0.001^^ <0.001^
Migraine severity (assessed by VAS)
Baseline, days/month	6.75 ± 1.35	6.96 ± 1.10	6.57 ± 1.35	0.50
3 month, days	5.53 ± 2.05	4.82 ± 1.70	3.32 ± 1.83
Change from baseline to 3 months	1.22 ± 1.46 (p<0.001)	2.14 ± 1.52 (p<0.001)	3.25 ± 2.21 (p<0.001)	0.030^ <0.001^^
Monthly acute antimigraine drug intake
Baseline, days/month	4.88 ± 2.29	4.96 ± 2.56	4.15 ± 2.42	0.18
3 month, days	2.78 ± 2.19	2.89 ± 2.67	1.06 ± 1.03
Change from baseline to 3 months	2.10 ± 2.22 (p<0.001)	2.07 ± 2.37 (p<0.001)	3.09 ± 1.92 (p<0.001)	0.013^^ 0.02^

^comparison between flunarizine and tSNS plus flunarizine. ^^ comparison between tSNS and tSNS plus flunarizine

Patient satisfaction 3 months after treatment

	tSNS % (n=51)	Flunarizine % (n=52)	tSNS plus flunarizine % (n=51)	P value
Satisfied	54.9 (28/51)	50 (26/52)	84.31 (43/51)	0.001
Not satisfied	45.10 (23/51)	50 (26/52)	15.69 (8/51)

Key safety findings

Adverse event	tSNS % (n=51)	Flunarizine % (n=52)	tSNS plus flunarizine % (n=51)
Any adverse event (all mild and transient)	5.88 (3/51)	34.62 (18/51)	37.25 (19/51)
Somnolence (or drowsiness)	1	9	12
Fatigue	0	1	0
Insomnia	0	1	0
Forehead paraesthesia	1	0	2
Pressure sensation (on stimulation site)	1	0	0
Skin allergy (at electrode contact site)	0	0	1
Dizziness	0	1	0
Weight gain	0	6	4

The incidence of adverse effects in tSNS group was statistically significantly lower than that in combination therapy (p<0.001) but no difference between flunarizine and combination therapy (p=0.89).

Study 10 Przeklasa-Muszyńska 2017

Study details

Study type	RCT
Country	China
Recruitment period	January 2015 to December 2016
Study population and number	n=91 patients with pharmacotherapy-refractory headaches. e-TNS and pharmacotherapy (n=61) versus pharmacotherapy (control n=30) Indication: migraine prophylaxis treatment of episodic primary headaches 60 patients with migraine headaches and 31 patients with other primary headaches (chronic daily and tension type headaches) Migraine duration (average): patients with migraine with or without aura 19 years other primary headaches 8 years headache type: migraine with aura: eTNS n=20, control n=12 migraine without aura: eTNS n=16, control n=8 other primary headache eTNS n=21, control n=10
Age and sex	Mean age 45 years in both eTNS and control groups. Female: migraine with aura 95%, migraine without aura 87%, other primary headaches 47%
Patient selection criteria	Inclusion criteria: patients 18 to 65 years of age, with migraine headache (according to International Headache Society criteria), other primary headaches (tension type, trigeminal autonomic cephalgias according to IHS criteria), previously treated with pharmacotherapy (prophylactic or symptomatic) according to valid recommendations of primary headaches therapy, no contraindications to electrotherapy, no serious disorders and patient's consent. Exclusion criteria: incomplete diagnostics of headaches, diagnostic criteria or migraine, tension, or chronic daily headache not met, history of arrhythmias, pacemaker implantation, epilepsy or other reasons precluding the use of electrotherapy, no previous treatment, patient not giving consent for participation and kept pain diaries.
Technique	tSNS alone Stimulation electrodes were placed on the forehead, covering the supratrochlear and supraorbital nerves. Treatment was delivered in a pain clinic between headache episodes using the prophylactic mode, low-frequency pulses (60 Hz). 10 stimulation courses were delivered 2 or 3 times a week and duration of each stimulation session was 20 minutes. The sessions were not administered during headache episodes and in patients who had strong pain on the day of stimulation. Pharmacotherapy (control group): pharmacological treatment as prophylactic therapy (antiepileptic, antidepressant, beta blockers), symptomatic therapy: non-opioid analgesics, or triptans were given. In both groups, pharmacological treatment was continued without any changes during study depending on individual needs. Medication use was not under medical supervision.
Follow up	30 days
Conflict of interest/source of funding	None declared. Study funded by the department of pain research and treatment, Jagiellonian university medical college, Krakow, Poland institute of pharmacology, polish academy of sciences. Devices were rented from the company.

