Approach

The aim of treatment is to provide prompt relief from acute attacks while using preventative therapy to suppress attacks for the duration of the cluster period, or over longer periods in patients with chronic cluster headache. Treatment also aims to shorten or stop the cluster period.

Treatment of cluster headache can be separated into acute treatments, transitional treatments, and preventative treatments. Lifestyle advice should also be given where appropriate; for example, avoiding alcohol and other triggers during a cluster period.

Therapy for acute attacks

Acute therapy is used to stop an individual attack. There is no evidence to support the use of non-steroidal anti-inflammatory drugs, opioids, or paracetamol for treating cluster headache and these agents should not be prescribed.

Acute treatment is focused on parenteral triptans and inhaled high-flow oxygen. As stated in guidelines from the UK's National Institute for Health and Care Excellence, all patients with cluster headache should be offered high-flow oxygen or subcutaneous or nasal triptans in the absence of contraindications.[24]

Non-invasive vagus nerve stimulation is a newer option for acute therapy for episodic cluster headache attacks. Intranasal lidocaine is a further option.

Triptans

  • Subcutaneous sumatriptan has been shown to be effective and is the preferred triptan for treating acute attacks.[1][26][28]​​[29][30]​​ [ Cochrane Clinical Answers logo ]

  • Intranasal zolmitriptan and sumatriptan are effective and well tolerated, with nearly 50% of patients given either drug reporting freedom from pain.[1][26][31][32] [ Cochrane Clinical Answers logo ]

  • Oral medications are generally less effective than subcutaneous and intranasal medications because symptoms begin with little or no warning and peak rapidly, but there is some evidence for effectiveness of oral zolmitriptan.[1][26][28]

  • Cardiovascular risk factors (e.g., coronary artery disease [CAD], cerebrovascular disease, or uncontrolled hypertension) can preclude the use of triptans.

High-flow oxygen

High-flow oxygen given at a rate of 12-15 L/minute for at least 15 minutes (or until the attack is terminated) using a non-rebreathing face mask has been shown to be effective in the acute treatment of cluster attacks, decreasing both the intensity and duration of attacks.[24][26][30]​​[33]​​​​ It is safe, can be used multiple times a day, and is not contraindicated in patients with hypertension or vascular disease.

Non-invasive vagus nerve stimulation

Non-invasive vagus nerve stimulation (using a handheld, patient-controlled, non-invasive vagus nerve stimulator that is applied to the skin of the neck) has been shown to be effective as an acute treatment for people with episodic cluster headache, but not those with chronic cluster headache.[1][30]​​​​[34]​​​​[35]

Lidocaine

Intranasal lidocaine can bring about rapid relief of cluster attacks in at least one third of patients. Application should be as close as possible to the sphenopalatine fossa.[9][11][26] Unlike triptans, it can be used for patients with cardiovascular disease, cerebrovascular disease, or uncontrolled hypertension.

Transitional therapy

Preventative medicines may take at least 2 weeks at the maximum dose to exert their full effect. Therefore, interventions that act more quickly, but are not appropriate for long-term use, are often started concurrently with standard preventative agents to suppress attacks rapidly. These interventions are referred to as transitional, short-term, or bridge therapies.[1][26]

A short tapering course of an oral corticosteroid such as prednisolone is commonly used.[1][26][30]​ Prednisolone was effective for the short-term prevention of episodic cluster headache in a double-blind randomised controlled trial.[36] However, corticosteroids must not be used on a regular basis (more than 2-3 times a year) due to the risk of long-term adverse effects.

Corticosteroids can also be delivered via greater occipital (or suboccipital) nerve blocks, which have been shown to be effective in reducing attack frequency.[30]​ A variety of corticosteroids have been used in studies, usually combined with a local anaesthetic.[1][26] A nerve block is often performed once at the start of a cluster period, and can be repeated on a regular basis (once every 3-4 months) for patients with chronic cluster headache.

