Cluster headache: Treatment and prognosis

INTRODUCTION — Cluster headache belongs to the trigeminal autonomic cephalalgias (TACs), a group of idiopathic headache disorders characterized by unilateral headaches accompanied by ipsilateral cranial autonomic symptoms.

The TACs include cluster headache, paroxysmal hemicrania, short-lasting unilateral neuralgiform headache attacks, and hemicrania continua. Cluster headache is the most common form.

The treatment and prognosis of cluster headache are discussed here. The epidemiology, clinical features, and diagnosis of cluster headache are discussed separately. (See "Cluster headache: Epidemiology, clinical features, and diagnosis".)

Other TACs are discussed elsewhere.

●(See "Paroxysmal hemicrania: Clinical features and diagnosis" and "Paroxysmal hemicrania: Treatment and prognosis".)

●(See "Short-lasting unilateral neuralgiform headache attacks: Clinical features and diagnosis" and "Short-lasting unilateral neuralgiform headache attacks: Treatment and prognosis".)

●(See "Hemicrania continua".)

APPROACH TO THERAPY — Acute therapy is useful for aborting individual attacks of cluster headache but does not reduce the duration of the cluster headache bout (ie, the period of recurrent attacks) [1]. Therefore, preventive therapy should be started without delay once a cluster headache bout begins.

Currently, the treatment of cluster headache is based on empirical data rather than a clear understanding of the biologic mechanisms that underlie the disease [2,3]. Drug treatment in cluster headache shows a placebo rate of about 30 percent, similar to that observed in migraine treatment [4]. In general, cluster headache treatment can be divided into acute therapy aimed at aborting individual attacks and prophylactic therapy aimed at preventing recurrent attacks during the cluster period (table 1) [5].

Acute headache treatment — For patients with acute cluster headache, we recommend initial treatment with either 100 percent oxygen or a triptan, in agreement with national guidelines and expert consensus (algorithm 1) [1,6].

●Oxygen should be tried first if available (eg, in a hospital or emergency clinic setting) since it is without side effects (see 'Oxygen' below)

●Otherwise, a triptan such as sumatriptan or zolmitriptan can be used as initial therapy (see 'Triptans' below)

For patients taking a triptan who are unable to administer or tolerate subcutaneous injections, we suggest treatment with intranasal triptans. Choices include intranasal sumatriptan 20 mg or intranasal zolmitriptan (5 mg). Intranasal triptans are administered contralateral to the side of headache. Intranasal triptan treatment has a slower onset of action than subcutaneous injection but is more convenient to use [6].

For patients who do not respond to or tolerate initial acute therapy with oxygen and/or triptans, alternative choices include intranasal lidocaine, oral ergotamine, and intravenous dihydroergotamine:

●Indirect evidence suggests that intranasal lidocaine for acute cluster headache is less effective than triptan treatment. However, it is easy to administer and generally lacks systemic side effects. (See 'Lidocaine' below.)

●Oral ergotamine may be effective for acute cluster headache, based upon clinical experience, and has the benefit of a longer half-life than triptans, making it an alternative that can be used as a short-term preventive treatment (evening dosage) for attacks that occur only at night. Intravenous dihydroergotamine may also be effective for acute cluster headache, based upon retrospective data. Note that ergots should not be given to patients who have received a triptan within the preceding 24 hours and are contraindicated in pregnancy. (See 'Ergots' below.)

Headache prevention — Preventive therapy should be started without delay at the onset of a cluster episode (algorithm 2). The goal is to suppress attacks over the expected duration of the cluster period. An effective preventive regimen is of utmost importance because patients typically have one to eight cluster headaches a day, and repeated use of abortive medications may result in toxicity and/or rebound.

●Verapamil – We recommend initial preventive therapy with verapamil for patients with chronic cluster headache (ie, continuous headaches or remission intervals of <3 months) and those with episodic cluster headache with relatively long-lasting active periods (ie, four weeks or longer) [1,6]. (See 'Verapamil plus glucocorticoids for patients with frequent attacks' below.)

●Glucocorticoids – For patients with episodic cluster headache whose active cluster periods are infrequent and last less than four weeks, we suggest initial preventive therapy with glucocorticoids alone. Glucocorticoids can rapidly suppress cluster attacks during the time required for the longer-acting preventive agents to take effect. Because of their potential for systemic side effects, glucocorticoids should only be used for short-term prophylaxis. (See 'Glucocorticoids for patients with infrequent attacks' below.)

