A headache is one of the most common types of pain. A headache is only rarely a symptom of a serious underlying disorder. The pain arises from tension in the meninges (the membranes around the brain), and in the blood vessels and muscles of the scalp. Headache may be felt all over the head or may occur on only one side, or in the forehead or the back of the neck. Sometimes the pain shifts from one area to another. The pain of headache may be superficial, throbbing, or sharp and may be accompanied by nausea, vomiting, and visual or other sensory disturbances.

Types

Many headaches are simply a response to some adverse stimulus, such as hunger. Headaches such as these usually clear up quickly. Tension headaches, due to tightening in the face, neck, and scalp muscles as a result of stress or poor pos-ture, are also common, and may last for days or weeks. Migraine can be a severe, incapacitating headache that is preceded or accompanied by visual and/or stomach disturbances (nauses and/or vomiting). Cluster headaches cause intense pain behind one eye.

Causes

Common causes of headache include hangover and noisy or stuffy environments. Some headaches are caused by the overuse of painkillers. Other possible causes include sinusitis, toothache, cervical osteoarthritis, and head injury. Food additives may also be a cause. Among the rare causes of headache are a brain tumour, hypertension (high blood pressure), temporal arteritis (inflammation of arteries in the face, neck, and scalp), an aneurysm (ballooning of a blood vessel), and increased pressure within the skull.

Treatment

Most headaches can be relieved by painkillers and rest. If a neurological cause is suspected, CT scanning or MRI (magnetic resonance imaging) may be performed.

Headache in detail - technical

Essentials

To manage headache can be a source of extreme frustration or undiluted pleasure—the difference simply reflects to what extent the practitioner is familiar with the subject. A formal nosology for headache disorders is to be found in the second edition of the International classification of headache disorders. Although it seems obvious, the key to successful management is establishment of a clear diagnosis. The general concept is that there are primary and secondary forms of headache, following the generic medical principle that clinical syndromes may be caused by something exogenous or secondary, or may manifest anew as the primary disease process. Such a system is outlined in Table 1. Mild secondary headache, such as that seen in association with upper respiratory tract infections, is common but only rarely worrisome. The clinical dilemma remains that, although life-threatening headache is relatively uncommon in Western society, it occurs and its detection requires suitable vigilance by doctors. Primary headache, in contrast, often confers considerable disability over time and although not life threatening certainly robs patients of quality of life. Some pointers to secondary headache are listed in Bullet list 1.

Primary headache syndromes

The primary headaches are a group of fascinating disorders in which headache and associated features are seen in the absence of any exogenous cause. The common syndromes (see Table 1) are tension-type headache, migraine, and cluster headache. Some other less well-known, indeed rarer, syndromes are mentioned because they are easily treated when diagnosed, and the most burdensome headache problems, the chronic daily headaches, are explicitly covered because concepts have altered considerably in this area in recent years.

Anatomy and physiology

The most common disabling primary headaches, migraine and cluster headache, have been studied extensively in recent times and they are now relatively well understood insofar as neurological disorders that involve the brain are concerned. In experimental animals the detailed anatomy of the connections of the pain-producing intracranial extracerebral vessels and the dura mater has built on the classic human observations of Wolff, Feindel, Penfield, McNaughton, and others. It is these structures, and not the brain itself, that are primarily involved in head pain, although it is not at all clear to what extent there is nociceptive activation as such, or simply the perception of that activation.

The key structures involved are:

• ◆ the large intracranial vessels and dura mater
• ◆ the peripheral terminals of the trigeminal nerve that innervate these structures
• ◆ the central terminals and second-order neurons of the caudal trigeminal nucleus and dorsal horns of C1 and C2: trigeminocervical complex

Bullet list 1 Warning signs in head pain

• • ◆ Sudden onset of pain
  • ◆ Fever
  • ◆ Marked change in pain character or timing of attacks
  • ◆ Neck stiffness
  • ◆ Pain associated with higher centre complaints
  • ◆ Pain associated with neurological disturbance, such as clumsiness or weakness
  • ◆ Pain associated with local tenderness, such as of the temporal artery
• ◆ higher-centre processing in the thalamus, ventroposteromedial and posterior thalamus, and cortex
• ◆ modulatory centres in the diencephalon and brainstem, such as periaqueductal grey matter, locus ceruleus, and parts of the hypothalamus

The innervation of the large intracranial vessels and dura mater by the trigeminal nerve is known as the trigeminovascular system. Cranial parasympathetic autonomic innervation provides the basis for symptoms such as lacrimation and nasal stuffiness, which are prominent in cluster headache and paroxysmal hemicrania, although they may also be seen in migraine. It is clear from human functional imaging studies that vascular changes in migraine and cluster headache are driven by these neural vasodilator systems so that these headaches should be regarded as neurovascular. The concept of a primary vascular headache should be abandoned because it does not explain the pathogenesis of what are complex central nervous system (CNS) disorders, or necessarily predict treatment outcomes. The term ‘vascular’ headache has no place in modern medical practice when referring to primary headache.

