Cluster Headache

01Pathophysiology

Cluster headache is classified as a trigeminal autonomic cephalalgia (TAC), a group of primary headache disorders characterized by unilateral head pain accompanied by ipsilateral cranial autonomic symptoms. The pathophysiology centers on the posterior hypothalamus, which acts as the pacemaker driving the cyclical and clock-like nature of attacks. Functional imaging studies consistently demonstrate activation of the ipsilateral posterior hypothalamic gray matter during cluster attacks, which explains the striking circadian and circannual periodicity of this disorder.

The hypothalamic activation triggers the trigeminovascular system. Nociceptive signals travel through the ophthalmic division of the trigeminal nerve (V1), producing the intense periorbital and temporal pain. This trigeminal activation simultaneously engages the trigeminal-autonomic reflex arc, a brainstem pathway connecting the trigeminal nucleus caudalis to the superior salivatory nucleus. Efferent parasympathetic fibers then travel via the greater petrosal nerve and the sphenopalatine ganglion to the lacrimal gland, nasal mucosa, and conjunctival vessels. This reflex arc is responsible for the ipsilateral autonomic features: lacrimation, conjunctival injection, nasal congestion, and rhinorrhea.

The associated miosis and ptosis (a partial Horner syndrome) result from sympathetic dysfunction, likely from compression or inflammation of the pericarotid sympathetic plexus within the cavernous sinus during the intense vasodilation and edema that accompany an attack. Unlike a complete Horner syndrome (which also includes anhidrosis of the entire face), cluster headache produces only a partial Horner syndrome because the sympathetic disruption occurs distal to the bifurcation of sudomotor fibers. This partial Horner syndrome can persist between attacks in long-standing cases.

A critical behavioral feature that directly follows from the intensity of hypothalamic and trigeminovascular activation is psychomotor agitation. Unlike migraine patients, who prefer to lie still in a dark room, cluster headache patients cannot remain still. They pace, rock, or even bang their heads. This restlessness is one of the most reliable clinical discriminators and is believed to result from the extreme severity of pain combined with autonomic arousal.

The reason alcohol reliably triggers attacks only during a cluster period (and not during remission) is thought to relate to hypothalamic sensitization. During an active cluster bout, the hypothalamic pacemaker is in a vulnerable, hyperexcitable state, and vasodilatory substances like alcohol lower the threshold for attack initiation.

02Classification and Clinical Manifestation

Classification

Clinical Manifestation by Phase

ICHD-3 Diagnostic Criteria Summary

03Diagnostic Workup

Cluster headache is fundamentally a clinical diagnosis. There is no laboratory test, imaging finding, or biomarker that confirms it. The diagnosis rests entirely on pattern recognition: the combination of strictly unilateral severe pain in the V1 distribution, ipsilateral cranial autonomic symptoms, attack duration of 15 to 180 minutes, and the behavioral hallmark of agitation.

The role of MRI brain with gadolinium contrast is not to confirm cluster headache but to exclude dangerous mimics, particularly lesions in the pituitary region, cavernous sinus, or posterior fossa that can produce a secondary cluster-like phenotype (so-called "symptomatic cluster headache"). Current guidelines recommend that every patient with a new diagnosis of cluster headache undergo at least one brain MRI. If the MRI is normal, it does not need to be repeated for each cluster period.

Vascular imaging (MRA or CTA of the head and neck) is warranted when the presentation includes features atypical for cluster headache, such as a persistent neurological deficit, neck pain, or a new Horner syndrome outside of an attack, which may suggest carotid artery dissection.

An indomethacin trial is a functional diagnostic test used when there is uncertainty between cluster headache and other trigeminal autonomic cephalalgias. A dramatic and complete response to indomethacin is diagnostic of paroxysmal hemicrania or hemicrania continua. Cluster headache does not respond to indomethacin. This distinction is a high-yield exam point.

If the patient is over 50 years old, ESR and CRP should be obtained to exclude giant cell arteritis, which can present with severe unilateral head pain and may occasionally mimic cluster headache.