Due to the relatively long time needed to titrate verapamil to an effective dose, glucocorticoids (or dihydroergotamine where available) may also be used as adjunctive therapy in the first two weeks of verapamil administration. An alternative is greater occipital nerve block ipsilateral to the headache attacks. (See 'Greater occipital nerve blocks' below.)

●Combination therapy – Although there is no high-quality evidence supporting the utility of combining various prophylactic drug treatments in cluster headache, clinical experience suggests that some patients do better with a combination of preventive medications rather than high-dose monotherapy. In most instances of combined therapy, verapamil is given in moderate doses (240 to 480 mg daily) in conjunction with another prophylactic medication, such as prednisone, topiramate, or lithium. (See 'Alternative preventive therapies' below.)

●Alternative options – Oral ergotamine has a relatively long half-life compared with triptans and can be used as preventive treatment (evening dosage) for attacks that occur only at night. (See 'Ergots' below.)

When chronic cluster headache is unresponsive to medical treatments, various surgical interventions and neurostimulation techniques are potential treatment options, though none are clearly established as effective. In such cases, it is particularly important to exclude potential causes of secondary cluster headache. Neurostimulation techniques, including sphenopalatine ganglion stimulation and vagus nerve stimulation, appear promising but remain investigational. Destructive surgical procedures are unproven and should be viewed with great caution. (See 'Surgery and neurostimulation' below.)

For patients with episodic cluster headache, prophylactic medications should be tapered after the expected duration of the cluster has passed. For patients with chronic cluster headache who have a good response to preventive pharmacotherapy, we suggest a dose reduction trial of preventive medications every three months. The drugs can be restarted at the lowest effective dose if symptoms recur.

ACUTE INTERVENTIONS

Initial acute therapy — We suggest initial treatment with oxygen inhalation for patients with acute cluster headache attack (algorithm 1) [7]. If oxygen is not effective or readily available, triptans (eg, subcutaneous sumatriptan) are an effective alternative initial treatment option and may be more convenient for some patients with no contraindications.

Oxygen — Although controlled trial evidence is limited [8], oxygen therapy is considered to be safe and effective for aborting cluster headache [6,8,9].

A double-blind randomized trial evaluated 76 patients, each treating four cluster headache attacks, and compared pure oxygen therapy (at 12 L/minute for 15 minutes) with air placebo [10]. By intention-to-treat analysis, pain-free status or adequate relief of attacks at 15 minutes was significantly more frequent with oxygen (78 percent of attacks versus 20 percent with placebo). Smaller randomized trials have also found that inhalation of pure oxygen is beneficial for acute cluster headache attacks [11-14].

In some patients, oxygen may be completely effective even when the pain is at maximal intensity, but in other cases the response is incomplete. Repeated or frequent oxygen intake in a short period of time should be avoided because there is evidence that the attack frequency may increase in some patients with overuse of oxygen [5].

Oxygen (100 percent) is administered via a nonrebreathing facial mask with a flow rate of at least 12 L/minute with the patient in a sitting, upright position [5]. The inhalation should continue for 15 minutes to prevent the attack from returning, although the pain may subside as soon as five minutes after starting oxygen. Higher oxygen flow rates, up to 15 L/minute, may sometimes be effective when standard rates are not [15]. From a practical standpoint, it is reasonable to increase the flow rate if a lower rate is ineffective, as higher flow rate does not increase the risk of side effects. The use of a demand valve mask may enhance the efficacy of oxygen [16].

Oxygen is generally safe and without side effects. However, patients with severe chronic obstructive pulmonary disease should not be treated with inhaled oxygen because of the risk for developing severe hypercapnia and CO₂ narcosis.

Unlike normobaric oxygen therapy, hyperbaric oxygen is not widely available, and there are few studies evaluating its utility for abortive treatment of cluster headache [13,17].

Triptans — Randomized, double-blind, placebo-controlled trials have established that triptans, particularly sumatriptan and zolmitriptan, are effective for the acute treatment of cluster headache [18].

●In trials involving a combined total of 183 patients with cluster headache, subcutaneous sumatriptan (6 mg) was beneficial (ie, pain-free within 20 minutes) in about 75 percent of patients, whereas the response rate for placebo was 25 to 35 percent [19,20]. Sumatriptan was safe with no evidence of tachyphylaxis or rebound in most patients, even after frequent use [21].

●Intranasal sumatriptan is effective for acute cluster headache. A randomized trial of 118 patients found a significantly higher response rate at 30 minutes with sumatriptan (20 mg) compared with placebo (57 versus 26 percent) and a higher 30 minute pain-free rate (47 versus 18 percent) [22]. Although the intranasal route is more convenient, the speed of response with intranasal sumatriptan in this study was slower than that seen with subcutaneous sumatriptan in other studies.

