Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

Review

Cephalalgia
2023, Vol. 43(8) 1–15
Management of cluster headache: ! International Headache Society 2023
Article reuse guidelines:
Treatments and their mechanisms sagepub.com/journals-permissions
DOI: 10.1177/03331024231196808
journals.sagepub.com/home/cep

Kuan-Po Peng1 and Mark J. Burish2

Abstract
Background: The management of cluster headache is similar to that of other primary headache disorders and can be
broadly divided into acute and preventive treatments. Acute treatments for cluster headache are primarily delivered via
rapid, non-oral routes (such as inhalation, nasal, or subcutaneous) while preventives include a variety of unrelated
treatments such as corticosteroids, verapamil, and galcanezumab. Neuromodulation is becoming an increasingly popular
option, both non-invasively such as vagus nerve stimulation when medical treatment is contraindicated or side effects
are intolerable, and invasively such as occipital nerve stimulation when medical treatment is ineffective. Clinically, this
collection of treatment types provides a range of options for the informed clinician. Scientifically, this collection provides
important insights into disease mechanisms.
Methods: Two authors performed independent narrative reviews of the literature on guideline recommendations,
clinical trials, real-world data, and mechanistic studies.
Results: Cluster headache is treated with acute treatments, bridge treatments, and preventive treatments. Common
first-line treatments include subcutaneous sumatriptan and high-flow oxygen as acute treatments, corticosteroids
(oral or suboccipital injections) as bridge treatments, and verapamil as a preventive treatment. Some newer acute
(non-invasive vagus nerve stimulation) and preventive (galcanezumab) treatments have excellent clinical trial data for
episodic cluster headache, while other newer treatments (occipital nerve stimulation) have been specifically tested in
treatment-refractory chronic cluster headache. Most treatments are suspected to act on the trigeminovascular system,
the autonomic system, or the hypothalamus.
Conclusions: The first-line treatments have not changed in recent years, but new treatments have provided additional
options for patients.

Keywords
medication, neuromodulation, CGRP, hypothalamus, trigeminovascular
Date received: 22 June 2023; revised: 30 July 2023; accepted: 1 August 2023

Introduction
with longer bouts (more than three weeks) or with
Cluster headache (CH) treatments are categorized as bouts of unpredictable duration, bridge treatments
acute (which manage an ongoing attack), bridge may be used while titrating up preventives so that the
(quick-acting preventives that cannot be used for long preventive is at a sufficient dose when the bridge
periods of time), and long-term preventive (hereafter
simply referred to as “preventive”). All CH patients
should be given acute treatments. For bridge and pre- 1
Department of Systems Neuroscience, University Medical Center
ventive treatments, usage depends on the type of CH. Hamburg-Eppendorf, Hamburg, Germany
For episodic cluster headache (eCH) patients with 2
Department of Neurosurgery, UTHealth Houston, Houston, Texas,
short bouts (three weeks or less, defined as the period USA
during which the patient experiences the CH attacks),
Corresponding author:
bridge treatments without preventives may be suffi- Mark J. Burish, Department of Neurosurgery, UTHealth Houston, 6400
cient, and the medications can be stopped when the Fannin Street, Suite 2010, Houston, TX, USA 77006.
patient is out of their headache bout. For eCH patients Email: mark.j.burish@uth.tmc.edu

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-
NonCommercial 4.0 License (https://creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and dis-
tribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.
sagepub.com/en-us/nam/open-access-at-sage).
2 Cephalalgia

medication wears off. When the headache bout is over, prescribed concurrently: in the authors’ personal expe-
the treatments can be down titrated and stopped. For riences, at the start of an eCH bout, we have at times
chronic cluster headache (cCH) patients, preventives prescribed two acute treatments (such as sumatriptan
are typically used year-round. Bridge medications are and oxygen), one bridge treatment (such as an occipital
still useful in cCH during headache flares; some cCH nerve block), and one preventive (such as verapamil).
patients, for example, can have a circannual pattern This review focuses on the first-line treatments
where the headaches are more frequent at certain (sumatriptan, oxygen, corticosteroids, and verapamil)
times of the year (1,2). as well as treatments that have emerged in the last few
The list of treatments for CH is shown in Table 1, years (galcanezumab as well as invasive and non-
along with their guideline recommendations and, in the invasive neuromodulation options). To arrive at these
case of acute treatments, their comparative effective- treatments, the authors independently examined the
ness from a recent network meta-analysis (3). Often American and European guidelines for CH (4,5), per-
multiple acute, bridge, and preventive medications are formed a narrative review of the literature for all of

Table 1. Treatment of Cluster Headache based on European and American society guidelines as well as probability ranking of acute
treatments from a network meta-analysis.

European Federation American


of Neurological Headache Ranking in
Societies Society Dose examined for Network
Recommendation Recommendation network meta-analysis Meta-Analysis

Acute
Oxygen (high flow) A A 12 L/min 1
Sumatriptan subcutaneous A A 6 mg 2
Sumatriptan nasal A B 20 mg 3
Oxygen (low flow) A A 6–7 L/min 4
Zolmitriptan nasal A A 5–10 mg 5
Octreotide subctaneous B C 100 mg 6
Noninvasive vagus nerve new* new* 3  2 min stimulations 7
stimulation (eCH only)
Zolmitriptan oral B B 5–10 mg NR
Lidocaine nasal B C 10% NR
——————————————— ———————— ——————— ——————————— ——————
Bridge Dose used in clinical trials
Prednisone A new* 100 mg daily for 5 days,
decreasing by 20 mg
every 3 days until off
Ipsilateral greater NR A Multiple; all use steroids
occipital nerve block
——————————————— ———————— ——————— ———————————
Preventive Dose used in clinical trials
Verapamil A C 360 mg divided TID
Lithium B C 900 mg divided TID
Melatonin C C 10 mg at bedtime
Topiramate B NR 100–400 mg divided BID***
Baclofen C NR 15–30 mg divided TID***
Warfarin NR C Titrated to INR 1.5–1.9
Galcanezumab (episodic only) new* new* 300 mg monthly
Valproic acid C unfavorable** 600–2000 mg divided BID
———————————————— ———————— ———————
Refractory
Occipital nerve stimulation new* new*
Sphenopalatine ganglion stimulation NR B
Hypothalamic deep brain stimulation NR unfavorable**
eCH, episodic cluster headache; NR, not rated; TID, three times daily.
*
Several treatments are too new to be placed in the guidelines. **American Headache Society guidelines have negative ratings for sodium valproate and
deep brain stimulation: both were found to be “probably ineffective.” ***Dose is based on clinical trials except for baclofen and topiramate, which are
based on open-label prospective studies.
Peng and Burish 3

these treatments, created summaries, and then com- formulations (24). The triptans can cause coronary
pared notes. vasoconstriction and thus chest tightness is a concern
that should be evaluated (24,28), however it should also
be noted that the majority of chest tightness cases
Acute pharmacological treatments
appear to be non-cardiac in etiology (24). Selective
Triptans: Sumatriptan and zolmitriptan 5HT1F agonists (which are called ditans instead of trip-
tans) are devoid of vasoconstrictive side effects and are
Triptans, as a class, consist of seven approved medica- effective in a preclinical model of CH (29). However,
tions that activate serotonin receptors 5HT1B, 5HT1D, human data on the efficacy of ditans in the treatment
and in some cases 5HT1F (6). Two of these triptans – of CH are still lacking. The current formulation of
sumatriptan and zolmitriptan – have fast non-oral ditans also lacks the fast-acting, non-oral form typical-
routes of administration, and these are the ones recom- ly needed for the treatment of CH.
mended for CH. Randomized clinical trials (RCTs) have Triptans are not recommended in patients with a his-
shown the effectiveness of subcutaneous sumatriptan tory of stroke or myocardial infarction or with uncon-
6 mg at 15 minutes and of intranasal sumatriptan 20 mg trolled hypertension. They should not be used in
and intranasal zolmitriptan 5–10 mg at 30 minutes (7–11). combination with monoamine oxidase inhibitor antide-
Of note oral zolmitriptan is also recommended for CH pressants as some triptans (including sumatriptan and
and is also effective at 30 minutes, but at a higher dose zolmitriptan) are degraded via monoamine oxidase
(10 mg) than for migraine (2.5–5 mg) (12,13). pathways (30). However, there is less concern combining
Real-world data also supports the use of sumatrip- triptans with other antidepressants; observational stud-
tan and zolmitriptan. In one study of 2031 attacks in 52 ies have shown a low risk for serotonin syndrome when
patients treated with subcutaneous sumatriptan, combining triptans with selective serotonin- or
attacks were aborted 88% of the time. Additionally serotonin-norepinephrine reuptake inhibitors (31,32).
42% of patients showed pain freedom at 15 minutes The mechanism of action of triptans is an interesting
in over 90% of their attacks (14). Other real-world historical debate; it was initially thought that the vaso-
studies have noted that patients can obtain relief constrictive function of these medications was the prima-
from less than 6 mg (15) and there appears to be no ry mechanism of action, while today the widely accepted
tachyphylaxis (14,16,17). Studies of predictors for trip- theory is that triptans primarily work in neurons through
tan response have shown that patients with eCH (as 5HT1B/5HT1D receptors to prevent the release of vasoac-
opposed to cCH), younger patients, and patients with tive peptides during trigeminovascular activation includ-
shorter and less frequent attacks may be more respon- ing calcitonin gene-related peptide (CGRP) (33,34),
sive to triptans (18–20). One genetic study identified a substance P (34), neurokinin A (34), and the activation
common GNB3 variant as a predictor of triptan of transient receptor potential vanilloid 1 (TRPV1) chan-
response in CH (21), though testing for GNB3 is not nels (35). An ongoing debate is where the triptans act, as
commonplace in clinical practice. 5HT1B and 5HT1D receptors are found peripherally in the
Real-world data further suggest a risk of medication trigeminal ganglion and centrally in the brainstem, thal-
overuse in cluster headache patients (both episodic and amus, and hypothalamus (36–39). The subcutaneous for-
chronic). Offending medications include triptans, ergot mulation may have a more rapid uptake into the central
derivatives, opioids, caffeine, non-steroidal anti-inflam- nervous system (40). Evidence from a recent human study
matory drugs (NSAIDs), and paracetamol, which suggests that one target is the junction between the axons
result in either an increased frequency of attacks or a of the first-order sensory neurons and the cell body of the
milder background headache (22). In one large case
second-order sensory neurons in the brainstem (41).
series of 430 patients, medication overuse headache
Furthermore, sumatriptan continued to produce a good
was found in 4% of patients and the headache was
response in a CH patient who underwent trigeminal gan-
most commonly a headache similar to that experienced
gliotomy, further supporting a potential central target of
by migraine patients with medication overuse head-
action (42).
ache: a bilateral, dull headache with no associated fea-
tures (23). A personal or family history of migraine is
quite common in cluster headache patients with medi-
Oxygen
cation overuse headache (23). Oxygen has been proposed as a CH treatment since
Side effects from triptans include sedation, dizziness, 1952 (43). Randomized clinical trials (RCTs) have
transient hypertension, flushing, and tightening sensa- shown that it is effective at high flow rates (7–12 L/
tions (in the chest, throat/neck, or extremities) (24–27) min) at 15 minutes (44–46), and should be used with a
and many of these effects are more common in subcu- non-rebreather mask or, in recent studies, a demand
taneous sumatriptan than in the nasal or oral valve mask (47,48). Predictors of oxygen
4 Cephalalgia

