Cluster Headache

Chronic and recurrent headaches are a complex problem. This is an understatement when it comes to Cluster Headache. This is by far the most severe form of headache and many experts believe it to be the most severe condition humans experience. This prompted eminent Professor and World expert on headache management Peter Goadsby to comment:

Cluster headache is probably the worst pain that humans experience. I know that’s quite a strong remark to make, but if you ask a cluster headache patient if they’ve had a worse experience, they’ll universally say they haven’t. Women with cluster headache will tell you that an attack is worse than giving birth. So you can imagine that these people give birth without anaesthetic once or twice a day, for six, eight, or ten weeks at a time, and then have a break. It’s just awful.

We know from the research that there are multiple areas of the brain active during cluster headache attacks. Like the other major headache types (migraine, tension headache and cervicogenic headache), cluster headache is characterised by a hyper-reflexive trigeminal nucleus. This uncouples from the hypothalamic grey matter (housing our internal body clock), which in all likelihood explains why there is such a strong timing component with cluster headache – with headaches occurring at similar times of the day repeatedly.

Imaging studies have shown in migraine that this constant ‘noise’ from the trigeminal nucleus seems to cause disruption and dysregulation of the hypothalamus just prior to an attack, along with other centres in the brainstem that control the ‘volume’ or intensity of pain.

It is likely that our sensitivity has evolved to naturally ‘cycle’ during the day (with our circadian cycle), during the week, during the month (typically hormonal cycles), and at different times of year with the seasons. One example of this cycling is menstrual migraine, where a hormonal change will  allow this underlying noise to become amplified.

In cluster headache changes in circadian rhythm with the time of year (attacks more prevalent in transition seasons – spring and autumn/fall), and changes with timing of sleep cycles (commonly a shift from slow wave sleep to REM sleep is a significant trigger overnight) can be potent triggers – but not everyone who undergoes theses changes gets pain, so what is causing the irritation? What is the source of the pain that this shifting sensitivity thresholds acts upon?

The answer, as for all head-pain must lie within the structure that generates pain signals for the head, face and neck – the trigeminal nucleus. ​It stands to reason, as one of the two major inputs into the trigeminal nucleus, that the upper cervical spine may be at least in part, responsible for the overstimulation, and hence, hyper-reflexivity, and a source of significant irritation during clusters. At the very least, you have nothing to lose (other than your headaches!)  by ruling it out as a source of the problem.

Call today and book your assessment to find out if your neck is overstimulating the trigeminal nucleus, and a potential cause of your symptoms.

Find out how we treat

Cluster Headache

here at the Melbourne Headache Centre

Diagnosis & Symptoms

Cluster headache is classified by the International Headache Society (IHS) under a collective called Trigeminal Autonomic Cephalalgias. This group includes SUNCT/SUNA, paroxysmal hemicrania and hemicrania continua. They are so called and classified due to the presence of distinctive autonomic features.

Cluster Headache is defined in the ICHD-3 as:

A. At least five attacks fulfilling criteria B-D

B. Severe or very severe unilateral orbital, supraorbital and/or temporal pain lasting 15-180 mins (when untreated).

C. Either or both of the following:

  1. at least one of the following symptoms or signs, ipsilateral to the headache:
    • conjunctival injection (bloodshot eyes) and/or lacrimation (flow of tears)
    • nasal congestion and/or rhinorrhoea (a runny nose)
    • eyelid oedema (retention of fluid or swelling)
    • forehead and facial sweating
    • miosis (constriction of the pupil) and/or ptosis (drooping of the upper eyelid over the eye)
  2. a sense of restlessness or agitation

D. Occurring with a frequency between one every other day and eight per day

E. Not better accounted for by another ICHD-3 diagnosis


Attacks occur in series lasting for weeks or months (so-called cluster periods) separated by remission periods usually lasting months or years.

About 10-15% of patients have Chronic cluster headache, without remission, or with remission less than 3 months over the course of 1 year.

Cluster headache sufferer reported associated features:

  • 91% report lacrimation
  • 84% report nasal rhinorrhoea
  • 59% report sweaty forehead
  • 48% were photophobic
  • 42% were phonophobic
  • While nausea (36%) and vomiting (17%) are lower than would be expected if they were the natural product of severe pain.
  • 70-90% report a sense of restlessness and agitation during an attack, pacing, rocking or banging their heads.
  • 14% report an aura preceding attacks by up to 60 minutes
  • Between attacks patients often report cutaneous allodynia in the area of their pain and also the occipital nerve on the same side.