●Intranasal zolmitriptan is another beneficial treatment option. In a randomized trial involving 69 patients available for intention-to-treat analysis, intranasal zolmitriptan at a dose of 5 or 10 mg was modestly effective at 30 minutes for acute cluster headache compared with placebo [23]. Similar results were observed in a second trial involving 52 patients [24]. Of note, the intranasal dose of 10 mg of zolmitriptan has not been approved by the US Food and Drug Administration and is not generally available [1].

●Oral zolmitriptan 10 mg was modestly effective for acute episodic (but not chronic) cluster headache within 30 minutes in another placebo-controlled trial [25].

Dosing and adverse effects — Triptans should be administered as soon as the headache begins, since early treatment is likely to be more effective. Subcutaneous sumatriptan is administered by injection in areas with adequate skin and subcutaneous thickness (eg, the abdomen, thighs, and upper arms). The injection site should be rotated to reduce the risk of local reactions.

With subcutaneous sumatriptan, a 3 mg dose is effective for some patients; the maximum single dose is 6 mg. Although the label for subcutaneous sumatriptan indicates that the dose may be repeated at one hour or more if needed, it is best to repeat the dose no sooner than two hours after the initial dose, since the attack frequency of cluster headache typically ranges between two and six per day. The repeat dose is generally the same strength in mg as the initial dose. The recommended maximum cumulative dose of subcutaneous sumatriptan is 12 mg in 24 hours, but some patients with frequent headaches may require more, as discussed below.

Intranasal triptans are administered contralateral to the pain side because patients with cluster headache and other trigeminal autonomic cephalalgias often have rhinorrhea or nasal congestion ipsilateral to the pain. The dose of intranasal triptan may be repeated once if needed at ≥2 hours.

While triptans are generally well tolerated, uncomfortable side effects include nonischemic chest pain and distal paresthesia. It is still recommended that triptans be avoided in patients with ischemic cardiovascular disease or stroke, Prinzmetal angina, or uncontrolled arterial hypertension. It may be prudent to give the first triptan dose under medical supervision for patients with risk factors but no known coronary heart disease. (See "Acute treatment of migraine in adults", section on 'Triptans'.)

Because of concerns about potential adverse effects, including medication overuse headache, some experts (and the drug label) recommend that subcutaneous sumatriptan 6 mg should be limited to no more than two doses in 24 hours and intranasal sumatriptan 20 mg to no more than two doses (40 mg) in 24 hours [26]. However, cluster headache attacks may occur more often than twice a day, underlining the need to start preventative medicine early. Until there has been sufficient time for the patient's preventative medication to take effect, more than two triptan doses per day may be required and should be prescribed [5,27]. Tachyphylaxis with recurrent use of triptans seems to be rare in cluster headache [27,28].

Alternative acute therapies — For patients who do not tolerate or respond to initial therapies, we suggest alternative agents such as lidocaine, ergots, or octreotide. We use patient risk factors and comorbid conditions to help select among these options.

Lidocaine — Small studies suggest that intranasal lidocaine is effective in at least one-third of patients, although the degree of benefit is modest [29-33].

Intranasal lidocaine (1 mL) is administered ipsilateral to the pain in a 4 percent topical solution [5]. The head position should be in extension by 45 degrees and rotated toward the symptomatic side by 30 to 40 degrees. Lidocaine nasal sprays are not commercially available in the United States but may be prepared by a compounding pharmacy.

Ergots — Oral ergotamine has been used as a treatment for cluster headache attacks since the 1940s [34]. Based upon clinical experience, oral ergotamine may be effective if started very early in the attack [6], but modern trials are lacking [35]. Ergotamine is available as a 2 mg sublingual tablet. The initial dose is 2 mg and may be repeated every 30 minutes with a maximum dose of 6 mg daily and 10 mg a week.

In a small clinical trial, intranasal dihydroergotamine (DHE; 1 mg) was no better than placebo for reducing the duration of cluster headache attacks [36]. However, several small retrospective studies reported that intravenous DHE given in the inpatient or outpatient setting may be effective for cluster headache [37-39]. DHE is given as a 1 mg intravenous bolus and may be repeated at one hour, with a maximum dose of 3 mg in 24 hours. The general limit per week is 6 mg, although this dose is sometimes exceeded in inpatients receiving a repetitive intravenous DHE protocol [38,39]. Because nausea is a common side effect, DHE is typically administered with an antiemetic medication (eg, metoclopramide, promethazine, chlorpromazine, or domperidone) [40,41].