Analysis

Follow-up issues: 4 patients with migraine headaches stopped taking part in the study because of intolerance to stimulation sensations, 57 patients in the TSNS group completed the study.

Study design issues: prospective study, patients were randomly selected (randomisation details not reported) and data were analysed blindly according to the subgroups. Patients did not change pharmacological treatment during study. Headaches (frequency, intensity and duration) and medication use were recorded in pain dairies before, during and 30 days after study period. NRS (0-10 with higher scores indicating strong pain) was used to assess pain before each treatment. Outcomes assessed were monthly number of episodes (days with pain), duration of episodes (in hours), and pain intensity in NRS during an episode.

Study population issues: the patient demographic characteristics and clinical features were not different between the groups.

Key efficacy findings

Number of patients analysed: 91 (tSNS, n=61 versus control, n=30)

Outcome measures (data are expressed as mean ± SD or percent [no of patients]).

	tSNS (n=57)			Control (n=30)
Clinical outcome	Migraine without aura (n=16)	Migraine with aura (n=20)	Other primary headaches (n=21)	Migraine without aura (n=8)	Migraine with aura (n=12)	Other primary headaches (n=10)
Frequency of headache episodes
Baseline, days/month	7.5	7.5	18	10	11	17
30 days	4	4	10	7.5	7	12.5
P value	<0.05	<0.01	<0.001	NS	NS	NS
Pain intensity during headache episode (assessed using NRS)
Baseline, days/month	8.5	9	7.5	9	8.8	8.5
30 days	5.5	5.5	4.5	7.6	7.8	7.2
P value	<0.001	<0.001	<0.001	NS	NS	NS
Duration of headache episodes
Baseline, hours	23	27	12	18	15	7
30 days, hours	9	13	7	13	12	6
P value	<0.05	<0.05	<0.001	NS	NS	NS
% decrease in pain intensity
In patients with high intensity pain NRS (7-10)	32.7 (p<0.001)	32.8 (p<0.001)	34% (p<0.001)	12.5	10	11.5
Subjective patient assessment, %	46% (p<0.01)	47% (p<0.001)	40% (p<0.05)	22.5%	26%	28.8%

Key safety findings

Adverse event	tSNS % (n=61)	Control (n=30)
Intolerance to stimulation sensations	n=4
Skin irritation at electrode application site (for up to 30 minutes)	n=4

Study 11 Magis (2015)

Study details

Study type	Case series (survey) prospective registry
Country	France, Belgium and Switzerland
Recruitment period	2009-2012
Study population and number	n=2,313 patients with headache (migraine according to ICHD-2 criteria)
Age and sex	Age 14-87 years; 70.9% (1,641/2,313) female
Patient selection criteria	Only patients using specific antimigraine drugs (triptans) and most likely to have migraine were included in the survey.
Technique	Patients had tSNS (with a Cefaly device) for a 40-day period using a unit rented via the internet. Stimulation was delivered with an external self-adhesive electrode placed on the forehead. Impulses with a frequency of 60 Hz and intensity of 16mA were generated. All were given leaflets advising them to have at least 1 session daily for 20 minutes to obtain an effect. For 40 days the minimum total time of use recommended was 800 minutes. A built-in electronic system recorded the total time of use.
Follow up	29 months
Conflict of interest/source of funding	Not reported

Analysis

Study design issues: patient satisfaction and self-reported adverse events were monitored by phone interviews. It was not clear at what time point the outcomes were assessed. A built-in electronic system allowed the total duration of use or compliance in patients who returned the device after the trial period to be assessed.