On initial presentation


incorrectly diagnosed
Suicidal ideation affects


of CH sufferers
High personal cost with


losing jobs due to CH

Cluster Headache is (fortunately) considered a rare disorder, which makes studying the epidemiology very difficult.

  • Estimated prevalence ranges from as low as 3 in 100,000 or 0.00003% up to 381 per 100,000 or 0.0039% of the population.
  • Pooled lifetime prevalence of 0.12%
  • Men affected 3 times more than women
  • Peak age of onset typically 20-25 y.o., however considerable delay to diagnosis with peak age of diagnosis at 30-34
  • Significant diagnostic delay with sufferers waiting up to 5 years before accurate diagnosis – only 21% being diagnosed properly on their first presentation.

Common Treatments

Characteristically, Cluster Headache is recalcitrant to treatments.

In acute medication trials the key outcome measure is being pain free at 2 hours. The average duration of an attack is 45-90mins so measuring the effectiveness of acute medication is almost impossible in a classic research sense.

In a similar vein preventive medication typically takes weeks to months to build an effective dose. With a ‘cluster’ duration averaging 6-12 weeks, a medication that takes 12 weeks to work can be difficult to measure with regards to efficacy. This can be easy to measure in Chronic Cluster Headache, but episodic will go into remission typically for periods of 12-24 months. If its 30 months is that natural variation or due to some effect of the preventive medication?

Watson Headache® Approach and Cluster Headache

A feature common to headache disorders is an overactive trigemino-cervical nucleus and cluster headache is no stranger, with the trigemino-cervical reflex having been evaluated. Combined with the efficacy of greater occipital nerve blocks (this is a nerve that comes out of the top of the neck) it makes sense to start with a thorough examination of the upper cervical spine to determine if there is a condition causing sensitivity there. Of course, it is far from being a ‘neck disorder’ as many other pieces of the puzzle must be present, not in the least of which is the hypothalamus controlling the stereotypic timing.

However, the neck to cluster headache, could be considered similar to the wind in a bush fire. It may not start it, but once you have an issue, its presence may cause a much bigger problem.

Oxygen Therapy for Cluster Headache:

High-flow oxygen therapy has been used for Cluster headache since 1952. The reason why oxygen helps acute cluster episodes is unclear, with theories ranging from vasoconstriction to hypoxic impact on trigeminal afferents being discredited.

The effectiveness has been reported varyingly from 53% to 82% with:

Oxygen Therapy Small

  • 13% achieving complete remission
  • 41% report it as ‘very effective’
  • 27% report it being ‘moderately effective’
  • 12% report minimally effective and
  • 7% report no effect

Some research has been conducted into mode of delivery.

Optimal oxygen flow rates seem to be between 6 and 15L/min.

The type of mask has undergone limited testing with O2ptimask and ‘Demand Valve Oxygen’ masks showing reduced need for rescue medication.

With cigarette smoking being a significant risk factor for cluster headache, being a current smoker prohibits prescription of oxygen therapy due to the risk of explosion.

Supplements for Cluster Headache

There is mounting evidence that a supplement based anti-inflammatory regime can be effective in Cluster Headache.

According to the following regime has shown benefit in up to 80% of sufferers: Vitamin D3, Omega 3 fish oil, calcium, magnesium glycinate, vitamin K2, Vitamin A, Vitamin B50, Zinc and Boron. I have not included the dosage but the original article can be found here. This constitutes general advice and is not a prescription. Should you choose to use this protocol, please do so in conjunction with a qualified health professional as some supplements are contraindicated in some health conditions.

Medication for Cluster Headache

For acute episodes:

The following drugs have a reported efficacy:

Triptans  – 54% reporting some benefit with oral sumatriptan.

Lidocaine – effective in aborting attacks in up to 30% of episodic cluster headache sufferers.

Ergotamine + caffeine (caffergot): only available through compounding pharmacies via prescription due to cardiac adverse side effects.