Migraine is an episodic syndrome of headache with sensory sensitivity, such as to light, sound, and head movement, probably due to dysfunction of aminergic brainstem/diencephalic sensory control systems. The first of the migraine genes has been identified for familial hemiplegic migraine, and includes mutations in the CACNA1A gene for the Ca_V2.1 (α_1A) subunit of the neuronal P/Q voltage-gated calcium channel, the Na⁺/K⁺ ATP pump α₂-subunit gene ATP1A2, and the voltage-gated sodium channel SCN1A. These findings and the clinical features of migraine suggest that it might be part of the spectrum of diseases known as channelopathies, or now ionopathies—disorders involving dysfunction of ion channel fluxes. Functional neuroimaging has suggested that brainstem regions in migraine, and the posterior hypothalamic grey matter site of the human circadian pacemaker cells of the suprachiasmatic nucleus in cluster headache, are good candidates for specific involvement in primary headache.

Secondary headache

It is imperative to establish in the patient presenting with any form of head pain whether there is an important secondary headache declaring itself. The headaches of subarachnoid haemorrhage, meningitis, giant cell arteritis, and raised intracranial pressure are important examples of medically sinister headaches requiring urgent attention. Perhaps the most crucial clinical feature to elicit is the length of the history. Patients with a short history require prompt attention and may require prompt investigation and management. Patients with a longer history generally require time and patience rather than alacrity. There are some important general features, including associated fever or sudden onset of pain (see Bullet list 1); these demand attention. Patients with a history of recent-onset headache or neurological signs need a positive diagnosis of a benign disorder or require brain imaging with CT or MRI. Patients with a history of recurrent headache over a period of 1 year or more, fulfilling International Headache Society (IHS) criteria for migraine (Bullet list 2), and with a normal physical examination, have positive brain imaging in only about 1/1000 images. In general it should be noted that a brain tumour is a rare cause of headache, and rarely a cause of isolated long-term histories of headache. A notable exception to the general rules about secondary headache is a pituitary tumour, which can trigger underlying primary headache biology, and should always be considered, especially in the differential diagnosis of trigeminal–autonomic cephalalgias (see below).

The management of secondary headache is generally self-evident: treatment of the underlying condition, such as an infection or mass lesion. An exception is the condition of chronic post-traumatic headache in which pain persists for long periods after head injury. This is an interesting generic problem that may beseen after CNS infection, trauma, both blunt and surgical, intracranial bleeds, and other precipitants. Although the syndrome is generally self-limiting up to 3 to 5 years after the event, treatment of the headache may be required if it is disabling (see ‘Chronic daily headache’, below).

Bullet list 2  Simplified diagnostic criteria for migraine

Clinical features

Migraine is an episodic brain disorder that affects about 12 to 15% of the population, and can be highly disabling. It has been estimated to be the most costly neurological disorder in the European Union at more than €27 billion per year and its cost to the economy of the United States of America is a staggering US$19.6 billion per year. Migraine presents with a headache generally accompanied by features, such as sensitivity to light, sound, or movement, and often with nausea, or less often vomiting (see Bullet list 2). None of the features is compulsory, and indeed, given that the migraine aura, visual disturbances with flashing lights or zigzag lines moving across the fields, or other neurological symptoms, are reported in only about 25% of patients, a high index of suspicion is required to diagnose migraine.

In a controlled study of patients presenting to general practitioners with a main complaint of headache over the previous 3 months, migraine was the diagnosis on more than 90% of occasions, so a high index of suspicion is well rewarded. A headache diary can often be helpful in making the diagnosis, although in reality the diary usually helps more in assessing disability or recording how often patients use acute attack treatments. Phenotyping remains an essentially clinical art, mixing experience and an understanding of the problems likely to present—good headache histories are taken, not given. In differentiating the two main primary headache syndromes seen in clinical practice, migraine at its most simple level is headache with associated features, and tension-type headache is headache that is featureless; furthermore, most disabling headache is probably migrainous in biology. By features is meant throbbing pain, or sensitivity to sensory stimuli—visual, auditory, olfactory—or to head movement itself.

Frequent migraine

If headache with associated features describes migraine attacks, then headachy describes the person who has migraines over a lifetime. It is important to realize that the word migraine can describe both the attacks using standard criteria (see Bullet list 2) and the disorder itself, which is more than just the attack. People who have migraines (migraineurs) inherit a tendency to have headache that is amplified at various times by their interaction with their environment, the much-discussed ‘triggers’. The brain of the migraineur seems more sensitive to sensory stimuli and to change; and this tendency is notably amplified in women during their menstrual cycle. People who have a migraine may have headache when they oversleep, when tired, when they skip meals, when they overexert, when stressed, or when they relax from a stressor. They are less tolerant to change, and part of successful management is to advise them to maintain regularity in their lives in the knowledge of this fluctuating biology. It is this biology that marks migraine and in clinical practice must over-ride the phenotype of individual headaches.

It has been said that migraine can never occur daily; this is simply not correct. Chronic migraine very definitely occurs and is probably the largest part of the group of headaches known collectively as chronic daily headache that presents to doctors (see below). After making a diagnosis, the second step in the clinical process is to be sure that the disease burden has been captured, how much headache patients have and, more important, what patients cannot do—what is their degree of disability? One can ask the patient directly to get a flavour for this, keep a diary, or get a quick but accurate estimate using the Migraine Disability Assessment Scale (MIDAS), which is well validated and very easy to use in practice.