Study population issues: no precise diagnosis of patients, authors' assumption was that all patients had migraine based on the fact that they used triptans as an abortive therapy. It was not clear if they had episodic or chronic migraine, migraine with or without aura; tension-type, medication overuse or cluster headaches. They did not have any regular neurological follow up.

Other issues: Authors state that some patients in whom the device was effective might not have purchased the device for financial reasons and this would have led to an overestimation of the proportion of non-satisfied subjects.

Key efficacy findings

Number of patients analysed: 2,313

Average rental period (from day the device was received until they were actually contacted to answer the questions): 58.2± 33.6 days.

Patient satisfaction and compliance: 53.4% (1236/2313) patients were satisfied and wanted to continue the treatment. They purchased the device. 46.6% (1077/2313) were not satisfied and returned the device, but the compliance showed that they used it only for 48.6% of the recommended time.

Key safety findings

	% (n)
1 or more adverse events (minor and fully reversible, some patients reported more than one event)	4.3 (99/2313)
Do not like the feeling and do not want to continue using the device	1.25 (29/2313)
Local pain/intolerance to paraesthesia (all stopped the treatment)	2.03 (47/2313)
Dental pain during/beginning of the session	0.09 (3/2313)
Cervical pain during sessions	0.04 (1/2313)
Cervical pain with nausea after the first 2 sessions	0.04 (1/2313)
Strong paraesthesia feeling on the left side	0.04 (1/2313)
Strong paraesthesia feeling on the right side	0.04 (1/2313)
More head pain when using the device during a headache	0.04 (1/2313)
Slight pain at one eyebrow during the first session	0.04 (1/2313)
Forehead skin burning sensation during a session	0.04 (1/2313)
Arousal and sleep changes	0.82 (19/2313)
Sleepiness during the session	0.52 (12/99)
Fatigue	0.13 (3/99)
Insomnia	0.17 (4/99)
Tension-type headache after stimulation	0.52 (12/2313)
Skin problems	0.38 (9/2313)
Reversible forehead skin irritation	0.22 (5/2313)
Transient local skin allergy	0.09 (2/2313)
Feeling of contusion on the forehead during a few days	0.09 (2/2313)
Other
Inability to keep eyes open during sessions	0.09 (2/2313)
Red eye after a session	0.04 (1/2313)
Eyes weeping during a session	0.04 (1/2313)
Feeling of stress during stimulation	0.09 (3/2313)
Pre-existing tinnitus increased during the session	0.04 (1/2313)
Tinnitus appearing after the session	0.04 (1/2313)
Wake up during night with a feeling of anxiety and tremor	0.04 (1/2313)
Vertigo during first session	0.04 (1/2313)
Short feeling of electrical shock	0.04 (1/2313)
Vomiting after a session	0.04 (1/2313)
Nausea and vertigo during session	0.04 (1/2313)
Nausea during sessions	0.04 (1/2313)
Forehead and cranial anaesthesia feeling during a few hours after a session	0.04 (1/2313)
Pressure feeling between the eyebrows during sessions	0.04 (1/2313)
Numbness at the back of the head after a session	0.04 (1/2313)
Subjective tachycardia during a session	0.04 (1/2313)
Migraine feeling during sessions	0.04 (1/2313)

Study 12 Danno D (2019)

Study details

Study type	Prospective case series ( UMIN-CTR(UMIN000033333)
Country	Japan (4 centres)
Recruitment period	April to December 2016
Study population and number	n=100 Indication: migraine prevention episodic migraine (n=60): 55 migraines without aura, 4 migraines with and without aura, and 1 migraine with aura chronic migraine (n=23): 2 migraines with aura not reported (n=17)
Age and sex	Mean 43.6 years; 68% (68/96) female
Patient selection criteria	Inclusion criteria: patients 18-75 years old, migraine with and without aura, including chronic migraine (ICHD 3beta) with at least 2 attacks and 4 days of migraine days per month, those using acute and prophylactic medicine who had not had the prescription changed in the last 3 months. Exclusion criteria: patients considered ineligible, who changed their prophylaxis, or received Botox or a nerve block within 3 months, with any secondary headache except for medication-overuse headache, severe neurological and/or psychiatric disorders, epilepsy, severe heart, liver, and/or renal dysfunction, allodynia, breast-feeding women, using opioids, with metal and/or electrical device in their body, using a cardiac pacemaker or implantable cardioverter defibrillator.
Technique	eTNS (with a Cefaly device) was delivered for 20 minutes every 24 hours. The electrical pulses are transmitted transcutaneously via a supraorbital bipolar self-adhesive electrode placed on the forehead covering the supratrochlear and supraorbital nerves bilaterally. Impulses were delivered at a pulse frequency 60 Hz, pulse width 300 microseconds, maximum intensity of 16 mA. Treatment period was 12 weeks, mean number of sessions was 75.6 days.
Follow up	12 weeks
Conflict of interest/source of funding	Authors declare no conflicts of interest.