responsiveness have been examined in small studies and recent RCT has shown the effectiveness of a 17-day
include eCH (as opposed to cCH), younger age, a lack taper consisting of 100 mg of prednisone daily for five
of nausea/vomiting, a lack of restlessness, and possibly days, followed by a 20 mg taper every three days until
male sex (20,49–51). Side effects are typically minimal discontinuation (61). This study is notable because it is
with oxygen such as lightheadedness or dry mouth (45), a higher dose than previous observational studies and
but there are two issues to note. First, patients can have perhaps a higher dose than what many clinicians pre-
a rebound headache when they finish the oxygen treat- viously used. The comparative effectiveness of oral
ment (52,53). Second, oxygen can ignite flammable versus suboccipital steroids is unclear precisely because
substances and patients should not use it near sparks this 100 mg prednisone clinical trial is new and is higher
or open flames (54). The rate of cigarette smoking is than most real-world comparisons.
higher in CH than in the general population, and For suboccipital steroid injections, the efficacy was
patients should not smoke near the oxygen. There are first reported in 1985 in a small (n ¼ 20) open-label
few contraindications, but caution should be given to study (62) and confirmed in two double-blind place-
patients with severe pulmonary disease as high-flow bo-controlled RCTs (63,64). Ambrosini et al. (63)
supplemental oxygen can reduce the respiratory drive recruited 23 CH patients (16 eCH in a new cluster
and cause hypercapnia (55). bout, defined as  7 days, and 7 cCH) and randomized
Like triptans, oxygen was originally thought to them to receive either suboccipital steroid injections or
work via vasoconstriction, but it is now thought to placebo injections. After one week of follow-up, 11 out
have a neuronal mechanism (43,56). However much of 13 patients in the betamethasone group became
less is known about the mechanism by which oxygen attack-free, compared with 0 out of 10 in the placebo
aborts CH attacks. Oxygen may modulate the auto- group. The effects were maintained for at least four
nomic system, as rodent studies show that oxygen weeks (63). Leroux et al. (64) investigated the efficacy
reduces pain signaling in the trigeminocervical complex of repeated injections of cortivazol as add-on therapy
when the superior salivatory nucleus is stimulated but for CH patients. A total of 15 cCH and 28 eCH
not when the trigeminal nerve (via the dura) is stimu- patients were recruited. The primary outcome was the
lated (57). Human studies activating the peripheral proportion of patients reporting an attack frequency of
portion of the cranial autonomic system using intrana- 2/day after injection. Twenty of 21 patients in the
sal kinetic oscillation stimulation, however, showed no cortivazol group had a met the primary outcome, com-
modulatory effects of oxygen on provoked parasympa- pared to 12 of 22 in the control group (64). Both
thetic output (58) (for more details of the peripheral double-blind placebo-controlled RCTs (63,64) and
autonomic pathway, see the discussion of the other non-controlled clinical trials (65–69) have consis-
trigeminal-autonomic reflex in the vagus nerve stimu- tently reported a good response rate, usually  80%,
lation section below). Furthermore oxygen inhalation although a head-to-head comparison between trials is
reduces plasma CGRP levels after spontaneous cluster not easy due to different outcome measures and cohort
attacks (59), although it is unclear whether oxygen characteristics.
directly inhibits CGRP release or whether this effect Corticosteroids, both oral and injected, are consid-
is secondary to headache termination. Oxygen also ered bridge treatments because of the safety risks of
appears to block protein extravasation in the dura extended use which include weight gain, osteoporosis,
(60). In short, there is still no consensus on how immunosuppression, peptic ulcer disease, avascular
oxygen works in CH. necrosis of the hip, and adrenal insufficiency. These
side effects are more commonly studied with oral ste-
roids, though suboccipital injections are limited to sev-
Bridge treatment eral times per year because of both local and systemic
issues with high-dose steroid injections. The local or
Corticosteroids
site-specific adverse events that may be associated
Corticosteroids have two routes of administration, with suboccipital injections include alopecia and cuta-
either as a daily oral tablet or as a suboccipital injec- neous atrophy (70). There are also acute side effects
tion. It is often referred to as a suboccipital steroid with both oral and injected steroids, which include
injection, rather than a greater occipital nerve (GON) insomnia, hyperglycemia (especially in diabetic
block, because the primary medication is the steroid patients), and mood changes (including mania in bipo-
and not the local anesthetic. However, the two terms lar patients).
are used interchangeably. The mechanism of corticosteroids in CH is unclear,
For oral tablets, the best studied is prednisone but partly because corticosteroids have multiple mecha-
the dose and titration schedule have long been debated nisms that could target CH. In the circadian system
due to a lack of large clinical trials. Fortunately, a corticosteroids signal daytime (and melatonin signals
Peng and Burish 5

nighttime); CH patients have higher corticosteroid and The mechanism of verapamil in CH is unclear, like
lower melatonin levels, and CH patients often have a steroids it has multiple potential mechanisms.
predictable clock-like pattern to their attacks (71). Verapamil appears to modulate a variety of serotoner-
Corticosteroids have also been shown to alter CGRP gic, noradrenergic, dopaminergic, muscarinic, and
levels in CH patients and endogenous opioid systems opioid receptors (34,78). Verapamil may block CGRP
for pain relief (72–74). The mechanism of suboccipital release through calcium channel blockade (77); along
steroid injections may be more localized than systemic, these lines, other calcium channel blockers like nimo-
based on the close functional connection between the dipine and nifedipine have been tried with possible
cervical (C2/C3) somatosensory and trigeminal sensory effectiveness (77). Furthermore verapamil may alter
systems (75) and the fact that occipital nerve blocks circadian rhythms, as patients using verapamil had
partially inhibit the trigeminal nerve (specifically the their headaches shifted forward by one hour (84) and
nociceptive blink reflex [76]), suboccipital steroid injec- one preclinical study found that verapamil alters core
tions may provide a more specific circadian gene expression in both the trigeminal gan-
trigeminal modulatory effect than systemic steroid glion and the hypothalamus in mice (85).
administration.
Anti-calcitonin gene-related peptide therapy
Preventive pharmacological treatments The idea of CGRP blockade as a treatment for head-
aches originally came from findings that plasma
Verapamil CGRP from the external jugular vein increases during
Verapamil was first proposed in 1983 (77), with the first attacks of both migraine (86) and CH (59), suggesting
RCTs in 1990 (for cCH [78]) and 2000 (for eCH [79]). trigeminovascular activation in both cases. Indeed,
Despite the relatively low number of patients enrolled migraine and CH share certain CGRP-related disease
in these trials (n ¼ 30 each), verapamil is widely consid- mechanisms. Intravenous infusion of CGRP triggers
ered the first-line preventive for CH. These trials are CH attacks (87) and migraine attacks (88), and anti-
supported by significant patient survey data showing CGRP monoclonal antibodies (CGRP-mAbs) have
that verapamil is one of the most effective preventives emerged in recent years to become indispensable in
(50,80). In real-world studies, however, the proposed the treatment of migraine (89). It would follow, then,
doses are typically higher than the 360 mg daily (divid- that all CGRP-mAbs would be effective for CH as
ed three times daily (TID)) from the RCTs. One study well as migraine, but the data are more complicated
found that CH patients used an average dose of 587 mg (Table 2). Galcanezumab, one of the four CGRP-
daily (81) leading to the idea of “neurologic doses” mAbs currently available, showed efficacy in reducing
which can be twice that of “cardiovascular doses” attack frequency in RCT for patients with eCH (90), but
(77,82). One possible reason for these high doses is not in patients with cCH (91). Three RCTs of fremane-
that verapamil is a substrate for the P-glycoprotein zumab in CH (clinicaltrials.gov number NCT03107052),
pump, which transports drugs out of the brain. As eCH (NCT02945046), and cCH (NCT02964338) were
verapamil is thought to act in the central nervous terminated early due to the results of interim futility
system, very high doses may be needed to overcome analyses, as was an eptinezumab study in eCH
this drug pump (53,77,82). (NCT04688775). Several clinical trials are still ongoing,
Side effects from verapamil include constipation including eptinezumab in cCH (NCT05064397), erenu-
which in the authors’ experience is fairly common at mab in cCH (NCT04970355), and rimegepant (an oral
higher doses. Others include lower extremity edema, small molecule CGRP antagonist rather than a mono-
bradycardia, fatigue, gastrointestinal upset, erectile clonal antibody) in CH (NCT05264714). In summary,
dysfunction, and, at high doses, gingival hyperplasia for eCH, one RCT showed efficacy and two others were
(77). One of the most concerning adverse side effects terminated early for futility. For cCH, no RCT has
is heart block, and an electrocardiogram (to monitor shown efficacy over placebo. Several case series
the PR interval) is recommended before treatment, one showed potential clinical efficacy in a subset of cCH
to two weeks after each dose increase, every one to two patients (92–94), but how to identify these patients
months thereafter out to six months, then every six who are likely to benefit from the treatment requires
months because delayed heart block has been docu- further investigation. One explanation as to why
mented (53,81,83). Additionally, verapamil is a CGRP-mAbs have not worked as well in CH as they
CYP3A4 inhibitor and can increase the serum concen- have in migraine may lie in the attack mechanism. Up to
tration of other medications sometimes used to treat 30% of migraine attacks may be CGRP-independent
CH such as ergot derivatives (ergotamine and dihydro- (95) and similar non-CGRP mechanisms may be at
ergotamine), gepants, and lithium. play in CH, as attacks triggered by infusion of
6 Cephalalgia