Principles of management of migraine

After diagnosis the management of migraine begins with an explanation of some aspects of the disorder to the patient:

• ◆ Migraine is an inherited tendency to have headache; this is caused by the patient’s genes, therefore it cannot be cured.
• ◆ Migraine can be modified and controlled by lifestyle adjustment and the use of medicines.
• ◆ Migraine is not life threatening or associated with serious illness, with the exception of women who smoke and use oestrogenic oral contraceptives, but migraine can make life a misery.
• ◆ Migraine management takes time and cooperation, e.g. when a headache diary has to be collected, or enquiry made about the disability.

Nonpharmacological management of migraine

This approach aims to help migrainous patients identify things making the problem worse and encourage them to modify these. Patients need to know that the brain sensitivity that is migraine varies, so that the effect of triggers will vary. Patient associations are often very helpful in supporting migraineurs to identify triggers. The knowledge that there is variability will remove considerable frustration on the patient’s part, and will ring true to most as they have had the experience. The crucial lifestyle advice is to explain to the patient that migraine is a state of brain sensitivity to change. This implies that these people need to regulate their lives: healthy diet, regular exercise, regular sleep patterns, avoiding excess caffeine and alcohol, and, as far as practicable, modifying or minimizing changes in stress. The balanced life with fewer highs and lows will benefit most people who have migraines.

Preventive treatments of migraine

Patients should be advised that they have an inherited, noncurable, but manageable problem. To start a preventive they need to have sufficient disability to wish to take a medicine to reduce the affects of the disease on their life. The basis of considering preventive treatment from a medical viewpoint is a combination of acute attack frequency and attack tractability that confers an unacceptable degree of disability. Patients with attacks unresponsive to abortive medications are easily considered for prevention, whereas patients with simply treated attacks may be less obvious candidates. Another important consideration is disease progress. If a patient diary shows a clear trend of an increasing frequency of attacks, it is better to initiate with prevention than wait for the problem to worsen.

A simple rule for frequency might be that for one to two headaches a month there is usually no need to start a preventive, for three to four it may be needed but not necessarily, and for five or more per month prevention should definitely be considered. Options available for treatment are covered in detail in Table 2 and vary somewhat by country. One problem with preventives is that they have fallen into use for migraine from other indications and often bring unwanted or intolerable side effects. It is not clear how preventives work, although it seems likely that they modify the brain sensitivity that underlies migraine. Another key clinical point is that generally each drug should be started at a low dose and gradually increased to a reasonable maximum if there is going to be a clinical effect.

Relatively little has been done in terms of systematic study of patients with more intractable forms of migraine. Neuromodulation approaches are promising, including stimulation of the occipital nerve, and functional imaging studies show that central processing of pain signals in migraine in the thalamus may be modified by such therapies. This is an exciting and developing area.

Acute attack therapies of migraine

Acute attack treatments for migraine can be usefully divided into disease-nonspecific treatments (analgesics and nonsteroidal anti-inflammatory drugs (NSAIDs)) and disease-specific treatments (ergot-related compounds and triptans). It is important to be aware that most acute attack medications seem to have a propensity to aggravate headache frequency and can induce a state of refractory daily or near-daily headache—medication overuse headache. As evidence is gathered, this seems to occur in patients with migraine: either a previous clear history or a family or personal history of headachiness. Codeine-containing compound analgesics are particularly troublesome when available in over-the-counter (OTC) preparations. Patients with migraine should be advised to avoid taking acute attack medicines on more than 2 days a week. A proportion of patients who stop taking regular analgesics will have substantial improvement in their headache with a reduction in frequency; however, for some it will not make any difference. It is crucial to emphasize to the patient that standard preventive medications often simply do not work in the presence of regular analgesic use.

Treatment strategies

Given the array of options to control an acute attack of migraine, how does one start? The simplest approach to treatment has been described as ‘stepped care’. In this model all patients are treated, assuming no contraindications, with the simplest treatment, such as aspirin 900 mg or paracetamol 1000 mg with an antiemetic. Aspirin is an effective strategy, has been proven so in double-blind controlled clinical trials, and is best used in its most soluble formulations. The alternative is a strategy known as ‘stratified care’, by which the physician determines, or stratifies, treatment at the start, based on the likelihood of response to levels of care. An intermediate option may be described as stratified care by attack. This is what many headache authorities suggest, and what patients often do when they have the option: they use simpler options for their less severe attacks, relying on more potent options when their attacks or circumstances demand them (Table 3).

Nonspecific acute migraine attack treatments

As simple drugs, such as aspirin and paracetamol, are cheap and can be effective, dosages should be adequate and the addition of domperidone (10 mg orally) or metoclopramide (10 mg orally) can be very helpful. NSAIDs can very useful when tolerated. Their success is often limited by inappropriate dosing, and adequate doses of naproxen 500 to 1000 mg orally or rectally, with an antiemetic, ibuprofen 400 to 800 mg orally, or tolfenamic acid 200 mg orally can be extremely effective.