Ergots are contraindicated during pregnancy because of the potential to induce hypertonic uterine contractions and vasospasm/vasoconstriction, which could cause adverse fetal effects. Ergotamine and DHE should not be given to patients who have received a triptan within the preceding 24 hours due to risk of arterial vasospasm.

Octreotide — Octreotide may be effective in the treatment of acute cluster headaches. In a randomized controlled trial of 57 patients, a single dose of subcutaneous octreotide (100 mcg) was superior to placebo in cluster headache response at 30 minutes (52 versus 36 percent, respectively) [42]. Minor gastrointestinal upset was the most common side effect of octreotide treatment.

Although there has been no direct comparison of octreotide with triptans, the results of this study suggest that octreotide is inferior to both subcutaneous sumatriptan 6 mg [19] and intranasal sumatriptan 20 mg [22] in terms of response rate and time to initial relief.

PREVENTIVE INTERVENTIONS — Verapamil is the agent of choice for the initial preventive therapy for most patients with cluster headache. Glucocorticoids are often used adjunctively to provide more rapid benefit during the initial titration of verapamil. In addition, glucocorticoids may be used alone as preventive therapy when cluster attacks are infrequent and short (algorithm 2). (See 'Verapamil plus glucocorticoids for patients with frequent attacks' below and 'Glucocorticoids for patients with infrequent attacks' below.)

Other agents that may be effective include galcanezumab, lithium, and topiramate. (See 'Alternative preventive therapies' below.)

Verapamil plus glucocorticoids for patients with frequent attacks — Verapamil is the drug of choice for prophylaxis of episodic and chronic cluster headache [3,43]. Verapamil may be given alone or with a short course of glucocorticoids to help provide rapid suppression cluster headache attacks during the initial titration.

●Verapamil dosing – We suggest starting verapamil at 80 mg three times daily for initial dosing and increasing the total daily dose by 120 mg every 14 days as tolerated. Both the regular- and sustained-release formulations are useful, but no direct comparative studies are available. Patients treated with regular-release verapamil should receive the total dose in three divided doses a day, while those treated with the sustained-release formulation should receive two divided doses a day.

Most patients respond to a total daily dose of 240 to 480 mg. However, clinical experience suggests that some patients require a total daily dose of up to 960 mg to obtain full prophylactic benefit [1,44]. In an early open-label trial, titration up to a total daily verapamil dose of 1200 mg was employed [45]. Thus, an adequate verapamil trial for most patients entails use of a total daily dose of 480 to 960 mg before the medication is regarded as a failure. The benefit of verapamil is usually seen within two to three weeks. When the bout is ended, verapamil must not be ended abruptly but should be gradually reduced over two to four weeks depending on the dose and finally stopped.

●Glucocorticoid coadministration – For most patients with cluster headaches who tolerate glucocorticoids, we typically give a short course of oral prednisone during the initial titration of verapamil. There are no data favoring one regimen of glucocorticoid administration over another. Our preferred regimen is prednisone 100 mg once a day for three days and then tapering by decreasing the dose 10 mg every third day. The dose and tapering regimen may be individualized according to cluster episode patterns, when known. The efficacy for adjunctive glucocorticoids was reported in a short-term treatment trial where patients started verapamil at 40 mg three times daily and increased every three days up to a total daily dose of 360 mg [46]. Those additionally assigned to daily prednisone 100 mg for five days and tapering by 20 mg every three days had fewer cluster attacks in the first week than those assigned to placebo (difference -2.4 attacks; 95% CI -4.8 to -0.03). Adverse effects were mild and similar in both groups.

The use of high-dose verapamil is associated with an increased incidence of electrocardiographic (ECG) abnormalities, including heart block and bradycardia [47,48]. Therefore, an ECG should be obtained after each dose increment above a total daily dose of 480 mg. Some experts recommend getting a pretreatment ECG to screen for baseline cardiac arrhythmia [49].

Other side effects of verapamil include edema, gastrointestinal discomfort, constipation, dull headache, and gingival hyperplasia. However, verapamil is usually well tolerated and can be used safely in conjunction with sumatriptan, ergotamine, glucocorticoids, and other preventive agents.

The efficacy of verapamil for prevention in cluster headache comes from observational experience and some trial data [3,45,50,51]. In one trial of 30 patients, daily verapamil at 360 mg in three divided doses reduced cluster headache attack frequency and analgesic consumption [43]. During the first week of treatment, the median number of daily attacks, the primary outcome measure, was similar for patients treated with verapamil and placebo (1.1 versus 1.7). However, in the second week, the median number of daily attacks was significantly lower for patients treated with verapamil (0.6 versus 1.7). Additionally, only those assigned to verapamil reported a reduction in headache frequency of >50 percent at two weeks (12 of 15 [80 percent] versus 0 of 15).