Analysis

Follow-up issues: 4 patients dropped out (3 due to adverse events and 1 due to allodynia) and 96 patients completed 12 weeks treatment; of these, 6 patients did not complete electronic headache dairy; 7 patients did not meet the inclusion criteria about the number of migraine attacks during baseline. Therefore, only 83 patients were included in the analysis.

Study design issues: study was done in 4 headache clinics; 4 week baseline period was followed by 12 weeks treatment; primary outcome was change in the number of migraine days at 12 weeks from baseline. Secondary outcomes include the changes of the number of migraine attacks (headaches for which patients used triptans with moderate to severe intensity and with nausea or vomiting), headache days, acute medicine consumption days and headache severity. Patient satisfaction was assessed using a questionnaire survey. Headaches and device use were recorded in an electronic headache diary.

Study population issues: medical co-morbidities were not reported.

Key efficacy findings

Number of patients analysed: 83 (chronic migraine n=23, episodic migraine n=60)

Outcome measures (data are expressed as mean ± SD)

Primary outcomes	Baseline	4 weeks	8 weeks	12 weeks
Average migraine days*	8.16±4.53	7.63±5.01 (p=0.118)	7.35±4.93 (p=0.0403)	6.84±4.41 (p=0.0035)
Average migraine attacks	5.33±3.95	5.04±4.32 (p=0.256)	4.33±3.19 (p=0.0017)	3.94±2.44 (p=0.0002)
Average headache days	11.48±5.70	11.41±6.27 (p=0.874)	10.4±5.83 (p=0.0201)	9.81±5.66 (p=0.009)
Average acute antimigraine drug intake	8.75±4.41	8.47±4.91 (p=0.487)	8.52±5.22 (p=0.569)	7.83±4.91 (p=0.0166)
Average severity of headache (NRS)	4.5±1.55	4.08±1.46 (p=0.058)	4.14±1.61 (p=0.099)	4.11±1.66 (p=0.122)
50% responder rate (reduction of migraine days by >50%)			19.3%
30% responder rate			28.9%

* an average reduction of 1.32 days.

Comparison of outcomes between patients with chronic and episodic migraine (at 12 weeks from baseline; data are expressed as mean ± SD)

	Chronic migraine N=23	Episodic migraine N=60	P value
Migraine days	0.90±0.70	1.06±1.13	0.543
Migraine attacks	0.75±0.38	0.96±0.70	0.173
Headache days	0.89±0.43	0.99±0.76	0.552
Acute antimigraine drug intake	0.85±0.65	0.96±0.46	0.406
Severity of headache (NRS)	1.01±0.27	0.97±0.28	0.555

There was no significant difference in outcomes between patients with (n=53) and without (n=30) prophylactic treatment or with (n=6) or without (n=77) medication overuse. P values not reported.