When simple analgesic measures fail or more aggressive treatment is required, the specific antimigraine treatments are required (Table 4). Although ergotamine remains a useful treatment, it can no longer be considered the treatment of choice in acute migraine. There are particular situations in which ergotamine is very helpful, but its use must be carefully controlled as ergotamine overuse produces dreadful headache in addition to a host of vascular problems. The triptans, serotonin 5HT_1B/1D-receptor agonists, have revolutionized the life of many patients with migraine and are clearly the most powerful option available to stop a migraine attack. They can be rationally applied by considering their pharmacological, physicochemical, and pharmacokinetic features, as well as the formulations that are available. Recent data suggest that combining a triptan with an NSAID can improve efficacy and reduce headache recurrence.

Tension-type headache

Clinical features

As its name suggests, tension-type headache (TTH) is the least understood headache form. TTH is diagnosed often and, although the phenotype is common, much of the disabling headache that goes under the name TTH is likely to be migrainous in terms of its biology. TTH has two forms—episodic TTH, where attacks occur on fewer than 15 days a month, and chronic TTH, where attacks, on average over time, are seen on 15 days or more a month. The latter is part of the broader clinical syndrome of chronic daily headache, but these terms are not equal.

TTH has been defined by the IHS for both its episodic and chronic forms, although the admixture of symptoms allowed has consistency problems. A useful clinical approach is to diagnose TTH when the headache is completely featureless: no nausea, no vomiting, no photophobia, no phonophobia, no osmophobia, no throbbing, and no aggravation with movement. Such an approach neatly divides migraine, which has one or more of these features and is the main differential diagnosis from TTH.

Pathophysiology

The pathophysiology of TTH is very poorly understood. This results from the fact that the name implies to most that it is a product of nervous tension, for which there is no clear evidence, and the definitions employed have undoubtedly admitted patients with migraine to the studies. Moreover, the concept that TTH in some way involves muscle contraction is incorrect because the evidence is that muscle contraction is no more likely than it is in migraine. It seems likely that TTH is due to a primary disorder of CNS pain modulation alone in contrast with migraine, which is a more generalized disturbance of sensory modulation.

Management

Adopting the clinical approach to TTH outlined above results in diagnosing a headache form that is usually less disabling, more often described by patients as irritating. Its episodic form is generally amenable to simple analgesics, paracetamol, aspirin, or NSAIDs, which can be purchased over the counter. There are clear clinical studies to demonstrate that triptans in TTH alone are not helpful, although, germane to the above discussion, triptans are effective in TTH where the patient also has migraine. For chronic TTH amitriptyline is the only treatment with a clear evidence base; the other tricylic antidepressants, selective serotonin reuptake inhibitors, or benzodiazepines have not been shown in controlled trials to be effective. Similarly, there is no controlled evidence for the use of electromyography (EMG) biofeedback, relaxation therapy, or acupuncture. Botulinum toxin has been shown reasonably clearly to be ineffective. Stress management has been shown to be an effective approach in a controlled trial.

Trigeminal-autonomic cephalgias

Cluster headache

Cluster headache is a rare form of primary headache with a population frequency of 0.1%. It is covered in specialist books. As a clinical anchor, it is about as common as multiple sclerosis in the United Kingdom, and must be regarded as a disorder best managed by neurologists or headache specialists. It is perhaps the most painful condition of humans; in the cohort of more than 1000 patients seen at the National Hospital for Neurology and Neurosurgery in London not a single one has had a more painful experience, including childbirth, multiple fractures of the limbs, and renal stones. It is one of a group of conditions known now as trigeminal–autonomic cephalalgias (TACs), and thus needs to be differentiated from other TACs and the short-lasting headaches without cranial autonomic symptoms, such as lacrimation or conjunctival injection (Table 5).

Bullet list 3 Diagnostic criteria for cluster headache

The core feature of cluster headache is periodicity, be it circadian or in terms of active and inactive bouts over weeks and months (Bullet list 3). The typical cluster headache patient is male, with a 3:1 predominance, who has bouts of one to two attacks of relatively short duration unilateral pain every day for 8 to 10 weeks a year. They are generally perfectly well in between. Patients with cluster headache tend to move about during attacks, pacing, rocking, or even rubbing their head for relief. The pain is usually retro-orbital, boring, and very severe. It is associated with ipsilateral symptoms of cranial (parasympathetic) autonomic activation: a red or watering eye, running or blocked nose, or cranial sympathetic dysfunction—eyelid droop. Cluster headache is likely to be a disorder involving neurons in or around the central pacemaker regions of the posterior hypothalamic grey matter. Although cluster headache patients may also experience nausea, photophobia, and phonophobia, the last two, particularly photophobia, tend to be ipsilateral to the pain only in TACs.

The TACs—cluster headache, paroxysmal hemicrania, and SUNCT syndrome—present a distinct group to be differentiated from short-lasting headaches that do not have prominent cranial autonomic syndromes, notably trigeminal neuralgia, idiopathic (primary) stabbing headache, and hypnic headache. By determining the cycling pattern, length of attack, frequency of attack, and timing of the attacks, most patients can be usefully classified. The importance of clinical classification of this group is threefold: first, the clinical phenotype determines the likely secondary causes that must be considered and appropriate investigations ordered; second, the appropriate classification gives clarity to the patient with a clear diagnosis and allows the physician to draw on available literature to comment on natural history; and third, the correct diagnosis determines therapy that can be very different in these conditions, being very good if the diagnosis is correct but largely ineffective if it is not (Table 6).