Intravenous dihydroergotamine may also be considered as transitional therapy; however, it is contraindicated in patients with cardiovascular or cerebrovascular risk factors and/or established cardiovascular disease (e.g., coronary artery disease, hypertension, or stroke) and should not be used in these patients.[1][26]

Preventative therapy

The main goal of preventative therapy is rapidly suppressing individual attacks and maintaining remission for the duration of the patient's typical cluster period. For patients with the episodic form, once the cluster period is over (that is, the patient has been headache-free for at least 2 weeks), the preventative therapy is gradually tapered off. For patients with the chronic form, preventative therapy is continued indefinitely; however, if the patient remains headache-free, reducing the dose is periodically attempted.

Verapamil

Verapamil is considered the first-line preventative therapy for both episodic and chronic cluster headache.[1][24][26][30]​​[37]​​

An ECG must be performed before verapamil is prescribed to exclude bradycardia and other conduction abnormalities, specifically PR interval prolongation. ECGs are then repeated before each dose increase to check for signs of developing heart block.[38]

Adverse effects include constipation, dizziness, and lower extremity swelling.

Patients with episodic cluster headache should start at the previous maximum efficacious dose of verapamil when a new cluster period begins, as long as the baseline ECG is normal.

Galcanezumab

Galcanezumab, a humanised monoclonal antibody against calcitonin gene-related peptide, was reported to reduce the weekly frequency of episodic cluster headache attacks measured in the 3-week period after the initial injection, compared with placebo.[1][39]​ Efficacy in preventing chronic cluster headache was not demonstrated in a randomised, placebo-controlled study, but a retrospective real-world study suggested benefit for some patients.[40][41]​ Galcanezumab is approved in the US for the treatment of episodic cluster headache. Galcanezumab is recommended in European guidelines as an off-label treatment option for patients with intractable episodic cluster headache (it is only approved for the treatment of migraine in Europe).[30]

Other pharmacological agents

Other preventative therapies that are used include lithium, melatonin, topiramate, gabapentin, and valproate, although evidence for efficacy is limited.

Lithium is reported as being effective, although evidence is contradictory.[1][9][26][29][30]​​ It has a narrow therapeutic range. Lithium toxicity can result in nausea, vomiting, tremors, confusion, and vision changes. Hypothyroidism and kidney dysfunction are potential long-term adverse effects of lithium.

Melatonin has been reported to be effective, but evidence is largely anecdotal.[1][26][42] Given its safety profile, some have suggested melatonin as first-line therapy in all patients requiring short-term prevention.[11][43] However, better-quality evidence for effectiveness is available for other therapies.

Topiramate appears to be relatively effective in clinical practice.[1][9][26][30]​​​ Potential adverse effects include paraesthesias, weight loss, cognitive slowing, hypohidrosis, nephrolithiasis, and acute angle-closure glaucoma. Topiramate is teratogenic, with an increased risk of cleft lip or palate.[44] In some countries, topiramate is contraindicated in women of childbearing age unless the conditions of a pregnancy prevention programme are fulfilled to ensure that women of childbearing potential: are using highly effective contraception; have a pregnancy test to exclude pregnancy before starting topiramate; and are aware of the risks associated with use of the drug.[45][46]

Evidence for gabapentin is scarce and clinical experience is not always as dramatic as that reported in the literature.[1][26][47]

Valproate semisodium has been deemed 'probably ineffective', but it may be considered in circumstances where all other treatments have failed.[26] Valproic acid and its derivatives are teratogenic:

  • These agents must not be used in female patients of childbearing potential unless other options are unsuitable, there is a pregnancy prevention programme in place, and certain conditions are met.

  • Precautionary measures may also be required in male patients owing to a potential risk that use in the 3 months leading up to conception may increase the likelihood of neurodevelopmental disorders in their children.

  • Regulations and precautionary measures for female and male patients may vary between countries, with some countries taking a more heightened precautionary stance, and you should consult your local guidance for more information.