Table 2. The current landscape of CGRP inhibitor trials for cluster headache. One CGRP receptor inhibitor not listed - atogepant -
is approved for migraine but is not currently being investigated for cluster headache.

Episodic cluster headache Chronic cluster headache

Effective In clinical trials Ineffective Effective In clinical trials Ineffective

Eptinezumab  *
Erenumab 
Fremanezumab  
Galcanezumab  
Rimegepant * *
*
These studies are open-label, while all others are randomized and double-blinded.

vasoactive intestinal peptide or pituitary adenylate reasons: 1) CGRP binds not only to the CGRP recep-
cyclase-activating polypeptide-38 did not lead to tor but also to the amylin-1 receptor (100), and 2) a
increased levels of CGRP (96). Indeed, the data using functional magnetic resonance study showed ligand-
CGRP as a pharmacological trigger gave a success rate and receptor-specific activation of networks in patients
of 89% in eCH patients during their active period (clus- receiving galcanezumab versus erenumab (104). While
ter bout) and only 50% in cCH patients (87). In addi- the mechanism of action is clear, the site of action of
tion, baseline plasma CGRP levels are higher in eCH CGRP-mAbs is, in contrast, still under debate. CGRP
patients, compared to chronic patients (97). Taken receptors have been found in intracranial arteries,
together, this suggests a greater role for a CGRP- peripheral neurons, and central neurons (100). In
independent pathway, especially in cCH patients. The fact, CGRP is the most abundant neuropeptide in the
effect of triptans in aborting CH attacks may involve trigeminal ganglion (105). Theoretically, due to the
the inhibition of CGRP release in trigeminal neurons large molecular size of the antibody, the percentage
(98,99). Whether the response to triptans correlates of CGRP-mAbs that cross the blood-brain barrier is
with the response to CGRP-mAbs is of clinical interest. minuscule and an effect (at least direct) on the central
Future studies investigating the effect of anti-CGRP nervous system is unlikely (106,107). Thus, the trigem-
treatment in CH patients that include biological meas- inal ganglion, located outside the blood-brain barrier,
urements of CGRP would help to better understand would presumably be the primary target for CGRP-
who benefits from the treatment. mAbs. However, the role of the trigeminal ganglion
Side effects for CGRP-mAbs are remarkably limited in CH may be less prominent than in migraine. In
given the widespread functions of CGRP in multiple one case report, a CH patient still experienced attacks
organ systems including cardiovascular, neurovascular, after trigeminal root resection (42). In another case
pulmonary, gastrointestinal, muscular, and immuno- series, ten CH patients underwent trigeminal root
logic (100). Common side effects from clinical trials resection. Five of them continued to have their regular
were upper respiratory tract infections and injection cluster attacks, while another five had periodic cranial
site reactions. Of all of the CGRP-mAbs, erenumab autonomic symptoms without headache (108). Both
has undergone the most clinical studies on adverse studies suggest the importance of a central pain gener-
events and potential side effects, finding a risk of con- ator in the pathophysiology of CH. Cumulative evi-
stipation, no risk of worsening angina in patients with dence from functional imaging studies (109,110) and
stable angina, and a better adverse event profile than neurophysiological studies (111,112) suggests that there
topiramate in a head-to-head trial (101–103). While is a central effect of CGRP-mAbs, albeit probably sec-
early trials of small molecule CGRP receptor inhibitors ondary to peripheral drug modulation. Some even sug-
were plagued by hepatotoxicity, hepatoxicity has not gest a differential central effect of CGRP-mAbs between
been a concern with the CGRP-mAbs or the newer responders and non-responders (109,110).
generation small molecule oral CGRP antagonists. There are several other options for CH prevention
The mechanism of action for CGRP-mAbs is (Table 1), and here we will briefly discuss two common-
thought to be clear, as these antibodies are highly spe- ly used ones, lithium and melatonin. The clinical trial
cific for either the CGRP ligand (eptinezumab, frema- dose of lithium was 900 mg divided TID (78), but it has
nezumab, and galcanezumab) or for the CGRP multiple side effects that include weight gain, diabetes
receptor (erenumab). It should be noted, however, insipidus, hypothyroidism, and hyperparathyroidism
that CGRP ligand blockers may have different effects (113), as well as a narrow therapeutic window with tox-
than CGRP receptor blockers for the following two icity linked to elevated lithium levels. Recommended
Peng and Burish 7

monitoring, from a variety of guidelines (114), includes The concept of treating CH with non-invasive vagus
lithium levels within a week of dose changes as well as nerve stimulation (nVNS) was based both on its
every 3–12 months thereafter, as well as renal and thy- known modulatory effect on the parasympathetic
roid labs every 6–12 months. Melatonin, in contrast, is system and on an early case report of headache
a widely available supplement with lower side effects improvement as an incidental finding in patients receiv-
(primarily sedation) that was effective for CH when ing invasive VNS for treatment-resistant depression
10 mg is taken at bedtime (115). Both of these medica- (118). In double-blind, randomized, sham-controlled
tions have known circadian effects (116) though lithium trials, nVNS was superior to placebo in aborting
also affects several neurotransmitters (34). acute cluster attacks exclusively in patients with eCH
(119,120). Despite the lack of efficacy as an acute treat-
ment in patients with cCH, early open-label studies
Neuromodulation suggested a potential prophylactic effect of nVNS –
more than 50% of cCH patients had a 50% reduction
Vagus nerve stimulation
in headache frequency (121). This prompted subse-
CH is characterized by prominent unilateral cranial quent studies on the use of nVNS as a preventive treat-
autonomic symptoms such as lacrimation, rhinorrhea, ment option. Two studies reported the clinical efficacy
and ptosis. Cranial autonomic symptoms are due to the of nVNS as an adjunctive prophylactic treatment in
activation of the trigeminal autonomic reflex which cCH patients receiving standard of care therapy
consists of two arcs: the trigeminal nociceptive input (122,123). eCH patients also benefit from nVNS as
and the parasympathetic output (117) (Figure 1). an adjunctive prophylactic treatment (122).