Managing cluster headache

Cluster headache is managed using acute attack treatments and preventive agents. Acute attack treatments are usually required by all cluster headache patients at some time, whereas preventives can seem almost life-saving for the patients with chronic cluster headache, and are often needed to shorten the active periods in patients with the episodic form of the disorder.

Preventive treatments

The options for preventive treatment in cluster headache depend on the bout length (Table 7). Patients with short bouts require medicines that act quickly but will not necessarily be taken for long periods, whereas those with long bouts or indeed those with chronic cluster headache require safe, effective medicines that can be taken for long periods. Verapamil is now widely considered as the first-line preventive treatment when the bout is prolonged, or in chronic cluster headache. By contrast, limited courses of oral corticosteroids or methysergide can be very useful strategies when the bout is relatively short.

Verapamil has been suggested as a useful option for the last decade and compares favourably with lithium. What has clearly emerged from clinical practice is the need to use higher doses than had initially been considered and certainly higher than those used routinely in cardiological indications. Although most patients will start on doses as low as 40 to 80 mg twice daily, doses up to 960 mg daily are often required. Side-effects, such as gingival hyperplasia, constipation, and leg swelling, are recognized, as are cardiac dysrhythmias. Verapamil can cause heart block by slowing conduction in the atrioventricular (AV) node, monitored clinically by the PR interval on the ECG. Given that the effects on the AV node take up to 10 days to manifest, 2-week intervals are recommended between dose changes on the first exposure, with ECGs before the next escalation, and routine 6-monthly ECGs after the dose has been established.

Acute attack treatment

Cluster headache attacks often peak rapidly and thus require a treatment with quick onset. Many patients with acute cluster headache respond very well to treatment with oxygen inhalation. This should be given as 100% oxygen at 10 to 12 litres/min for 15 to 20 min. It is important to have a high flow and high oxygen content. Injectable sumatriptan 6 mg is effective, rapid in onset, and has no evidence of tachyphylaxis. Sumatriptan 20 mg and zolmitriptan 5 mg nasal sprays are effective in acute cluster headache in controlled trials, and offer a useful option. Sumatriptan is not effective when given pre-emptively as 100 mg orally three times daily, and there is no evidence that it is useful when used orally in the acute treatment of cluster headache; indeed it can be associated with medication-overuse headache problems.

Surgical treatment

The surgical treatment of cluster headache has been completely revolutionized with the introduction of neurostimulation therapies. Surgical treatment of cluster headache is reserved for the most refractory patients, typically with chronic cluster headache. Destructive procedures, such as pterygopalatinectomy or radiofrequency lesions of the trigeminal ganglion, have been used. The former, with no clear effects, and the latter being helpful, but often at significant cost, including ocular complications or anaesthesia dolorosa. Trigeminal rhizotomy has also been employed, with all the complications of radiofrequency lesions and the occasional death. Set against this the functional imaging work describing activations in the posterior hypothalamic region directly lead to deep brain stimulation approaches in the same region which seem to be highly effective. Occipital nerve stimulation is a further very promising and largely noninvasive approach to the management of intractable chronic cluster headache, which may become the treatment of choice in this setting over the next 5 to 10 years.

Sjaastad and Dale (1976) first reported eight cases of a frequent unilateral severe but short-lasting headache without remission coining the term chronic paroxysmal hemicrania (CPH). The mean daily frequency of attacks varied from 7 to 22 with the pain persisting from 5 to 45 min on each occasion. The site and associated autonomic phenomena were similar to cluster headache, but the attacks of CPH were suppressed completely by indometacin.

The essential features of paroxysmal hemicrania that we have seen from a substantial cohort of patients are:

• ◆ unilateral, very severe pain
• ◆ short-lasting attacks, typically 20 min in length
• ◆ very frequent attacks (usually >5/day with a mean of 10)
• ◆ marked autonomic features ipsilateral to the pain
• ◆ robust, quick (<72 h), excellent response to indometacin

The pathophysiology of paroxysmal hemicrania (PH) is marked by activations on PET (positron emission tomography) in the contralateral posterior hypothalamus and contralateral ventral midbrain. The posterior hypothalamic activity is shared with cluster headache, SUNCT, and hemicrania continua, whereas the ventral midbrain activity is only seen in hemicrania continua, which remarkably is also an indometacin-sensitive primary headache.

The therapy of PH may be complicated by gastrointestinal side effects seen with indometacin, in which topiramate may be helpful. Secondary PH is more likely if the patient requires high doses (>200 mg/day) of indometacin and raised cerebrospinal fluid pressure should be suspected in apparent bilateral PH. It is worth noting that indometacin reduces cerebrospinal fluid pressure by an unknown mechanism. It is appropriate to image patients, with MRI if practical, when a diagnosis of PH is being considered.