Glucocorticoids for patients with infrequent attacks — For patients with episodic cluster headache who have active cluster periods that are infrequent and last less than four weeks, we suggest initial preventive therapy with glucocorticoids alone. Glucocorticoids can rapidly suppress cluster attacks during the time required for the longer-acting preventive agents to take effect.

Glucocorticoids have not been evaluated by rigorous randomized controlled trials for the treatment of cluster headache. However, in several open-label studies and case series, approximately 70 to 80 percent of patients with cluster headache responded to glucocorticoid therapy [52,53]. These studies used different regimens, such as prednisone 30 mg daily and higher or dexamethasone 8 mg daily given in two divided doses. In addition, clinical experience suggests that glucocorticoids are beneficial for reducing the frequency of cluster headache [1].

There are no data favoring one regimen of glucocorticoid administration over another [54]. We suggest using oral prednisone 100 mg once a day for at least three days and then tapering by decreasing the dose 10 mg every third day. Intravenous and oral administration of glucocorticoids can be successfully combined [55].

Some patients do not respond to other preventive treatments and are attack-free only with glucocorticoids. Consequently, such patients require continuous glucocorticoid therapy, which is problematic due to the deleterious side effects of prolonged systemic glucocorticoid use. These commonly include weight gain, Cushingoid facies, easy bruising and skin fragility, cataracts, aseptic necrosis of the femoral or humeral heads, hypertension, diabetes, infection, and osteoporosis. (See "Major adverse effects of systemic glucocorticoids".)

Alternative preventive therapies — For patients who do not respond to or tolerate initial therapy with verapamil and/or glucocorticoids, we use an alternative therapy, using patient risk factors and comorbid conditions to help select among available options. Combination therapy by adding an alternative agent may be beneficial for patients who report partial response to initial therapy. Several classes of medications and some interventional options have shown benefit for patients with cluster headache.

Galcanezumab — Galcanezumab, a human monoclonal antibody that binds to the calcitonin gene-related peptide (CGRP) ligand, is modestly effective for the prevention of episodic cluster headache, as demonstrated in a randomized controlled trial that randomly assigned 106 patients with episodic cluster headache in a 1:1 ratio to subcutaneous galcanezumab 300 mg or placebo once a month for two months [56]. At three weeks, the mean number of cluster headaches per week for the galcanezumab and placebo groups decreased by 8.7 and 5.2 (absolute difference 3.5), respectively, from a baseline of approximately 17.5 for both groups. However, effectiveness of galcanezumab assessed at 12 weeks in a subsequent trial was not confirmed for the prevention of chronic cluster headache [57].

Although its role in cluster headache preventive treatment is not yet defined, galcanezumab is most likely to be reserved for patients with prior episodic cluster headache periods lasting longer than one month when first-line preventative medications are ineffective, poorly tolerated, or contraindicated. It may be less useful for patients with refractory chronic cluster headache.

The dose of galcanezumab for episodic cluster headache is 300 mg (given as three consecutive doses of 100 mg each) at the onset of the cluster period and then monthly until the end of the cluster period. The drug is administered subcutaneously in the thigh, back of the upper arm, or buttocks. Injection site reactions were the most common adverse events in clinical trials.

Lithium — As with glucocorticoids, lithium (lithium carbonate) has not been evaluated by rigorous randomized controlled trials for the treatment of cluster headache. Evidence of effectiveness is limited:

●In a double-blind controlled trial of less than 30 patients, substantial improvement in episodic cluster headache at one week occurred more often with slow-release lithium carbonate (800 mg daily) than with placebo (8 of 13 patients [62 percent] versus 6 of 14 [43 percent]), but the difference was statistically nonsignificant [58]. A number of methodologic issues limit the strength of this study, including small patient numbers, very short-term outcome assessment at one week, and early study termination.

●A comparative, double-blind crossover trial found that lithium and verapamil showed similar effectiveness for cluster headache prevention [51]. However, verapamil was associated with more rapid improvement and better tolerability.

●In a review of older open-label studies, the response rate with lithium (600 to 1500 mg daily) was as high as 78 percent overall for cluster headache, and 63 percent for episodic cluster headache [59].

Because lithium has a narrow therapeutic window, it is generally used for patients with chronic cluster headache only when other drugs are ineffective or contraindicated.