Non-invasive vagus nerve stimulation

For patients with chronic cluster headache, non-invasive vagus nerve stimulation as an adjunct to standard of care was shown to significantly reduce cluster headache attack frequency compared with standard of care alone.[1][30]​​[48][49]​​

Invasive neuromodulation

Surgery should only be considered for refractory chronic cluster headache once all other options have been tried, as procedures carry risks of serious adverse events. Patients can be considered for invasive neuromodulation with methods such as occipital nerve stimulation (ONS), sphenopalatine ganglion (SPG) stimulation, or deep brain stimulation (DBS) of the posterior hypothalamic region.[30]​​[50] The mechanisms by which neuromodulation works in cluster headache remain unknown, but they are thought to involve modulation of the brain's pain matrix, trigeminal cervical complex, and trigeminal autonomic reflex.

Decompression or destruction of the trigeminal nerve were used previously for refractory cluster headache, but morbidity is significant for a low rate of pain cessation. These surgical techniques should not be considered for cluster headache.[51]

Occipital nerve stimulation

The rationale for ONS lies in the important role the trigeminocervical complex has in cluster headache. ONS was shown in cohort studies to be a promising therapeutic option for intractable chronic cluster headache, with over two-thirds of patients showing a good response.[26][52][53][54][55] There is evidence that ONS is effective even in the longer-term treatment of chronic cluster headache.[56][57] Adverse effects or complications include infection, lead breakage and migration requiring replacement, battery depletion and replacement, skin discomfort, and painful overstimulation.

Invasive sphenopalatine ganglion stimulation

The SPG is a key component of the trigemino-autonomic loop that is responsible for the production of the autonomic features in cluster attacks. An implantable miniaturised device has been developed specifically to treat cluster headache. Evidence shows that the device is useful both in the treatment of acute attacks and potentially as a preventative treatment.[58][59][60] There is no evidence available on its long-term effectiveness.

Deep brain stimulation

The rationale for DBS in chronic cluster headache stems from the imaging findings of activation of the ipsilateral posterior hypothalamic region in cluster attacks.[61] There are numerous published cases, with two-thirds of patients achieving a 50% or more improvement in their headache frequency over a mean follow-up of 2.2 years.[62][63][64][65] One prospective open-label study of 21 patients has shown sustained efficacy over a median of 18 months.[66] The potential adverse effect profile of DBS includes intracerebral bleed, stroke, death, infection, and seizure (although all are very rare), and so it should only be considered for patients in whom all other options, including peripheral neurostimulation with ONS, have failed.

Management of cluster headache in pregnancy

Specialist advice should be sought if acute or preventative treatment for cluster headache is needed during pregnancy. No medication is completely free of risk, and decisions should be made on an individual basis, balancing the risk of the untreated headache disorder as a threat to the health of the mother and unborn child against the risk of the treatment.[67][68]

High-flow oxygen therapy is the preferred option for acute treatment of cluster headache in pregnant women.[67] Triptans are thought likely to be safe in pregnancy, but are contraindicated in patients with moderate or severe hypertension.[68][69][70]

Evidence for the safety of verapamil in pregnancy is limited, but calcium-channel blockers are generally thought to be safe, and nifedipine is used widely as an antihypertensive treatment in pregnancy, with limited adverse effects and established safety.[67][68][70]

Topiramate and lithium are contraindicated for preventative treatment in pregnancy, especially in the first trimester during organ development, because of the risk of major congenital malformations.[70][71] However, it has been suggested that they may be considered in the second and third trimesters for pregnant patients with severe cluster headache when other treatments are ineffective, if the benefit to the mother clearly exceeds the possible risk to the fetus.[72]

Prednisolone and gabapentin may be used with caution (avoided during the first trimester and dose kept as low as possible), but there is some suggestion of an increased risk of fetal abnormalities.[72]

Valproic acid and its derivatives may cause major congenital malformations, including neurodevelopmental disorders and neural tube defects, after in utero exposure. These drugs are contraindicated in pregnancy; however, if it is not possible to stop them, treatment may be continued with appropriate consultant care.

There is no safety information about galcanezumab in pregnancy, and it is not currently recommended for pregnant women; patients should stop taking it 6 months before trying to become pregnant.[68]

There have been no reports of pregnancy-related adverse effects of non-invasive neuromodulation or greater occipital nerve block.[67][68]

Use of this content is subject to our disclaimer