Figure 1. Schematic representation of peripheral and central pathways summarizing potential targets for neuromodulation in the
treatment of cluster headache. Trigeminal pain during cluster headache attacks (red) leads to the activation of the trigeminal pain
pathway (black) and the activation of the parasympathetic output (green). The spinal trigeminal nucleus in the trigeminal cervical
complex has a bidirectional connection to the hypothalamus, which may initiate cluster headache attacks and serve as a modulatory
center for trigeminal nociception. Notably, a direct trigeminohypothalamic tract has been demonstrated in animal studies, but not yet
in humans. Neuromodulatory treatments and their potential targets are shown in this figure. Other brain areas relevant to the
pathophysiology of cluster headache, such as the insula, cerebellum, or occipital cortex, are not depicted in this figure because there is
insufficient evidence that they are potential targets of currently available neuromodulation treatments for cluster headache. Solid lines:
monosynaptic connections; dashed lines: multisynaptic connections. HT: hypothalamus; NTS: nucleus tractus solitarius; S1: primary
sensory cortex; SPG: sphenopalatine ganglion; SSN: superior salivatory nucleus; TCC: trigeminal cervical complex; TG: trigeminal
ganglion; Thal: thalamus. Figure created with BioRender.com.
8 Cephalalgia

Side effects from nVNS include a brief pulling Occipital nerve stimulation
sensation of the face or lip (possibly from platysma
The success of suboccipital steroid injections in the
muscle contraction). Exclusion criteria used in the
treatment of CH, whether episodic or chronic, has led
study, which could be considered suggestions for clin-
to the idea of treating cCH patients with GON stimu-
ical contraindications, included an abnormal electro-
lation, as regular repeated nerve blocks are either inac-
cardiogram (as the right vagus nerve innervates the
cessible in the long term or carry risks because they
sinoatrial node of the heart [124]); implanted electrical
need to be performed too frequently. The first cohort
or neuromodulatory devices; cervical spine hardware
study was conducted in 2007, and in a cohort of eight
or other metal near the neck; and diseases of the
refractory cCH patients, six achieved responses
head (including seizures or syncope in the last year, (25%–95%) in either frequency or intensity reduction
brain tumors, or significant head trauma), the neck (134). Subsequent open-label studies reported similar
(carotid endarterectomy or severe carotid artery dis- efficacy with approximately 50–70% reduction in
ease), or the heart (uncontrolled hypertension, severe either frequency or intensity of cluster attacks
coronary artery disease, recent myocardial infarction, (135–137). The first RCT was conducted in 2021 in
or congestive heart failure). medically refractory cCH patients (defined as failure,
Despite the clinical efficacy of nVNS, its mechanism intolerance, or contraindications to verapamil, lithium,
of action is not fully understood. nVNS appears to and at least one other preventive) (138). This RCT
modulate the trigeminal autonomic reflex, including showed a more than 40% improvement in headache
the reduction of parasympathetic output (VNS reduces frequency at 21–24 weeks post-operatively, however
lacrimation in healthy control humans [125]), the the improvement was the same in the experimental
reduction of trigeminal sensitivity in both an animal group (100% stimulation) and the “sham” group
model (126) and a human model (127), and the link (30% stimulation). A placebo effect thus cannot be
between the two (when the superior salivary nucleus ruled out in this study, though such a strong placebo
is stimulated to increase firing in the trigeminal effect in such a medically refractory group raises the
nucleus of rodents, the addition of VNS reduces the question of whether the sham protocol is already suf-
trigeminal nucleus firing [128]). It is not clear which ficient to stimulate the occipital nerve in the current
part of this pathway is crucial for the clinical effect of paradigm. Such cases are not uncommon with neuro-
nVNS. Provocation of cranial autonomic symptoms by modulatory devices: the sham device for nVNS can
induction of the trigeminal autonomic reflex in CH effectively reduce parasympathetic output and is not
patients does not trigger CH attacks, suggesting ideal for sham stimulation (125). Adverse events were
that the parasympathetic output is a downstream similar in both open-label studies and the RCT and
phenomenon in CH attacks, and its suppression may were primarily hardware-related issues (lead migration,
only reduce the cranial autonomic symptoms but not battery depletion, infection, and wound healing issues).
headache (129). Therefore, the clinical efficacy of In the absence of adverse events, GON stimulation
nVNS may involve a mechanism that is parallel to appears to be an effective long-term treatment option,
peripheral parasympathetic suppression. Potential can- with an open-label study showing that 66.7% remained
didates for this mechanism include a direct trigeminal responders (>50% improvement) for an average of 6.7
nociceptive modulation or the involvement of other years of follow-up (139).
central structures such as the hypothalamus, thalamus, Patients with severe psychiatric comorbidity were
periaqueductal gray, cortex, or cerebellum (130,131). less likely to respond to occipital nerve stimulation
A recent functional imaging study showed altered (140). In addition, response to suboccipital injections
activation in the spinal trigeminal nuclei, pons, does not reliably predict the response to occipital nerve
parahippocampal gyri, and hypothalamus (132). The stimulation (140,141), making it unclear who is a good
hypothalamus has long been implicated as a cluster candidate for this relatively invasive treatment.
generator (133), and its involvement provides a poten- Suggested criteria for surgical treatment of CH, i.e.,
tial mechanism by which nVNS may modulate disease occipital nerve stimulator placement, include cCH (as
activity and subsequently reduce headache frequency. opposed to eCH) for at least two years, headache fre-
Similarly, involvement of the trigeminal nucleus pro- quency between four per week and daily, failure of at
vides an explanation for how nVNS may suppress least three if not all preventive medications including
acute attacks by directly modulating trigeminal pain verapamil (138,142,143), and an unremarkable work-
processing. However, why patients with eCH and up (which includes an MRI brain and, potentially,
cCH respond somewhat differently to nVNS remains MRA head and neck, pituitary hormone testing,
unknown. chest X-ray, and sleep study [144]).
Peng and Burish 9

Pain in CH is primarily felt near the eye in the tri- autonomic reflex, but this may not fully explain how it
geminal nerve distribution, raising the question of how works against headache. In SPG stimulation, low-
stimulation of the GON would be effective. Animal frequency stimulation leads to increased parasympa-
studies suggest a functional link between the trigeminal thetic outflow, whereas high-frequency stimulation, as
system and the C2/3 spinal system, where the GON the paradigm used in SPG stimulation clinical trials,
originates, at the trigeminocervical complex (TCC). leads to decreased parasympathetic outflow (155).
Stimulation of the GON leads to hyperexcitable dural Schytz et al. (155) successfully induced cluster-like
afferents (145) and increased metabolic activity in the attacks in three out of six CH patients with low-
trigeminal nucleus caudalis (146). Anatomically, fibers frequency stimulation and unexpectedly, in one out of
from the C2/C3 ganglion also partially converge on the six patients with high-frequency stimulation. Guo et al.
marginal part of two of the three subnuclei of the spinal (156) repeated low-frequency SPG stimulation in 20
trigeminal nucleus (the nucleus caudalis and nucleus CH patients in a double-blind, randomized, sham-
interpolaris), providing an anatomical basis for cross- controlled, crossover design, and low-frequency stimu-
talk between the two systems at the TCC (147). In lation effectively induced autonomic symptoms but not
healthy humans, stimulation of the GON with capsai- headache. Intranasal kinetic oscillation is another
cin alters the pain thresholds of the dermatomes of all method that effectively triggers the trigeminal auto-
three branches of the trigeminal nerve; reciprocally, nomic reflex and leads to prominent parasympathetic
stimulation at V1 modulates pain thresholds in the output, i.e., lacrimation (125). However, it also failed
occipital region (C2/3 dermatome) (75). Furthermore, to trigger headache attacks in patients with CH (129).
the clinical efficacy of the suboccipital steroid injection Taken together, these two studies support a dissocia-
outlasts the half-life of the injected drug, suggesting a tion between the parasympathetic output and cluster
pain modulatory effect rather than direct pain inhibi- headache triggers (129,156). Therefore, the acute
tion (148), and the TCC is most likely the site of the pain-aborting/inhibitory effect of SPG stimulation,
modulation. and the long-term disease-modulating effect, i.e.,
reduction in attack frequency, may involve a mecha-
Sphenopalatine ganglion simulation nism that is more complex than inhibition of the para-
sympathetic autonomic output. In animal models, the
As mentioned above, the trigeminal autonomic reflex
SPG projects not only to the lacrimal gland and nasal
plays a central role in the pathophysiology of cranial
mucosa via parasympathetic postganglionic fibers, but
autonomic symptoms (117). While vagus nerve stimu-
also directly to the trigeminal ganglion, providing
lation targets several components of the trigeminal
an anatomical basis for direct modulation of the
autonomic reflex, the sphenopalatine ganglion (SPG)
trigeminal nociception by SPG stimulation (157).
stimulation targets the parasympathetic portion of
Alternatively, SPG stimulation shows modulatory
the trigeminal autonomic reflex (149). Like the GON,
effects on cerebral perfusion and alteration of blood-
the SPG is a peripheral structure; peripheral stimula-
brain barrier permeability, but how this may affect
tion carries a lower risk than the deep brain stimula-
pain perception requires further investigation (158).
tion, which will be discussed in the next section. The
efficacy of SPG stimulation as both an acute and pro-
phylactic treatment has been demonstrated in two Deep brain stimulation
randomized, double-blind, sham-controlled trials For patients with medically refractory CH, deep brain
(150,151). In the study by Schoenen et al. (150), 68% stimulation (DBS) of the posterior hypothalamus was
of cCH patients reported clinically significant improve- tried before other neuromodulatory treatment options
ment, defined as a 50% reduction in either attack fre- were available. DBS is based on functional imaging
quency or pain intensity. An American study was studies showing posterior hypothalamic activation
conducted and reported similar efficacy of SPG stimu- during CH attacks, so this site has been considered a
lation in cCH patients (151). Long-term follow-up of potential headache generator in CH (159,160). In small
SPG stimulation showed continued efficacy, 61–68% case series, posterior hypothalamic DBS effectively
remained either acute responders (intensity reduction) reduced the attack intensity and attack frequency
or preventive responders (frequency reduction) (161–163). One double-blind, randomized, sham-
(152,153). Despite the well-validated clinical efficacy, controlled study (164) failed to show the superiority
the company that manufactured the SPG stimulator of DBS during the one-month randomized phase, but
had financial issues and the stimulator is currently at the end of the open-label phase (one year after the
unavailable (154). randomized phase), 6/11 had a good response, similar
The mechanism of SPG stimulation is not fully to the previous studies (161–163). However, the poten-
understood. SPG stimulation modulates the trigeminal tially fatal risk of intracranial hemorrhage renders this
10 Cephalalgia