Verapamil, Lithium, Topiramate, Botox

Corticosteroids – either orally or via Greater Occipital Nerve block

CGRP medications:

  • Galcanezumab (Emgality) – phase III CGAM study (300mg vs placebo) showed modest reduction in frequency across weeks 1-12. Placebo reduced attacks by 5.2, compared to emgality 8.7, which seems great until you minus the placebo response from emgality – i.e. placebo was 5.2, emgality is 3.5 (the difference between the groups. This makes a 20 cent placebo (injectable saline) 1.5 times more effective than a $1500 per month drug.
  • Fremanezumab – phase III trial discontinued due to lack of effectiveness in chronic, while episodic results are pending.

Neuromodulation – Electrical Stimulation

Non-invasive vagus nerve stimulator demonstrates a modest benefit, and maybe be useful though devices can be prohibitively expensive. One device with 30 charges cost over US$500, and is not rechargeable (even though the demo model is!). With multiple attacks per day lasting up to 8-12 weeks .a sufferer would need to have several devices sitting there waiting, possibly for years.

Sphenopalatine ganglion stimulation – this is a surgically implanted device targets the part of the brain that is responsible for the runny nose, tearing and conjunctival injection part of CH. The acute responder rate from two studies is 32-35%. At 24 months, 35% of chronic cluster headache patients had a frequency reduction of 50%, however in 12% of sufferers there was a 50% increase.

Hypothalamic deep brain stimulation – showed no benefit over placebo.

Occipital nerve stimulator – lacking placebo controlled trials.

Non-medical treatments

There has been significant interest in psilocybin (magic mushrooms) and cannabis for cluster headache, but research is difficult due to the illegal nature of the medications.


The underlying causes for cluster headache, like other primary headache disorders can be broken down into several categories; Brain regions involved or pathophysiology, triggers and pre-disposing factors.

Pathophysiology of Cluster Headache

The distinctive characteristics of cluster revolve around three main areas of the brain. The trigeminocervical complex mediates the pain expression of cluster headache. Another diagnostic feature of cluster headache is the associated symptoms involving lacrimation, runny nose and redness of the eye, indicating activation of the trigeminal-autonomic reflex. Lastly, another stereotypic feature of cluster headache is the timing features – occurring at consistent times of the night/day which are believed to be related to the hypothalamus – home to the suprachiasmatic nucleus (SCN) or master body clock.

Trigeminocervical complex (TCC) and cluster headache

A key requirement for any headache disorder (migraine, tension-type headache, cervicogenic headache and cluster headache) is that signals relating to head (also neck and face) pain are sent to our sensory cortex from a nucleus deep within the lower part of the brainstem (actually extending down into the neck to C3) that receives inputs from the head and face (trigeminal) and neck (cervical) nerves.

What causes the signalling from the TCC? There are two options.

The first example would be an injury (peripheral input) in the region where the pain is being generated (trigeminal nerve behind the eye). Notwithstanding the fact that for something to cause that amount of pain behind your eye you would be in a life threatening situation (i.e. burst aneurysm, penetrating injury to the eye/temple), there have been cases where attempts to alleviate the pain by severing the nerves that would be sending these signals (trigeminal nerve root) has been attempted – sadly with no effect.

The second option is that the relay centre for this input from the periphery is overstimulated – or sensitised.

Cluster Headache and CGRP

During spontaneous attacks of cluster headache, blood was sampled from the external jugular vein. The results demonstrated an increase in CGRP (and VIP – important for the autonomic symptoms), which was then returned to levels deemed normal by cessation of the attack and concomitant use of oxygen or sumatriptan. The paper published by Peter Goadsby and Lars Edvinsson in 1994 then goes on to discuss at length how CGRP is a marker for the trigeminovascular system and implicated in intracranial and cerebral blood vessel pain.

There are two large problems with this assertion that have not been adequately discussed. Firstly, the external jugular vein does not drain intracranial blood vessels. The internal jugular vein does. The source of raised CGRP in external jugular venous blood is predicted by tracing back the tributaries:

  • Posterior external jugular vein – forms from superficial veins in the posterior neck and posterior scalp.
  • Anterior jugular vein – arises from the chin in the region of the hyoid bone or suprahyoid neck.
  • Suprascapula vein – drains the posterior structures of the scapula including the supraspinatus and infraspinatus muscles.
  • Transverse cervical vein – posterior neck musculature.

Secondly, CGRP is always described as a potent vasodilator, yet cerebral blood flow remains relatively unaltered – how can this be reconciled?