Sjaastad and colleagues (1989) reported three male patients whose brief attacks of pain in and around one eye were associated with sudden conjunctival injection and other autonomic features of cluster headache. The attacks lasted only 15 to 60 s and recurred 5 to 30 times per hour, and could be precipitated by chewing or eating certain foods, such as citrus fruits. They were not abolished by indometacin. Brain imaging has suggested that they share with cluster headache and paroxysmal hemicrania the feature on activation studies of involvement of the posterior hypothalamic region. Of the patients recognized with this problem, males dominate slightly and the paroxysms of pain may last between 5 and 300 s, although longer, duller interictal pains are recognized, as are longer attacks with a sawtooth pattern. The conjunctival injection seen with SUNCT is often the most prominent autonomic feature and tearing may be very obvious.

If one of either conjunctival injection or tearing is absent, or neither is present but another cranial autonomic symptom is seen, the term SUNA is used. The two key clinical features of SUNCT/SUNA are that the attacks can be triggered with no refractory period to triggering. The latter serves as a very useful distinction between SUNCT/SUNA and trigeminal neuralgia. SUNCT/SUNA can be treated very frequently with lamotrigine, and if that is unhelpful topiramate or gabapentin. Carbamazepine often has a useful but incomplete effect. Given what has been reported, cranial MRI with pituitary and posterior fossa views is highly recommended when SUNCT/SUNA is considered as a diagnosis.

Other primary headaches

Primary stabbing headache

Short-lived jabs of pain, defined by the Headache Classification Committee of the IHS as primary stabbing headache, are well documented in association with most types of primary headache.

The following are the essential clinical features:

• ◆ Pain confined to the head, although rarely is it facial
• ◆ Stabbing pain lasting from 1 to many seconds and occurring as a single stab or a series of stabs
• ◆ Recurring at irregular intervals (hours to days).

These pains have been called ice-pick pains or jabs and jolts. They generally respond to indometacin 25 to 50 mg two to three times daily. The symptoms tend to wax and wane and after a period of control on indometacin it is appropriate to withdraw treatment and observe the outcome. Most patients will not want treatment when the nature of the problem is explained and they are reassured that the attacks are not sinister in any way.

Primary cough headache

Sharp pain in the head on coughing, sneezing, straining, laughing, or stooping has long been regarded as a symptom of organic intracranial disease, commonly associated with obstruction of the cerebrospinal fluid pathways. The presence of an Arnold–Chiari malformation or any lesion causing obstruction of cerebrospinal fluid pathways or displacing cerebral structures must be excluded before cough headache is assumed to be benign. Cerebral aneurysm, carotid stenosis, and vertebrobasilar disease may also present with cough or exertional headache as the initial symptom. The term ‘benign Valsalva’s manoeuvre-related headache’ covers the headaches provoked by coughing, straining, or stooping but ‘cough headache’ is more succinct and so widely used that it is unlikely to be displaced.

The following are the essential clinical features of primary cough headache:

• ◆ Bilateral headache of sudden onset, lasting minutes, precipitated by coughing
• ◆ May be prevented by avoiding coughing
• ◆ Diagnosed only after structural lesions, such as posterior fossa tumour, have been excluded by neuroimaging

Indometacin is the medical treatment of choice in cough headache. Raskin followed up an observation of Sir Charles Symonds reporting that some patients with cough headache are relieved by lumbar puncture. This is a simple option when compared with prolonged use of indometacin. The mechanism of this response remains unclear.

Primary exertional headache

The relationship of this form of headache to cough headache is unclear and certainly much is shared. Indeed the relationship to migraine also requires delineation.

The following are the clinical features:

• ◆ Pain specifically brought on by physical exercise
• ◆ Bilateral and throbbing in nature at onset and may develop migrainous features in those patients susceptible to migraine
• ◆ Lasts from 5 min to 24 h
• ◆ Prevented by avoiding excessive exertion, particularly in hot weather or at high altitude.

The acute onset of headache with straining and breath-holding, as in weightlifter’s headache, may be explained by acute venous distension. The development of headache after sustained exertion, particularly on a hot day, is more difficult to understand. Anginal pain may be referred to the head, probably by central connections of vagal afferents, and may present as exertional headache, so-called cardiac cephalgia. The link to exercise is the important clinical clue. Phaeochromocytoma may occasionally be responsible for exertional headache. Intracranial lesions or stenosis of the carotid arteries may have to be excluded as discussed for benign cough headache. Headache may be precipitated by any form of exercise and often has the pulsatile quality of migraine. The most obvious form of treatment is to take exercise gradually and progressively whenever possible. Indometacin at daily doses varying from 25 to 150 mg is generally very effective in benign exertional headache. Indometacin 50 mg, ergotamine tartrate 1 to 2 mg orally, dihydroergotamine by nasal spray, or methysergide 1 to 2 mg orally given 30 to 45 min before exercise are useful prophylactic measures.

Primary sex headache

Sex headache may be precipitated by masturbation or coitus and usually starts as a dull bilateral ache while sexual excitement increases, suddenly becoming intense at orgasm. The term ‘orgasmic cephalgia’ is not accurate because not all sex headache require orgasm. Two types of primary sex headache are recognized: a dull ache in the head and neck that intensifies as sexual excitement increases, and a sudden severe (‘explosive’) headache occurring at orgasm. Low cerebrospinal fluid volume headache may also be precipitated by a sexual activity and is considered as a form of new daily persistent headache (see below).