central nervous system treatment option less preferable Conclusion


to the peripheral nervous system alternatives such as
Common first-line treatments for CH – subcutaneous
SPG and GON stimulation (163). DBS is currently
sumatriptan, oxygen, prednisone, suboccipital steroid
rarely used to treat refractory CH. Nevertheless, the
injections, and verapamil – remain the mainstay for
potential success of hypothalamic DBS in the treat- treatment of this disorder. The mechanisms of these
ment of CH provides proof of concept that this treatments, either how they work or where they work,
region may be directly involved in the generation of are still incompletely understood. Additionally, the
cluster attacks. A neurophysiological study in patients advent of neuromodulation and the introduction of
receiving hypothalamic DBS showed that DBS anti-CGRP treatment have brought new opportunities
increased the thermal pain threshold in dermatome for CH patients. However, patients still suffer from
V1 ipsilateral to the stimulation, suggesting a direct inadequate access to these treatment options. In addi-
pain modulatory effect of DBS (165). In addition, tion, some patients, particularly those with cCH,
both nVNS and CGRP-mAbs showed (possibly indi- respond poorly to existing treatment options. A better
rect) hypothalamic modulatory effects (109,110,132), understanding of the biological differences between eCH
further supporting the hypothalamus as a potential and cCH is essential to unravel the factors that deter-
treatment target. mine clinical response.

Clinical implications
• Cluster headache is treated with acute treatments (to manage an ongoing attack), bridge treatments (short-
term preventives to be used while uptitrating preventives), and long-term preventive treatments.
• Common first-line treatments include fast-acting triptans and high-flow oxygen as acute treatments, cor-
ticosteroids (oral or suboccipital injections) as bridge treatments, and verapamil as a long-term preventive
treatment.
• The mechanisms of oxygen, corticosteroids and verapamil are uncertain, whereas those of triptans and
anti-CGRP treatments are clear but their site of action remains controversial. The mechanisms of neuro-
stimulation devices are thought to target parts of the cluster headache pathways.

Declaration of conflicting interests Mark J. Burish https://orcid.org/0000-0002-8931-6436


The authors declared the following potential conflicts of
interest with respect to the research, authorship, and/or pub- References
lication of this article: 1. Jürgens TP, Koch HJ, May A. Ten years of chronic
K-PP has received honoraria as speaker from TEVA and serves cluster–attacks still cluster. Cephalalgia 2010; 30:
as Associate Editor of Cephalalgia and Cephalalgia Reports. 1123–1126.
MJB was an unpaid medical advisor for Praxis Precision 2. Barloese M, Lund N, Petersen A, et al. Sleep and chro-
Medicines (in lieu of compensation a fee was paid to the nobiology in cluster headache. Cephalalgia 2015; 35:
University of Texas Health Science Center at Houston) and 969–78.
was an unpaid consultant for Beckley Psytech limited (in lieu 3. Medrea I, Christie S, Tepper SJ, et al. Network meta-
of compensation a donation was made to the Will Erwin analysis of therapies for cluster headache: Effects of
Headache Research Foundation). It should be noted, howev- acute therapies for episodic and chronic cluster.
er, that Dr. Burish is employed by the University of Texas Headache 2022; 62: 482–511.
Health Science Center at Houston and receives research fund- 4. Robbins MS, Starling AJ, Pringsheim TM, et al.
ing from the Will Erwin Headache Research Foundation. He Treatment of cluster headache: The American
is an unpaid member of the medical advisory board of Headache Society Evidence-Based Guidelines.
Clusterbusters, and was a site investigator for a cluster head- Headache 2016; 56: 1093–1106.
ache clinical trial funded by Lundbeck. 
5. Evers S, Afra J, Frese A, et al. Cluster headache and
other trigemino-autonomic cephalgias. In: NE Gilhus,
Funding MP Barnes, M Brainin (eds) European Handbook of
Neurological Management. Vol 1. 2nd ed: Chichester,
The authors received no financial support for the research,
UK: Blackwell Publishing Ltd, 2011, pp.179–190.
authorship, and/or publication of this article.
6. Ahn AH, Basbaum AI. Where do triptans act in the
treatment of migraine? Pain 2005; 115: 1–4.
ORCID iDs 7. Group SCHS. Treatment of acute cluster headache with
Kuan-Po Peng https://orcid.org/0000-0002-2718-3738 sumatriptan. New Eng J Med 1991; 325: 322–6.
Peng and Burish 11

8. Ekbom K, Monstad I, Prusinski A, et al. Subcutaneous 26. Tepper SJ, Rapoport AM, Sheftell FD. Mechanisms of
sumatriptan in the acute treatment of cluster headache: action of the 5-HT1B/1D receptor agonists. Arch Neurol
a dose comparison study. The Sumatriptan Cluster 2002; 59: 1084–1088.
Headache Study Group. Acta Neurol Scand 1993; 88: 27. Ala M, Ghasemi M, Mohammad Jafari R, et al. Beyond
63–69. its anti-migraine properties, sumatriptan is an anti-
9. van Vliet JA, Bahra A, Martin V, et al. Intranasal suma- inflammatory agent: A systematic review. Drug Dev
triptan in cluster headache: randomized placebo-controlled Res 2021; 82: 896–906.
double-blind study. Neurology 2003; 60: 630–633. 28. Lionetto L, Negro A, Casolla B, et al. Sumatriptan suc-
10. Cittadini E, May A, Straube A, et al. Effectiveness of cinate: pharmacokinetics of different formulations in
intranasal zolmitriptan in acute cluster headache: a ran- clinical practice. Expert Opin Pharmacother 2012; 13:
domized, placebo-controlled, double-blind crossover 2369–2380.
study. Arch Neurol 2006; 63: 1537–1542. 29. Vila-Pueyo M, Page K, Murdock PR, et al. The selective
11. Rapoport AM, Mathew NT, Silberstein SD, et al. 5-HT(1F) receptor agonist lasmiditan inhibits trigeminal
Zolmitriptan nasal spray in the acute treatment of clus- nociceptive processing: Implications for migraine and
ter headache: a double-blind study. Neurology 2007; 69: cluster headache. Br J Pharmacol 2022; 179: 358–370.
821–826. 30. Leone M, Proietti Cecchini A. Long-term use of daily
12. Bahra A, Gawel MJ, Hardebo JE, et al. Oral zolmitrip- sumatriptan injections in severe drug-resistant chronic
tan is effective in the acute treatment of cluster head- cluster headache. Neurology 2016; 86: 194–195.
ache. Neurology 2000; 54: 1832–1839. 31. Wenzel RG, Tepper S, Korab WE, et al. Serotonin
13. Marmura MJ, Silberstein SD, Schwedt TJ. The acute syndrome risks when combining SSRI/SNRI drugs
treatment of migraine in adults: the american headache and triptans: is the FDA’s alert warranted? Ann
society evidence assessment of migraine pharmacothera- Pharmacother 2008; 42: 1692–1696.
pies. Headache 2015; 55: 3–20. 32. Orlova Y, Rizzoli P, Loder E. Association of copre-
14. Gobel H, Lindner V, Heinze A, et al. Acute therapy for scription of triptan antimigraine drugs and selective
cluster headache with sumatriptan: findings of a one- serotonin reuptake inhibitor or selective norepinephrine
year long-term study. Neurology 1998; 51: 908–911. reuptake inhibitor antidepressants with serotonin syn-
15. Gregor N, Schlesiger C, Akova-Ozturk E, et al. Treatment drome. JAMA Neurol 2018; 75: 566–572.
of cluster headache attacks with less than 6 mg subcutane- 33. Edvinsson L. Blockade of CGRP receptors in the intra-
ous sumatriptan. Headache 2005; 45: 1069–1072. cranial vasculature: a new target in the treatment of
16. Ekbom K, Krabbe A, Micieli G, et al. Cluster headache headache. Cephalalgia 2004; 24: 611–622.
attacks treated for up to three months with subcutane- 34. Costa A, Antonaci F, Ramusino MC, et al. The neuro-
ous sumatriptan (6 mg). Sumatriptan Cluster Headache pharmacology of cluster headache and other trigeminal
Long-term Study Group. Cephalalgia 1995; 15: 230–236. autonomic cephalalgias. Curr Neuropharmacol 2015; 13:
17. Centonze V, Bassi A, Causarano V, et al. Sumatriptan 304–323.
overuse in episodic cluster headache: lack of adverse 35. Evans MS, Cheng X, Jeffry JA, et al. Sumatriptan
events, rebound syndromes, drug dependence and inhibits TRPV1 channels in trigeminal neurons.
tachyphylaxis. Funct Neurol 2000; 15: 167–170. Headache 2012; 52: 773–784.
18. Yalinay Dikmen P, Ari C, et al. Cluster analysis 36. Dussor G. Serotonin, 5HT1 agonists, and migraine: new
revealed two hidden phenotypes of cluster headache. data, but old questions still not answered. Curr Opin
Front Neurol 2022; 13: 898022. Support Palliat Care 2014; 8: 137–142.
19. Giani L, Cecchini AP, Astengo A, et al. Cluster head- 37. Classey JD, Bartsch T, Goadsby PJ. Distribution of
ache not responsive to sumatriptan: A retrospective 5-HT(1B), 5-HT(1D) and 5-HT(1F) receptor expression
study. Cephalalgia 2021; 41: 117–121. in rat trigeminal and dorsal root ganglia neurons: rele-
20. Schurks M, Rosskopf D, de Jesus J, et al. Predictors of vance to the selective anti-migraine effect of triptans.
acute treatment response among patients with cluster Brain Res 2010; 1361: 76–85.
headache. Headache 2007; 47: 1079–1084. 38. Edvinsson JCA, Maddahi A, Christiansen IM, et al.
21. Schurks M, Kurth T, Stude P, et al. G protein beta3 Lasmiditan and 5-Hydroxytryptamine in the rat trigem-
polymorphism and triptan response in cluster headache. inal system; expression, release and interactions with
Clin Pharmacol Ther 2007; 82: 396–401. 5-HT(1) receptors. J Headache Pain 2022; 23: 26.
22. Paemeleire K, Evers S, Goadsby PJ. Medication-over- 39. Lanfumey L, Hamon M. 5-HT1 receptors. Curr Drug
use headache in patients with cluster headache. Curr Targets CNS Neurol Disord 2004; 3: 1–10.
Pain Headache Rep 2008; 12: 122–127. 40. Muzzi M, Zecchi R, Ranieri G, et al. Ultra-rapid brain
23. Paemeleire K, Bahra A, Evers S, et al. Medication-over- uptake of subcutaneous sumatriptan in the rat:
use headache in patients with cluster headache. Implication for cluster headache treatment. Cephalalgia
Neurology 2006; 67: 109–113. 2020; 40: 330–336.
24. Tepper SJ, Millson D. Safety profile of the triptans. 41. Peng KP, Jürgens T, Basedau H, et al. Sumatriptan
Expert Opin Drug Saf 2003; 2: 123–132. prevents central sensitization specifically in the trigemi-
25. Ferrari MD, Goadsby PJ, Burstein R, et al. Migraine. nal dermatome in humans. Eur J Pain 2022; 26:
Nat Rev Dis Primers 2022; 8: 2. 2152–2161.
12 Cephalalgia