The initial dose of lithium for adults is usually 300 mg once daily, with dose increases every four to five days based upon lithium levels; the typical maintenance dose is 900 to 1200 mg/day given in two to four divided doses of the regular formulation; the sustained-release formulation can be given in two divided doses or once daily. The lithium plasma level should be monitored and should not exceed 1.2 mEq/L (1.2 mmol/L); most patients respond to levels from 0.6 to 0.8 mEq/L (0.6 to 0.8 mmol/L) [60].

Major side effects of lithium are hypothyroidism, tremor, and renal dysfunction. In addition, lithium rarely can cause cardiac arrhythmias. Therefore, it is reasonable to obtain an ECG with rhythm strip at baseline for those age >40 years or any age with cardiovascular risk factors, repeating as indicated. Regular monitoring of serum lithium concentration, as well as liver, renal, and thyroid function, is required.

Topiramate — In the author's clinical experience, topiramate is an effective add-on medication for cluster headache prevention, most often in combination with verapamil. This strategy can sometimes be used to avoid the need for high-dose verapamil monotherapy or long-term glucocorticoid treatment. However, data from available open-label studies are limited and conflicting regarding the effectiveness of topiramate for the prophylaxis of cluster headache [61-63].

The recommended starting dose of topiramate is 25 mg/day. It is titrated at weekly intervals in 25 mg increments according to clinical response and tolerance, up to the recommended total daily dose of 100 mg/day given in two divided doses. The main side effects are cognitive disturbances, paresthesia, and weight loss. It is contraindicated in nephrolithiasis. Topiramate is associated with an increased risk of fetal malformations and is contraindicated during pregnancy.

Greater occipital nerve blocks — In some cases, greater occipital nerve anesthetic blockade or local glucocorticoid injection are effective, at least temporarily, for patients with refractory chronic cluster headache [64-68]. In a systemic review of clinical trials and observational studies that included 595 patients with cluster headache treated with greater occipital nerve blocks, the response rates ranged from 47 to 100 percent [69]. Onset of response varied from several hours to days, and duration of effect persisted for days to months in some patients. Common adverse effects included injection site pain, head/neck pain, and lightheadedness. Serious adverse effects were uncommon and included alopecia at the injection site, and avascular hip necrosis. These data are limited by small numbers in each study, variable definitions and duration of response, and different agents and doses used for treatment.

Other medications — The following medications may have some usefulness as alternative options for patients with medically refractory cluster headache:

●Pizotifen – The antiserotonergic drug pizotifen (3 mg daily) was associated with improvement in cluster headache in a single-blind, placebo-controlled, nonrandomized crossover trial of 28 subjects [35,70]. However, a review of seven small studies suggested that pizotifen has only a modest effect [71]. Its use is further limited by side effects such as tiredness and weight gain.

●Valproate – Evidence regarding the utility of valproate (valproic acid) in cluster headache prophylaxis is limited and conflicting. Open-label and retrospective studies suggested benefit [72,73]. In contrast, a two-week double-blind randomized controlled trial involving 96 patients with cluster headache found no statistically significant difference between treatment with valproate (1000 to 2000 mg daily) and placebo [74]. Valproate can be used in a dose between 5 and 20 mg/kg daily.

●Capsaicin – Repeated ipsilateral intranasal application of capsaicin was reported to be effective in approximately two-thirds of patients with cluster headache in two open-label studies [75,76] and one double-blind, placebo-controlled trial [77]. Since capsaicin can induce pain, sneezing, and nasal secretions upon mucosal application, it is uncertain whether most patients were truly blind to treatment assignment.

●Triptans – Limited data from small uncontrolled studies and case reports suggest that two longer-acting triptans (oral naratriptan [78-80] and frovatriptan [81]) are beneficial for cluster headache prevention. While sumatriptan is effective for treating acute cluster headache (see 'Triptans' above), a multicenter placebo-controlled trial of 169 patients found that oral sumatriptan was not effective for preventing cluster headache attacks [82].

●Ergotamine – Some experts advocate the combination drug caffeine (250 mg) and ergotamine (2 mg) given as a rectal suppository to prevent cluster headache attacks during the night [2]. Ergotamine alone is not available as a suppository.

●Melatonin – Oral melatonin (10 mg) was more effective than placebo in a double-blind, randomized controlled study involving 20 patients, but the response rate was relatively low (5 of 10 patients) [83]. In contrast, a crossover study involving nine patients with refractory cluster headache found that adjunctive melatonin did not produce any benefit compared with placebo [84].

●Indomethacin – Although short-term indomethacin treatment is considered ineffective for cluster headache [85], there are reports of occasional patients meeting diagnostic criteria for cluster headache who have responded to indomethacin when used for several weeks at higher doses (≥225 mg daily) [86].