Patient satisfaction at 12 weeks (assessed using questionnaire)

	% (n)
Satisfied	65.6 (63/95)
Reasons
Improvement in headache	44.8%
Reduction in consumption of acute medications	35.4%
Effective as acute treatment	40.4%
Headache while using device	(47/63)
Device use
Device used at night	87.5%
Installing electrodes easy	97.9%
Willing to purchase device	52.4%
Would 'for sure purchase' device	9.4%
'Probably purchase' device	26%

Key safety findings

Adverse events	n=100
Sleepiness	2
Strong stimulation	1
Tingling at the stimulation site	1
Discomfort at the stimulation site	1
Fatigue	1
Headache	1

Study 13 Birlea M (2019)

Study details

Study type	Prospective case series
Country	USA (single centre)
Recruitment period	February 2015 to April 2017
Study population and number	n=73 patients with chronic migraine Indication: migraine prevention Patients with preventive medications n=25
Age and sex	Mean 40.45 years; female 85% (49/58)
Patient selection criteria	Inclusion criteria: patients aged 18 to 65 years, with a history of chronic migraine (ICHD-3 beta) with or without medication overuse headache, having at least 15 headache days and 8 migraine days during the baseline month, preventive antimigraine medications started 3 months before baseline and no change in dose during study. Beck's Depression Inventory score less than 24, not pregnant, not-lactating, and not less than 6 months postpartum. Exclusion criteria: patients having treatment with Botox within the last 4 months or not able to tolerate the paraesthesia induced by the stimulation.
Technique	eTNS (with a Cefaly device) was delivered for 20 minutes once or twice daily for 3 months. The electrical pulses are transmitted transcutaneously via a supraorbital bipolar self-adhesive electrode placed on the forehead covering the supratrochlear and supraorbital nerves bilaterally. Impulses were delivered at a frequency of 60 Hz, pulse width 300 microseconds, and maximum intensity of 16 mA.
Follow up	3 months
Conflict of interest/source of funding	One author received research grants, honoraria and personal fees. Another author is an employee of Cefaly technology. The study was funded by the manufacturer.

Analysis

Follow-up issues: 21% (15/73) patients were not included in the data analysis (3 withdrew from study early, 3 changed prophylactic medications, 7 failed screening for headache and migraine days and 2 were lost to follow up).

Study design issues: study was done in 1 clinic; 4 week baseline period was followed by 3 months treatment; Primary outcomes were mean changes in frequency of headache days and overall acute headache medication intake. Secondary outcomes were changes in the number of migraine days, 50 and 30% responder rates for migraine days, headache days, monthly headache hours, frequency of headache episodes and headache intensity. Headaches (intensity, time, other symptoms), acute antimigraine medication intake and adverse events were recorded in an electronic headache diary. Headache intensity was measured on a scale of 1 to 3, with higher scores representing severe headache. Data analysis was done on an ITT basis.

Intake of acute antimigraine rescue medication was allowed at any time of the study and was reported in the headache diary. Daily number of sessions were not similar across patients. Device recorded the total number of treatment sessions and the total time of use.

Study population issues: 33 patients had medication overuse headache.

Key efficacy findings

Number of patients analysed: 58 (ITT analysis)

58.6% (34/58) non-continuous headache, and 41.4% (24/58) continuous headache (with CM and not interrupted by pain-free periods of more than 3 hours on more than 5 days/month unless these are attributed to drug treatment).

Outcome measures (ITT analysis)