42. Matharu MS, Goadsby PJ. Persistence of attacks of 59. Goadsby PJ, Edvinsson L. Human in vivo evidence for
cluster headache after trigeminal nerve root section. trigeminovascular activation in cluster headache.
Brain. 2002;125:976–84. Neuropeptide changes and effects of acute attacks ther-
43. Haane DY, Dirkx TH, Koehler PJ. The history of apies. Brain 1994; 117: 427–434.
oxygen inhalation as a treatment for cluster headache. 60. Schuh-Hofer S, Siekmann W, Offenhauser N, et al.
Cephalalgia 2012; 32: 932–939. Effect of hyperoxia on neurogenic plasma protein
44. Cohen AS, Burns B, Goadsby PJ. High-flow oxygen for extravasation in the rat dura mater. Headache 2006;
treatment of cluster headache: a randomized trial. 46: 1545–1551.
JAMA 2009; 302: 2451–2457. 61. Obermann M, Nagel S, Ose C, et al. Safety and efficacy
45. Dirkx THT, Haane DYP, Koehler PJ. Oxygen treat- of prednisone versus placebo in short-term prevention
ment for cluster headache attacks at different flow of episodic cluster headache: a multicentre, double-
rates: a double-blind, randomized, crossover study. J blind, randomised controlled trial. Lancet Neurol 2021;
Headache Pain 2018; 19: 94. 20: 29–37.
46. Fogan L. Treatment of cluster headache. A double-blind 62. Anthony M. Arrest of attacks of cluster headache by
comparison of oxygen v air inhalation. Arch Neurol 1985; local steroid injection of the occipital nerve. In: Rose,
42: 362–363. F. Clifford (ed.) Migraine: Clinical and Research
47. Tepper SJ, Duplin J, Nye B, et al. Prescribing oxygen Advances. 5th International Symposium. Basel, New
for cluster headache: a guide for the provider. Headache York: Karger, 1984, pp.169–173
2017; 57: 1428–1430. 63. Ambrosini A, Vandenheede M, Rossi P, et al.
48. Petersen AS, Barloese MC, Lund NL, et al. Oxygen Suboccipital injection with a mixture of rapid- and
therapy for cluster headache. A mask comparison long-acting steroids in cluster headache: A double-
trial. A single-blinded, placebo-controlled, crossover blind placebo-controlled study. Pain 2005; 118: 92–96.
study. Cephalalgia 2017; 37: 214–224. 64. Leroux E, Valade D, Taifas I, et al. Suboccipital steroid
49. Rozen TD, Fishman RS. Inhaled oxygen and cluster injections for transitional treatment of patients with
headache sufferers in the United States: use, efficacy more than two cluster headache attacks per day: a rand-
and economics: results from the United States Cluster omised, double-blind, placebo-controlled trial. Lancet
Headache Survey. Headache 2011; 51: 191–200. Neurol 2011; 10: 891–897.
50. Schor LI, Pearson SM, Shapiro RE, et al. Cluster head- 65. Afridi SK, Shields KG, Bhola R, et al. Greater occipital
ache epidemiology including pediatric onset, sex, and nerve injection in primary headache syndromes – pro-
ICHD criteria: Results from the International Cluster longed effects from a single injection. Pain 2006; 122:
Headache Questionnaire. Headache 2021; 61: 1511–1520. 126–129.
51. Rozen TD. Inhaled oxygen for cluster headache: effica- 66. Gantenbein AR, Lutz NJ, Riederer F, et al. Efficacy
cy, mechanism of action, utilization, and economics. and safety of 121 injections of the greater occipital
Curr Pain Headache Rep 2012; 16: 175–179. nerve in episodic and chronic cluster headache.
52. Geerlings RP, Haane DY, Koehler PJ. Rebound follow- Cephalalgia 2012; 32: 630–634.
ing oxygen therapy in cluster headache. Cephalalgia 67. Lambru G, Abu Bakar N, Stahlhut L, et al. Greater
2011; 31: 1145–1149. occipital nerve blocks in chronic cluster headache: a
53. Wei DY, Goadsby PJ. Cluster headache pathophysiol- prospective open-label study. Eur J Neurol 2014; 21:
ogy - insights from current and emerging treatments. 338–343.
Nat Rev Neurol 2021; 17: 308–324. 68. Wei J, Robbins MS. Greater occipital nerve injection
54. Oude Nijhuis JC, Haane DY, Koehler PJ. A review of versus oral steroids for short term prophylaxis of cluster
the current and potential oxygen delivery systems and headache: a retrospective comparative study. Headache
techniques utilized in cluster headache attacks. 2018; 58: 852–858.
Cephalalgia 2016; 36: 970–979. 69. Gaul C, Roguski J, Dresler T, et al. Efficacy and safety
55. Abdo WF, Heunks LM. Oxygen-induced hypercapnia of a single occipital nerve blockade in episodic and
in COPD: myths and facts. Crit Care 2012; 16: 323. chronic cluster headache: A prospective observational
56. Mo H, Chung SJ, Rozen TD, et al. Oxygen therapy in study. Cephalalgia 2017; 37: 873–880.
cluster headache, migraine, and other headache disor- 70. Shields KG, Levy MJ, Goadsby PJ. Alopecia and cuta-
ders. J Clin Neurol 2022; 18: 271–279. neous atrophy after greater occipital nerve infiltration
57. Akerman S, Holland PR, Lasalandra MP, et al. Oxygen with corticosteroid. Neurology 2004; 63: 2193–2194.
inhibits neuronal activation in the trigeminocervical 71. Benkli B, Kim SY, Koike N, et al. Circadian features of
complex after stimulation of trigeminal autonomic cluster headache and migraine: a systematic review,
reflex, but not during direct dural activation of trigem- meta-analysis, and genetic analysis. Neurology 2023;
inal afferents. Headache 2009; 49: 1131–1143. 100: e2224–e2236.
58. Schroder CF, Basedau H, Moeller M, et al. Oxygen 72. Neeb L, Anders L, Euskirchen P, et al. Corticosteroids
inhalation has no effect on provoked cranial autonomic alter CGRP and melatonin release in cluster headache
symptoms using kinetic oscillation stimulation in episodes. Cephalalgia 2015; 35: 317–326.
healthy volunteers. Cephalalgia 2023; 43. DOI: 73. Figuerola ML, Levin G, Leston J, et al. Opioid and
10.1177/3331024231161269. sympathetic nervous system activity in cluster headache
Peng and Burish 13