One answer is that that the source of increased CGRP is from the neck.

Cluster Headache and the Upper Cervical Spine

Headaches with features of Cluster Headache have been described following traumatic injury to the cervical spine as far back as 1949 (Hunter and Mayfield), and with cervical meningioma (Kuritzky 1984).

Further, injections of local anaesthetic and corticosteroid into the greater occipital nerve can alter the presentation, and even abort attacks of cluster headache, supporting the possibility, that posterior cervical and occipital structures are implicated in the pathogenesis of cluster headache.

Cluster Headache and Circadian Clock

There are a number of factors that point to circadian dysregulation in episodic cluster headache;

  • Stereotypical timing of episodes
  • Spike in cluster onset after longest, and shortest days of the year, and with changes to/from daylight savings.
  • Melatonin has a lower nocturnal peak and abnormalities in cortisol, testosterone, and orexin all indicate hypothalamic dysfunction.
  • Posterior hypothalamus is activated during spontaneous and triggered cluster attacks

Cluster Headache Triggers

  • The most common trigger is the time of year and time of night with peak attack times are between 12 midnight and 3am
  • 73% have a positive history (current or past) of smoking cigarettes
  • Other triggers:
    • 5 3 -63% report alcohol to trigger episodes during a bout
    • 36% report weather changes
    • Some report odours such as perfume, paint and cigarette smoke




Differential Diagnosis

Cluster Headache sufferers endure a significant delay before diagnosis, with sufferers seeing an average of 4.6 practitioners over 5 years before being diagnosed. This is largely due to lack of awareness of the diagnosis to be able to differentiate it from pathological headaches, but more commonly from other primary headaches, most notably, other trigeminal autonomic cephalalgias (TAC’s) and migraine.

Other TAC’s

Hemicrania Continua

The key distinguishing feature is in the name. This is a unilateral headache, often more diffuse than cluster headache (orbito-temporal and focal) and is unremitting. Response to indomethacin is diagnostic for HC (along with paroxysmal hemicrania) and can distinguish HC from CH.

Paroxysmal Hemicrania

This is possibly the closest to cluster headadache, but often more frequent (more than 5 episodes per day) and often shorter (duration 2-30mins), but clearly there can be overlap. The key feature is responsiveness to indomethacin.


SUNCT – Short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing

SUNA – Short-lasting unilateral neuralgiform headache attacks with cranial autonomic symptoms

The key differentiation is the duration of attacks is from 1 to 600 seconds, occurring as single stabs, a series of stabs, or in a saw tooth pattern.

Other Primary Headaches

Hypnic headache

This is a strictly nocturnal headache recurring each night, with a pain duration of 15 minutes to 4 hours, but the similarity to cluster headache ends here. There are no autonomic symptoms and the pain is moderate with Hypnic headache, like migraine, tend to be more comfortable staying still (rather than the agitation seen cluster headache).


Migraine typically have longer duration headaches (defined as a minimum of 4 hours) and rarely are accompanied by autonomic symptoms. The intensity of pain is less and migraine sufferers typically want to avoid movement and keep still, whereas cluster headache sufferers are typically agitated and restlessness during an attack (pacing, rocking or banging their heads).

Trigeminal Neuralgia

If the pain involves the maxillary branch (V2) of the trigeminal nerve it can present suddenly and sharply under the eye, however associated autonomic signs (lacrimation, runny/stuffy nose/sinus, conjunctival injection) are rare. The episodes for TN last seconds, and rarely up to 1 minute, and is of a typical electric zapping quality. CH last a minimum 15 minutes, and have a burning, tearing, brushing, hot iron driven into eye quality. Cluster headache is typically occurring on a circadian periodicity whereas TN is often triggered with mechanical stimulus in the facial region.

Chronic Rhinosinusitis

Due to the locus of pain and presence of nasal discharge and feeling of blocked stuffy nasal passage many people will be subject to unnecessary interventions such as antibiotics (which can then have a negative impact on gut health leading to more inflammation) and surgery (introducing another ‘trauma’ to the trigeminal system).

Cluster-like Headache as Secondary Symptoms

Arterial Dissection

Pituitary Gland Disease

Initial presentations are often cleared with contrast enhanced MRI and MRA (magnetic resonance angiography)

Learn More

Learn more about our unique treatment, and what to expect during out consutations