The following are the essential clinical features of a sex headache:

• ◆ Precipitation by sexual excitement
• ◆ Bilateral at onset
• ◆ Prevented or eased by ceasing sexual activity before orgasm

Headaches developing at the time of orgasm are not always benign, and consideration of a diagnosis of subarachnoid headache is essential. Sex headache affects men more often than women and may occur at any time during the years of sexual activity. It may develop on several occasions in succession, and then not trouble the patient again, despite no obvious change in sexual technique. In patients who stop sexual activity when the headache is first noticed it may subside within a period of 5 min to 2 h, and it is recognized that more frequent orgasm can aggravate established sex headache. About one-third of the patients with sex headache have a history of exertional headaches, but there is no excess of cough headache in patients with sex headache. In about 50% of patients, sex headache will settle in 6 months. Migraine is reported in about 25% of patients with sex headache.

Primary sex headaches are usually irregular and infrequent in occurrence, so management can often be limited to reassurance and advice about ceasing sexual activity if a milder, warning headache develops. When the condition recurs regularly or frequently, it can be prevented by the administration of propranolol; the dosage required varies from 40 mg to 200 mg daily. An alternative is the calcium channel blocking agent diltiazem 60 mg three times daily, which this author finds particularly useful in such patients. Ergotamine (1–2 mg) or indometacin (25–50 mg) taken about 30 to 45 min before sexual activity can also be helpful.

Hypnic headache

This syndrome was first described by Raskin in patients aged from 67 to 84 who had headache of a moderately severe nature which typically came on a few hours after going to sleep. These headaches last from 15 min to 30 min, are typically generalized, although may be unilateral, and can be throbbing. Patients may report falling back to sleep only to be awoken by a further attack a few hours later, with up to three repetitions of this pattern over the night. In Dodick’s series of 19 patients, 16 (84%) were female and the mean age at onset was 61 ± 9 years. Headaches were bilateral in two-thirds and unilateral in one-third, and in 80% of cases mild or moderate. Three patients reported similar headaches when falling asleep during the day. None had photophobia or phonophobia and nausea is unusual.

Patients with this form of headache generally respond to a bedtime dose of lithium carbonate (200 600 mg) and in those who do not tolerate this verapamil or methysergide at bedtime may be alternative strategies. Some investigators reported that one to two cups of coffee or caffeine 60 mg orally at bedtime was helpful. This is a simple approach that is effective in about one-third of patients. An important secondary cause of hypnic headache is hypertension, which should be carefully pursued and appropriately investigated as treatment of the blood pressure will arrest the headache problem.

Primary thunderclap headache

Sudden-onset severe headache may occur in the absence of sexual activity; the differential diagnosis includes the sentinel bleed of an intracranial aneurysm, cervicocephalic arterial dissection, and cerebral venous thrombosis. Headaches of explosive onset may also be caused by the ingestion of sympathomimetic drugs or tyramine-containing foods in a patient who is taking monoamine oxidase inhibitors, and can also be a symptom of a phaeochromocytoma. Whether thunderclap headache can be the presentation of an unruptured cerebral aneurysm is unclear. Day and Raskin (1986) reported on a woman with three episodes of sudden-onset, very severe headache who was found to have an unruptured aneurysm of the internal carotid artery, with adjacent areas of segmental vasospasm. In the absence of a CT scan or cerebrospinal fluid evidence of subarachnoid haemorrhage, studies indicate that such patients do very well, and there indeed seems to be a form of benign or primary thunderclap headache.

Wijdicks et al. (1988) followed up 71 patients whose CT scans and cerebrospinal fluid findings were negative for an average of 3.3 years: 12 patients had further such headaches and 31 (44%) later had regular episodes of migraine or TTH. Factors identified as precipitating the headache were sexual intercourse in 3 cases, coughing in 4, and exertion in 12, while the remainder had no obvious cause. A history of hypertension was found in 11 and of previous headache in 22. Markus (1991) compared the presentation of 37 patients with subarachnoid haemorrhage and 189 with a similar thunderclap headache and normal cerebrospinal fluid examination, and could not discern any characteristic to distinguish the two conditions.

Investigation of any sudden-onset severe headache, be it in the context of sexual excitement or isolated thunderclap headache, should be driven by the clinical context. The first presentation should be vigorously investigated with CT and cerebrospinal fluid examination, and where possible MRI/MR venography/MR angiography. Formal cerebral angiography should be reserved for when no primary diagnosis is forthcoming, and the clinical situation is particularly suggestive of intracranial aneurysm. Bearing in mind the entity of diffuse multifocal reversible cerebral vasospasm, which may be seen in apparent primary thunderclap headache without there being an intracranial aneurysm, caution in interpretation of the findings is crucial.