under verapamil or prednisone treatment. Headache 90. Goadsby PJ, Dodick DW, Leone M, et al. Trial of gal-
1994; 34: 257–260. canezumab in prevention of episodic cluster headache.
74. Neeb L, Hellen P, Hoffmann J, et al. New Eng J Med 2019; 381: 132–141.
Methylprednisolone blocks interleukin 1 beta induced 91. Dodick DW, Goadsby PJ, Lucas C, et al. Phase 3 ran-
calcitonin gene related peptide release in trigeminal gan- domized, placebo-controlled study of galcanezumab in
glia cells. J Headache Pain 2016; 17: 19. patients with chronic cluster headache: Results from 3-
75. Basedau H, Nielsen T, Asmussen K, et al. Experimental month double-blind treatment. Cephalalgia 2020; 40:
evidence of a functional relationship within the brain- 935–948.
stem trigeminocervical complex in humans. Pain 2022; 92. Membrilla JA, Torres-Ferrus M, Alpuente A, et al.
163: 729–734. Efficacy and safety of galcanezumab as a treatment of
76. Busch V, Jakob W, Juergens T, et al. Functional con- refractory episodic and chronic cluster headache: Case
nectivity between trigeminal and occipital nerves series and narrative review. Headache 2022; 62:
revealed by occipital nerve blockade and nociceptive 1395–1405.
blink reflexes. Cephalalgia 2006; 26: 50–55. 93. Mo H, Kim B-K, Moon H-S, et al. Real-world experi-
77. Petersen AS, Barloese MCJ, Snoer A, et al. Verapamil ence with 240 mg of galcanezumab for the preventive
and cluster headache: still a mystery. a narrative review treatment of cluster headache. J Headache Pain 2022;
of efficacy, mechanisms and perspectives. Headache 23: 132.
2019; 59: 1198–1211. 94. Ruscheweyh R, Broessner G, Goßrau G, et al. Effect of
78. Bussone G, Leone M, Peccarisi C, et al. Double blind calcitonin gene-related peptide (-receptor) antibodies in
comparison of lithium and verapamil in cluster head- chronic cluster headache: Results from a retrospective
ache prophylaxis. Headache 1990; 30: 411–417. case series support individual treatment attempts.
79. Leone M, D’Amico D, Frediani F, et al. Verapamil in the Cephalalgia 2020; 40: 1574–1584.
prophylaxis of episodic cluster headache: a double-blind 95. Alpuente A, Gallardo VJ, Asskour L, et al. Salivary CGRP
study versus placebo. Neurology 2000; 54: 1382–1385. can monitor the different migraine phases: CGRP (in)
80. Lademann V, Jansen JP, Evers S, et al. Evaluation of dependent attacks. Cephalalgia 2021; 42: 186–196.
guideline-adherent treatment in cluster headache. 96. Pellesi L, Chaudhry BA, Vollesen ALH, et al.
Cephalalgia 2016; 36: 760–764. PACAP38- and VIP-induced cluster headache attacks
81. Cohen AS, Matharu MS, Goadsby PJ. are not associated with changes of plasma CGRP or
Electrocardiographic abnormalities in patients with markers of mast cell activation. Cephalalgia 2022; 42:
cluster headache on verapamil therapy. Neurology 687–695.
2007; 69: 668–675. 97. Snoer A, Vollesen ALH, Beske RP, et al. Calcitonin-
82. Tfelt-Hansen P, Tfelt-Hansen J. Verapamil for cluster gene related peptide and disease activity in cluster head-
headache. Clinical pharmacology and possible mode of ache. Cephalalgia 2019; 39: 575–584.
action. Headache 2009; 49: 117–125. 98. Durham PL, Russo AF. Regulation of calcitonin gene-
83. Lanteri-Minet M, Silhol F, Piano V, et al. Cardiac related peptide secretion by a serotonergic antimigraine
safety in cluster headache patients using the very high drug. Neuroscience 1999; 19: 3423–3429.
dose of verapamil (>/¼720 mg/day). J Headache Pain 99. Goadsby PJ, Edvinsson L. The trigeminovascular
2011; 12: 173–176. system and migraine: Studies characterizing cerebrovas-
84. Barloese M, Haddock B, Lund NT, et al. Chronorisk in cular and neuropeptide changes seen in humans and
cluster headache: A tool for individualised therapy? cats. Ann Neurol 1993; 33: 48–56.
Cephalalgia 2018; 38: 2058–2067. 100. Russo AF, Hay DL. CGRP physiology, pharmacology,
85. Burish MJ, Han C, Mawatari K, et al. The first-line and therapeutic targets: migraine and beyond. Physiol
cluster headache medication verapamil alters the circa- Rev 2023; 103: 1565–1644.
dian period and elicits sex-specific sleep changes in mice. 101. Yuan H, Spare NM, Silberstein SD. Targeting CGRP
Chronobiol Int 2021; 38: 839–850. for the prevention of migraine and cluster headache: a
86. Goadsby PJ, Edvinsson L, Ekman R. Vasoactive peptide narrative review. Headache 2019; 59: 20–32.
release in the extracerebral circulation of humans during 102. Reuter U, Ehrlich M, Gendolla A, et al. Erenumab
migraine headache. Ann Neurol 1990; 28: 183–187. versus topiramate for the prevention of migraine – a
87. Vollesen ALH, Snoer A, Beske RP, et al. Effect of infu- randomised, double-blind, active-controlled phase 4
sion of calcitonin gene-related peptide on cluster head- trial. Cephalalgia 2022; 42: 108–118.
ache attacks: a randomized clinical trial. JAMA Neurol 103. Depre C, Antalik L, Starling A, et al. A randomized,
2018; 75: 1187–1197. double-blind, placebo-controlled study to evaluate the
88. Iljazi A, Ashina H, Zhuang ZA, et al. Hypersensitivity effect of erenumab on exercise time during a treadmill
to calcitonin gene-related peptide in chronic migraine. test in patients with stable angina. Headache 2018; 58:
Cephalalgia 2021; 41: 701–710. 715–723.
89. Edvinsson L, Haanes KA, Warfvinge K, et al. CGRP as 104. Basedau H, Sturm LM, Mehnert J, et al. Migraine
the target of new migraine therapies — successful trans- monoclonal antibodies against CGRP change brain
lation from bench to clinic. Nat Rev Neurol 2018; 14: activity depending on ligand or receptor target - an
338–350. fMRI study. eLife 2022; 11: e77146.
14 Cephalalgia

105. Edvinsson L. The CGRP pathway in migraine as a double-blind, sham-controlled ACT1 study. Headache
viable target for therapies. Headache 2018; 58: 33–47. 2016; 56: 1317–1332.
106. Johnson KW, Morin SM, Wroblewski VJ, et al. 121. Nesbitt AD, Marin JC, Tompkins E, et al. Initial use of
Peripheral and central nervous system distribution of a novel noninvasive vagus nerve stimulator for cluster
the CGRP neutralizing antibody [(125)I] galcanezumab headache treatment. Neurology 2015; 84: 1249–1253.
in male rats. Cephalalgia 2019; 39: 1241–1248. 122. Gaul C, Diener H-C, Silver N, et al. Non-invasive vagus
107. Noseda R, Schain AJ, Melo-Carrillo A, et al. nerve stimulation for PREVention and Acute treatment
Fluorescently-labeled fremanezumab is distributed to of chronic cluster headache (PREVA): A randomised
sensory and autonomic ganglia and the dura but not controlled study. Cephalalgia 2016; 36: 534–546.
to the brain of rats with uncompromised blood brain 123. Marin J, Giffin N, Consiglio E, et al. Non-invasive
barrier. Cephalalgia 2020; 40: 229–240. vagus nerve stimulation for treatment of cluster head-
108. Lin H, Dodick DW. Tearing without pain after trigem- ache: early UK clinical experience. J Headache Pain
inal root section for cluster headache. Neurology 2005; 2018; 19: 114.
65: 1650–1651. 124. Johnson RL, Wilson CG. A review of vagus nerve stim-
109. Basedau H, Sturm L-M, Mehnert J, et al. Migraine ulation as a therapeutic intervention. J Inflamm Res
monoclonal antibodies against CGRP change brain 2018; 11: 203–213.
activity depending on ligand or receptor target – an 125. M€ oller M, Schroeder CF, May A. Vagus nerve stimula-
fMRI study. eLife 2022; 11: e77146. tion modulates the cranial trigeminal autonomic reflex.
110. Ziegeler C, Mehnert J, Asmussen K, et al. Central Ann Neurol 2018; 84: 886–892.
effects of erenumab in migraine patients: An event- 126. Akerman S, Simon B, Romero-Reyes M. Vagus nerve
related functional imaging study. Neurology 2020; 95: stimulation suppresses acute noxious activation of trige-
e2794–e2802. minocervical neurons in animal models of primary
111. Casillo F, Sebastianelli G, Di Renzo A, et al. The mono- headache. Neurobiol Disease 2017; 102: 96–104.
clonal CGRP-receptor blocking antibody erenumab has 127. Peng K-P, May A. Noninvasive vagus nerve stimulation
different effects on brainstem and cortical sensory- modulates trigeminal but not extracephalic somatosen-
evoked responses. Cephalalgia 2022; 42: 1236–1245. sory perception: functional evidence for a trigemino-
112. de Tommaso M, Delussi M, Gentile E, et al. Effect of vagal system in humans. Pain 2022; 163: 1978–1986.
single dose Erenumab on cortical responses evoked by 128. Oshinsky ML, Murphy AL, Hekierski H, Jr., et al.
cutaneous a-delta fibers: A pilot study in migraine Noninvasive vagus nerve stimulation as treatment for
patients. Cephalalgia 2021; 41: 1004–1014. trigeminal allodynia. Pain 2014; 155: 1037–1042.
113. McKnight RF, Adida M, Budge K, et al. Lithium tox- 129. M€ oller M, Haji AA, Hoffmann J, et al. Peripheral prov-
icity profile: a systematic review and meta-analysis. ocation of cranial autonomic symptoms is not sufficient
Lancet 2012; 379: 721–728. to trigger cluster headache attacks. Cephalalgia 2018;
114. Nikolova VL, Pattanaseri K, Hidalgo-Mazzei D, et al. 38: 1498–1502.
Is lithium monitoring NICE? Lithium monitoring in a 130. Lai Y-H, Huang Y-C, Huang L-T, et al. Cervical non-
UK secondary care setting. J Psychopharmacol 2018; 32: invasive vagus nerve stimulation for migraine and clus-
408–415. ter headache: a systematic review and meta-analysis.
115. Leone M, D’Amico D, Moschiano F, et al. Melatonin Neuromodulation 2020; 23: 721–731.
versus placebo in the prophylaxis of cluster headache: a 131. Villar-Martinez MD, Goadsby PJ. Non-invasive neuro-
double-blind pilot study with parallel groups. modulation of the cervical vagus nerve in rare primary
Cephalalgia 1996; 16: 494–496. headaches. Front Pain Res (Lausanne) 2023; 4: 1062892.
116. Burish MJ, Chen Z, Yoo SH. Emerging relevance of 132. Moller M, Mehnert J, Schroeder CF, et al. Noninvasive
circadian rhythms in headaches and neuropathic pain. vagus nerve stimulation and the trigeminal autonomic
Acta Physiol (Oxf) 2019; 225: e13161. reflex: An fMRI study. Neurology 2020; 94: e1085–e1093.
117. Moller M, May A. The unique role of the trigeminal 133. Goadsby P, May A. PET demonstration of hypotha-
autonomic reflex and its modulation in primary head- lamic activation in cluster headache. Neurology 1999;
ache disorders. Curr Opin Neurol 2019; 32: 438–442. 52: 1522.
118. Mauskop A. Vagus nerve stimulation relieves chronic 134. Burns B, Watkins L, Goadsby PJ. Treatment of medi-
refractory migraine and cluster headaches. Cephalalgia cally intractable cluster headache by occipital nerve
2005; 25: 82–86. stimulation: long-term follow-up of eight patients.
119. Goadsby PJ, de Coo IF, Silver N, et al. Non-invasive Lancet 2007; 369: 1099–1106.
vagus nerve stimulation for the acute treatment of epi- 135. Fontaine D, Blond S, Lucas C, et al. Occipital nerve stim-
sodic and chronic cluster headache: A randomized, ulation improves the quality of life in medically-intractable
double-blind, sham-controlled ACT2 study. Cephalalgia chronic cluster headache: Results of an observational pro-
2018; 38: 959–969. spective study. Cephalalgia 2017; 37: 1173–1179.
120. Silberstein SD, Mechtler LL, Kudrow DB, et al. Non– 136. Fontaine D, Christophe Sol J, Raoul S, et al. Treatment
invasive vagus nerve stimulation for the acute treatment of refractory chronic cluster headache by chronic occip-
of cluster headache: findings from the randomized, ital nerve stimulation. Cephalalgia 2011; 31: 1101–1105.
Peng and Burish 15