The available data suggest no utility for the following agents:

●Botulinum toxin – There is no good evidence that botulinum toxin injections have any prophylactic benefit in the treatment of cluster headache [87].

●Hyperbaric oxygen – A small placebo-controlled, double-blind trial found that hyperbaric oxygen was ineffective in preventing cluster headache attacks [88].

SURGERY AND NEUROSTIMULATION — There are several promising but unproven methods using neurostimulation to treat medically refractory cluster headache, including sphenopalatine ganglion stimulation, occipital nerve stimulation, noninvasive vagus nerve stimulation, and deep brain stimulation. All are investigational and require further study to confirm long-term benefit and safety.

●Noninvasive vagus nerve stimulation (VNS) may reduce the frequency of cluster headache attacks, but evidence is inconsistent. In an open-label trial of 97 patients with chronic cluster headache, patients assigned to VNS had statistically significant reductions in the number of attacks per week compared with those assigned to medical treatment [89]. However, a double-blind, randomized trial of 92 subjects with episodic or chronic cluster headache found no difference between noninvasive VNS and sham stimulation in the proportion of treated attacks achieving pain-free status within 15 minutes of treatment initiation [90].

●Deep brain stimulation (DBS) of the posterior inferior hypothalamic region is a promising but unproven investigational treatment for intractable cluster headache [91-95]. Only small numbers of patients have been studied. In an analysis of individual patient data of 40 patients from prior reports, DBS was associated with an overall 77 percent mean reduction of headache frequency among the 75 percent of responders [96]. However, given the investigational status of DBS and the potential for rare but life-threatening adverse effects such as intracerebral hemorrhage [97], stringent criteria for patient selection have been proposed [98]. These require strictly diagnosed, unilateral, chronic cluster headache present daily for at least 24 months and refractory to all state-of-the-art drugs alone and in combination.

●Sphenopalatine ganglion stimulation is performed by surgically implanting a neurostimulator that can be activated by the patient using a handheld remote controller [99]. In a trial of 32 patients with cluster headache, sphenopalatine ganglion stimulation was superior to sham stimulation for pain relief at 15 minutes (67 versus 7 percent) and for several secondary endpoints, including reduction in attack frequency [100]. Serious adverse events involved neurostimulator lead misplacement in three patients and neurostimulator lead migration in two. An open-label follow-up found that effectiveness was maintained for up to 24 months in a majority of responders [101]. Other open-label studies also suggest that sphenopalatine ganglion stimulation is beneficial for a subset of patients with cluster headache [102,103]. The medical device company that previously marketed a sphenopalatine ganglion neurostimulator announced insolvency in 2019 and the device is no longer available.

●Occipital nerve stimulation (ONS) is a technique using electronic stimulation of subcutaneous electrodes implanted in the bilateral occipital region via an implanted pulse generator. Small observational studies have reported benefit for some patients with cluster headache [104-106]. Efficacy in observational studies varies from 36 to 90 percent according to specific outcome measures, surgical approaches, and electrode placement [107]. In a study of 131 patients with medically refractory chronic cluster headache, ONS was associated with an improvement in frequency and severity of cluster attacks [108]. At 24 weeks, 45 percent of patients reported at least a 50 percent reduction in attack frequency. However, infection and lead migration may limit the effectiveness of this therapy. Adverse events included localized pain, neck stiffness, lead migration, and electrode damage. Additional data are needed to help guide patient selection, to optimize the stimulation protocol, and to establish long-term safety of device implantation.

Historically, a number of destructive surgical procedures have been used to treat chronic cluster headache [109-111]. Procedures aimed at the sensory trigeminal nerve have targeted the trigeminal ganglion, supraorbital nerve, infraorbital nerve, and trigeminal root. Newer procedures have targeted the occipital nerves. Procedures directed at autonomic pathways have targeted the greater superficial petrosal nerve, nervus intermedius, and sphenopalatine ganglion.

Surgical methods included radiofrequency thermocoagulation, glycerol ganglio-rhizolysis, alcohol injection, cocaine injection, gamma knife radiosurgery, and nerve sectioning. However, these treatments should be considered only with great caution because reliable long-term observational data are scarce and because complications have included trigeminal neuralgia, anesthesia dolorosa, hypesthesia, meningitis, and cerebrospinal fluid leak [112,113].

The available evidence, while conflicting, illustrates that trigeminal denervation may not be effective in preventing the headache attacks or autonomic symptoms of chronic cluster headache.

●In a prospective study of 10 patients utilizing gamma knife treatment, outcomes graded as excellent, good, or failure were observed in 3, 3, and 4 patients, respectively [114].