Primary outcomes	Baseline (mean ± SD)	3 months (mean ± SD)	Mean change absolute (relative) between baseline and 3 months
Average migraine days
All patients (n=58)	19.02 ± 6.36	15.67 ± 8.70	-3.35 (-19.02%) (p<0.001)
Patients with non-continuous headache (n=34)	15.21 ± 4.06	11.76 ± 6.32	-3.44 (-23.38%) (p<0.001)
Patients with continuous headache (n=24)	24.42 ± 4.97	21.21 ± 8.69	-3.21 (-12.83%) (p<0.05)
Average number of headache days
All patients (n=58)	22.55 ± 5.38	19.43 ± 8.74	-3.12 (-16.21%) (p<0.001)
Patients with non-continuous headache (n=34)	18.71 ± 3.64	14.29 ± 7.08	-4.42 (-24.38%)
Patients with continuous headache (n=24)	28.00 ± 0.00	26.71 ± 4.79	1.29 (-4.62%)
Average number of headache episodes
All patients (n=58)	22.66 ± 5.35	19.53 ± 8.76	-3.12 (-16.03%) (p<0.001)
Patients with non-continuous headache (n=34)	18.89 ± 3.76	14.44 ± 7.20	-4.45 (-24.21%) (p<0.001)
Patients with continuous headache (n=24)	28.00 ± 0.00	26.75 ± 4.80	-1.25 (-4.46%)
Average number of moderate/severe headache days
All patients (n=58)	16.07 ± 7.65	12.97 ± 9.01	-3.10 (-21.78%) (p<0.001)
Patients with non-continuous headache (n=34)	12.21 ± 5.37	9.56 ± 6.18	-2.65 (-23.44%) (p<0.05)
Patients with continuous headache (n=24)	21.54 ± 7.10	17.79 ±10.23	-3.75 (-19.42%) (p<0.05)
Cumulative hours of headache on headache days
All patients (n=58)	249.64 ± 124.29	222.42 ± 154.74	-27.22 -(-14.95%) (p<0.05)
Patients with non-continuous headache (n=34)	166.00 ± 75.24	128.92 ± 94.73	-37.08 (-22.53%) (p<0.01)
Patients with continuous headache (n=24)	368.14 ± 71.89	354.88 ±123.42	-13.26 (-4.21%)
Average headache intensity
All patients (n=58)	1.98 ± 0.44	1.87 ± 0.47	-0.12 (-4.81%) (p<0.05)
Patients with non-continuous headache (n=34)	1.92 ± 0.42	1.87 ± 0.40	-0.05 (-1.12%)
Patients with continuous headache (n=24)	2.08 ± 0.45	1.87 + 0.57	-0.21 (-10.04%) (p<0.05)
Average acute antimigraine drug intake
All patients (n=58)	26.33 ± 30.25	18.22 ± 18.50	-8.11 (-30.81%) (p<0.001)
Patients with non-continuous headache (n=34)	21.68 ± 18.71	16.21 ± 14.72	-5.47 (-25.24%) (p<0.05)
Patients with continuous headache (n=24)	32.92 ± 41.08	21.06 ± 22.88	-11.85 (-36.01%) (p<0.01)
Percentage of 50% responder rate (reduction of migraine days by >50%)
All patients (n=58)		18.97%
Patients with non-continuous headache (n=34)		29.41%
Patients with continuous headache (n=24)		4.17%
Percentage of 30% responder rate (reduction of migraine days by >30%)
All patients (n=58)		24.14%
Patients with non-continuous headache (n=34)		38.24%
Patients with continuous headache (n=24)		4.17%
Mean number of eTNS sessions
All patients (n=58)		139.32
Patients with non-continuous headache (n=34)		127.74
Patients with continuous headache (n=24)		153.78

Key safety findings

Minor adverse events during eTNS treatment	47 events (n=26)
Related to eTNS device
Skin irritation at electrode site on forehead	1
Worsening headaches and vertigo (stopped using device)	1
Not related to treatment
Worsening myasthenia gravis and migraine headaches (patient hospitalised)	1

Study 14 MAUDE database

Study details

Study type	MAUDE adverse event report
Country	USA
Study population and number	n=1
Age and sex	Not reported
Patient selection criteria	Not reported
Technique	Patient tested tSNS (with a Cefaly device)
Follow up
Conflict of interest/ source of funding	Not reported

Key safety findings

Number of patients analysed: 1

Event date: 08/01/2014

Event type: Injury, reported by a physician

Event description: Cefaly device was prescribed to a patient by a neurologist to help decrease the frequency of headaches which the patient had for many years. Patient purchased and tested the device as described in the manufacturer's instructions on a day that she was completely well. The intensity of the power was set to the lowest setting.

About 17 minutes after starting the device on the lowest setting, the patient began to experience weakness in jaw muscles and upper extremities (proximally more than distally) and to a lesser extent, proximal lower extremity muscles. Patient also developed significant dizziness. These symptoms increased until the use of the device was stopped. After shutting the device off, the muscle weakness and dizziness did not immediately stop. It took a period of about 2 hours for complete resolution of these symptoms.

How are you taking part in this consultation?

Interventional procedure overview of transcutaneous electrical stimulation of the supraorbital nerve for treating and preventing migraine