04Management and Treatment

Acute treatment focuses on speed. Because cluster attacks are short (15 to 180 minutes) and excruciating, slow-onset medications like oral analgesics are ineffective. The two first-line acute treatments are high-flow 100% oxygen and subcutaneous sumatriptan. Oxygen is administered at 12 to 15 L/min via a non-rebreather mask with a reservoir bag. The patient should be seated and leaning slightly forward. It is safe, has no significant contraindications, and can be repeated without limit. Subcutaneous sumatriptan at 6 mg acts within minutes and is the most rapidly effective pharmacologic agent. These two treatments can be used together. Oral triptans are generally too slow for cluster headache and are not recommended as first-line. Intranasal sumatriptan (20 mg) or intranasal zolmitriptan (5 mg) are alternatives with intermediate onset.

Opioids and barbiturate-containing analgesics have no role in cluster headache management. This is a commonly tested point. NSAIDs and acetaminophen are also ineffective for acute cluster attacks.

Transitional (bridge) prevention is initiated at the start of a cluster period to suppress attacks quickly while a slower-acting maintenance preventive is being titrated. Prednisone is the most commonly used transitional agent, started at 60 to 80 mg/day and tapered over 2 to 3 weeks. Corticosteroids should not be used long-term due to cumulative adverse effects. A greater occipital nerve block with local anesthetic (such as lidocaine or bupivacaine with or without methylprednisolone) on the ipsilateral side can also provide rapid, short-term relief.

Maintenance prevention is the cornerstone of long-term cluster management during a cluster period. Verapamil is the first-line maintenance preventive. It is started at 240 mg/day in two or three divided doses and titrated upward in increments of 80 mg every 1 to 2 weeks. Doses up to 960 mg/day may be necessary for refractory cases. Because verapamil at higher doses carries a risk of PR interval prolongation and heart block, an ECG must be performed at baseline and repeated after each dose increase. This ECG monitoring requirement is a frequently tested detail.

Lithium is an alternative, particularly in chronic cluster headache. It requires monitoring of serum lithium levels (target 0.6 to 1.0 mEq/L), thyroid function, and renal function. Lithium has a narrow therapeutic window, and toxicity can be precipitated by dehydration, NSAIDs, ACE inhibitors, and thiazide diuretics.

A notable contraindication: triptans and ergotamines should not be used within 24 hours of each other due to the risk of coronary vasospasm. Additionally, triptans are contraindicated in patients with uncontrolled hypertension, coronary artery disease, or a history of stroke.

05Differential Diagnosis and Distractors

06Traps and High-Yield Pearls

The single most common trap on this topic is confusing cluster headache with migraine. Both cause severe unilateral headaches, and both can have autonomic features. The examiners rely on two reliable discriminators: (1) the patient's behavior during the attack, and (2) the duration of pain. A cluster headache patient is agitated and restless, unable to stay still, pacing the room. A migraine patient wants to lie in a dark, quiet room without moving. If the vignette describes a patient pacing or rocking during the headache, think cluster headache. If the patient is lying down with the lights off, think migraine. Attack duration is the second anchor: 15 to 180 minutes points to cluster; 4 to 72 hours points to migraine.

The second major trap involves differentiating among the trigeminal autonomic cephalalgias. The examiners distinguish these conditions primarily by attack duration and indomethacin responsiveness. Cluster headache attacks last 15 to 180 minutes and do not respond to indomethacin. Paroxysmal hemicrania attacks last 2 to 30 minutes and respond completely to indomethacin. Hemicrania continua is continuous rather than episodic and also responds completely to indomethacin. If a vignette describes a patient with a cluster-like headache that resolves completely with indomethacin, the answer is paroxysmal hemicrania, not cluster headache.

A third common error involves acute treatment selection. Students may select oral sumatriptan or oral analgesics for an acute cluster attack. This is incorrect. The speed of onset matters: cluster attacks are brief and intense, so treatment must work within minutes. The correct first-line agents are high-flow oxygen (12 to 15 L/min) and subcutaneous sumatriptan (6 mg). Oral formulations are too slow.

Finally, do not forget the ECG monitoring requirement with verapamil. If a question asks about the next step after starting or increasing verapamil, the answer is to obtain an ECG to evaluate for PR prolongation or heart block. This is a test of your knowledge of drug monitoring, not headache pathophysiology, and it appears with regularity.

The core competency being tested across all cluster headache questions is pattern recognition of a distinctive clinical syndrome: a young male smoker with episodic, strictly unilateral periorbital pain lasting under 3 hours, accompanied by ipsilateral tearing and nasal congestion, who is agitated and restless during the attack. Recognizing this pattern instantly is what the examiners expect.