●In a retrospective case series of 17 patients with intractable chronic cluster headache treated with total or partial section of the trigeminal nerve root, complete or near-complete relief of symptoms at a mean follow-up of nearly seven years was reported in 15 patients (88 percent) [112].

●In a case report, complete surgical section of the trigeminal sensory root had no long-term effect on the frequency or intensity of cluster attacks [115].

MORBIDITY AND PROGNOSIS — Cluster headache is the most prominent of the trigeminal autonomic cephalalgias. It is recognized as so vicious that some patients attempt suicide if the disease is not diagnosed or successfully treated [5]. Cluster headache has been associated with a reduced quality of life and an increased risk of depression [116,117].

Limited data suggest that cluster headache is a lifelong condition, although the proportion of patients experiencing cluster headache attacks gradually declines with increasing age [35,118,119]. Nevertheless, even 15 years after cluster headache onset, attacks still occur in 80 percent of patients [118]. Cluster headache also results in marked functional disability, even when appropriate treatments are used [120,121].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Migraine and other primary headache disorders".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or email these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Beyond the Basics topic (see "Patient education: Headache treatment in adults (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Acute treatment

•Initial therapy – For patients with acute cluster headache, we suggest initial treatment with oxygen (algorithm 1) (Grade 2C). Triptans (such as subcutaneous sumatriptan 6 mg) are a reasonable alternative when oxygen is not effective or readily available and may be more convenient for some patients. However, triptans are contraindicated in some patients (eg, those with ischemic cardiovascular disease or stroke, or uncontrolled arterial hypertension) (table 2). (See 'Acute headache treatment' above and 'Oxygen' above and 'Triptans' above.)

•Alternative therapies – For patients with acute cluster headache who do not respond to or tolerate initial therapy with oxygen and/or triptans, alternatives include intranasal lidocaine, oral ergotamine, and intravenous dihydroergotamine. These agents are associated with potential toxicity and/or have less evidence of efficacy than initial treatment options. (See 'Lidocaine' above and 'Ergots' above.)

Preventive therapy should be started without delay once a cluster episode begins with the goal of suppressing attacks over the expected duration of the cluster period (algorithm 2). (See 'Headache prevention' above.)

●Preventive treatment

•Patients with frequent and/or longer cluster periods – For patients with chronic cluster headache (ie, continuous headaches or remission intervals of <3 months) and those with episodic cluster headache with relatively long-lasting active periods (ie, four weeks or longer), we use combination therapy. Our approach is as follows (see 'Verapamil plus glucocorticoids for patients with frequent attacks' above):

-We suggest preventive therapy with verapamil rather than other agents (Grade 2C). Verapamil is generally well tolerated and effective in this setting. The starting dose is usually 240 mg daily in three divided doses. Most patients respond to a total dose of 240 to 480 mg daily. Titration to a total dose of up to 960 mg daily may be necessary for some patients to obtain full prophylactic benefit.

-We also suggest a short course of glucocorticoids to provide rapid benefit during the initial titration of verapamil (Grade 2C). We use oral prednisone 100 mg once a day for at least three days and then tapering by decreasing the dose 10 mg every third day. Patients who experience good relief with symptomatic treatment and those with relative contraindications to glucocorticoids may choose to forego such treatment.

•Patients with infrequent and shorter cluster periods – For patients with episodic cluster headache who have active cluster periods that are infrequent and last less than four weeks, we suggest initial preventive therapy with glucocorticoids alone (Grade 2C). We prefer oral prednisone 100 mg once a day for at least three days, followed by a taper with a dose reduction of 10 mg every three days. (See 'Glucocorticoids for patients with infrequent attacks' above.)

•Alternative options – We reserve galcanezumab for patients with prior cluster headache periods lasting longer than one month when first-line preventive medications are ineffective, poorly tolerated, or contraindicated. (See 'Galcanezumab' above.)

Alternative medications such as lithium, topiramate, and interventional procedures may be useful for some patients who do not respond to initial preventive therapies. (See 'Lithium' above and 'Topiramate' above and 'Greater occipital nerve blocks' above and 'Other medications' above.)

Neurostimulation and other surgical options used to treat medically refractory cluster headache are investigational and require further study to confirm long-term benefit and safety. Destructive surgical procedures aimed at the sensory trigeminal nerve or autonomic pathways are unproven and should be viewed with great caution. (See 'Surgery and neurostimulation' above.)

●Prognosis – Limited data suggest that cluster headache is a lifelong condition, but attacks gradually decline with increasing age. Cluster headache results in severe functional consequences even when appropriately treated. (See 'Morbidity and prognosis' above.)