137. Magis D, Gerard P, Schoenen J. Invasive occipital nerve 151. Goadsby PJ, Sahai-Srivastava S, Kezirian EJ, et al.
stimulation for refractory chronic cluster headache: Safety and efficacy of sphenopalatine ganglion stimula-
what evolution at long-term? Strengths and weaknesses tion for chronic cluster headache: a double-blind,
of the method. J Headache Pain 2016; 17: 8. randomised controlled trial. Lancet Neurol 2019; 18:
138. Wilbrink LA, de Coo IF, Doesborg PGG, et al. Safety 1081–1090.
and efficacy of occipital nerve stimulation for attack 152. Barloese M, Petersen A, Stude P, et al. Sphenopalatine
prevention in medically intractable chronic cluster head- ganglion stimulation for cluster headache, results from a
ache (ICON): a randomised, double-blind, multicentre, large, open-label European registry. J Headache Pain
phase 3, electrical dose-controlled trial. Lancet Neurol 2018; 19: 6.
2021; 20: 515–525. 153. Jürgens TP, Barloese M, May A, et al. Long-term effec-
139. Leone M, Proietti Cecchini A, Messina G, et al. Long- tiveness of sphenopalatine ganglion stimulation for clus-
term occipital nerve stimulation for drug-resistant ter headache. Cephalalgia 2017; 37: 423–434.
chronic cluster headache. Cephalalgia 2017; 37: 756–763. 154. May A. The need for continued care after sponsor clo-
140. Miller S, Watkins L, Matharu M. Predictors of response sure. Lancet Neurol 2020; 19: 205.
to occipital nerve stimulation in refractory chronic 155. Schytz HW, Barlose M, Guo S, et al. Experimental acti-
headache. Cephalalgia 2018; 38: 1267–1275. vation of the sphenopalatine ganglion provokes cluster-
141. Kinfe TM, Schuss P, Vatter H. Occipital nerve block like attacks in humans. Cephalalgia 2013; 33: 831–841.
prior to occipital nerve stimulation for refractory chron- 156. Guo S, Petersen AS, Schytz HW, et al. Cranial para-
ic migraine and chronic cluster headache: myth or pre- sympathetic activation induces autonomic symptoms
diction? Cephalalgia 2015; 35: 359–362. but no cluster headache attacks. Cephalalgia 2018; 38:
142. Leone M, May A, Franzini A, et al. Deep brain stimu- 1418–1428.
lation for intractable chronic cluster headache: pro- 157. Ivanusic JJ, Kwok MMK, Ahn AH, et al. 5-HT(1D)
receptor immunoreactivity in the sphenopalatine gangli-
posals for patient selection. Cephalalgia 2004; 24:
on: implications for the efficacy of triptans in the treat-
934–937.
ment of autonomic signs associated with cluster
143. Leone M, Franzini A, Cecchini AP, et al. Stimulation of
headache. Headache 2011; 51: 392–402.
occipital nerve for drug-resistant chronic cluster head-
158. Levi H, Schoknecht K, Prager O, et al. Stimulation of
ache. Lancet Neurol 2007; 6: 289–291.
the sphenopalatine ganglion induces reperfusion and
144. Mitsikostas DD, Ashina M, Craven A, et al. European
blood-brain barrier protection in the photothrombotic
Headache Federation consensus on technical investiga-
stroke model. PLOS ONE 2012; 7: e39636.
tion for primary headache disorders. J Headache Pain
159. May A, Bahra A, Büchel C, et al. Hypothalamic acti-
2015; 17: 5.
vation in cluster headache attacks. Lancet 1998; 352:
145. Bartsch T, Goadsby PJ. Stimulation of the greater
275–278.
occipital nerve induces increased central excitability of
160. Schulte LH, Haji AA, May A. Phase dependent hypo-
dural afferent input. Brain 2002; 125: 1496–1509. thalamic activation following trigeminal input in cluster
146. Goadsby PJ, Knight YE, Hoskin KL. Stimulation of the headache. J Headache Pain 2020; 21: 30.
greater occipital nerve increases metabolic activity in the 161. Franzini A, Ferroli P, Leone M, et al. Stimulation of the
trigeminal nucleus caudalis and cervical dorsal horn of posterior hypothalamus for treatment of chronic intrac-
the cat. Pain 1997; 73: 23–28. table cluster headaches: first reported series.
147. Garcıa-Magro N, Martin YB, Negredo P, et al. The Neurosurgery 2003; 52: 1095-1099; discussion 9-101.
greater occipital nerve and its spinal and brainstem 162. Leone M, Franzini A, Bussone G. Stereotactic stimula-
afferent projections: A stereological and tract-tracing tion of posterior hypothalamic gray matter in a patient
study in the rat. J Comp Neurol 2018; 526: 3000–3019. with intractable cluster headache. N Engl J Med 2001;
148. Bartsch T, Goadsby PJ. The trigeminocervical complex 345: 1428–1429.
and migraine: current concepts and synthesis. Curr Pain 163. Schoenen J. Hypothalamic stimulation in chronic clus-
Headache Rep 2003; 7: 371–376. ter headache: a pilot study of efficacy and mode of
149. Li C, Fitzgerald MEC, Del Mar N, et al. Disinhibition action. Brain 2005; 128: 940–947.
of neurons of the nucleus of solitary tract that project to 164. Fontaine D, Lazorthes Y, Mertens P, et al. Safety and
the superior salivatory nucleus causes choroidal vasodi- efficacy of deep brain stimulation in refractory cluster
lation: Implications for mechanisms underlying choroi- headache: a randomized placebo-controlled double-
dal baroregulation. Neurosci Lett 2016; 633: 106–111. blind trial followed by a 1-year open extension.
150. Schoenen J, Jensen RH, Lanteri-Minet M, et al. J Headache Pain 2010; 11: 23–31.
Stimulation of the sphenopalatine ganglion (SPG) for 165. Jürgens TP, Leone M, Proietti-Cecchini A, et al.
cluster headache treatment. Pathway CH-1: a random- Hypothalamic deep-brain stimulation modulates ther-
ized, sham-controlled study. Cephalalgia 2013; 33: mal sensitivity and pain thresholds in cluster headache.
816–830. Pain 2009; 146: 84–90.
Copyright of Cephalalgia is the property of Sage Publications Inc. and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.

You might also like