Hemicrania continua

Two patients were initially reported with this syndrome, a woman aged 63 years and a man of 53, who developed unilateral headache without obvious cause. Both patients were relieved completely by indometacin whereas other NSAIDs were of little or no benefit. Newman and colleagues (1992) reviewed the 24 previously reported cases and added 10 of their own, including some with pronounced autonomic features resembling cluster headache. They divided their case histories into remitting and unremitting forms. Of the 34 patients reviewed, 22 were women and 12 men with the age of onset ranging from 11 to 58 years. The symptoms were controlled by indometacin 75 to 150 mg daily. The following are the essential features of hemicrania continua:

• ◆ Unilateral pain
• ◆ Pain is continuous but with exacerbations that may be severe
• ◆ Complete resolution of pain with indometacin
• ◆ Exacerbations may be associated with autonomic features

Apart from analgesic overuse as an aggravating factor, and a report in an HIV-infected patient, the status of secondary hemicrania continua is unclear. Antonaci and colleagues (1998) proposed the ‘Indotest’ by which the intramuscular injection of indometacin 50 mg could be used as a diagnostic tool. In hemicrania continua, pain was relieved in 73 ± 66 min and the pain-free period was 13 ± 8 h. A placebo-controlled modification of this test is preferred where possible to the open-label version. Using the latter method in conjunction with PET, it has been shown that there is activation of the contralateral posterior hypothalamus and ipsilateral dorsal rostral pons in association with the headache of hemicrania continua, as well as activation of the ipsilateral ventrolateral midbrain. The alternative is a trial of oral indometacin, initially 25 mg three times daily, then 50 mg three times daily, and 75 mg three times daily. One should allow up to 2 weeks for any dose to have a useful effect. Acute treatment with sumatriptan has been employed and reported to be of no benefit. Cyclooxygenase II (COX-II) antagonists seem effective, although undesirable now, and topiramate is helpful in some patients, as is greater occipital nerve injection.

Chronic daily headache

Each of the above primary headache forms can occur very frequently. When a patient experiences headache on 15 days or more a month one can apply the broad diagnosis of chronic daily headache (CDH). CDH is not one thing but a collection of very different problems with different management strategies. Crucially not all daily headache is simply TTH (Table 8). This is the most common clinical misconception in headache, confusing the clinical phenotype with the headache biotype. Population-based estimates of daily headache are remarkable, demonstrating that about 5% of Western populations have daily or almost daily headache. Daily headache may again be primary or secondary, and it seems clinically useful to consider the possibilities in this way when making management decisions (Table 8). It should be said that population-based studies bear out clinical practice in that a large group of refractory daily headache patients overuse various OTC preparations.

Chronic migraine

Although it is widely accepted that some of the primary headaches, TTH, cluster headache, and PH, have chronic varieties, this question seems to have become unnecessarily troublesome for migraine. Few headache authorities would argue that migraine can never ever be chronic in terms of frequency, but the issue of whether patients with frequent headache, some of which fulfils standard criteria for migraine and some for TTH, have a single migrainous biology is a very vexed one. Given that TTH describes a phenomenology that is indistinct at best, it seems unlikely that all its phenotypes will have a single biological generator.

The concept behind chronic migraine is that some patients who inherit a migrainous biology end up with CDH. The typical patient will have daily headache of a dull, nonspecific type, punctuated by more severe attacks that would often, in isolation, fulfil standard criteria for migraine. In headache speciality clinics this group is dominant, with about 90% of patients in referral headache clinics having chronic migraine, usually with medication overuse. It could be suggested that they have a biologically more difficult problem and this is the basis for their over-representation in referral centres.

If one applies the concepts outlined for TTH (see above) then the diagnosis of chronic TTH (CTTH) is made when the patient has 15 days or more a month of entirely featureless generalized dull or pressure-like pain. When any of the attacks on some days have migrainous features—nausea, photophobia, phonophobia, throbbing, or aggravation with movement—then chronic migraine is more likely to be the diagnosis. Clearly both chronic migraine and CTTH exist. Moreover, some patients must simply have coexisting CTTH and episodic migraine; however, it is simply impossible on clinical or other grounds to determine whom they are. The approach outlined probably overdiagnoses chronic migraine, taking that to be a biological entity, and underdiagnoses the coexistence of CTTH and episodic migraine. The converse would be true—if one diagnoses them all as CTTH and episodic migraine then chronic migraine is missed. In clinical practice the concept of chronic migraine is particularly helpful. Given that the lifestyle advice is identical for both TTH and migraine, and that the range of therapeutic options for preventive treatment in migraine is so much greater, the clinician loses absolutely nothing diagnosing chronic migraine, and the patient has much to gain. For research there are other imperatives.

Management

The management of CDH can be very rewarding. Most patients overusing analgesics respond very sensibly when the problem is explained.

The keys to managing daily headache are:

• ◆ Exclude treatable causes (see Table 8)
• ◆ Obtain a clear analgesic history
• ◆ Make a diagnosis of the primary headache type involved

Medication overuse

Medication overuse is defined as consuming an acute attack therapy on 10 days or more per month, expect for paracetamol where 15 days is allowed under current guidance. It is essential that analgesic overuse be reduced and eliminated if one is to see the underlying headache phenotype and start to manage the problem. Patients can reduce their use by, as an example, 10% every week or two, depending on their circumstances, or if they wish, and there is no contraindication, by immediate cessation of use. Either approach can be facilitated by first keeping a careful diary over a month or two to be sure of the size of the problem. A small dose of an NSAID, such as naproxen 500 mg twice daily if tolerated, will take the edge off the pain as analgesic use is reduced, as does a greater occipital nerve injection. It is a useful aside that NSAID overuse does not seem to be a common issue in daily headache when they are dosed once or twice daily, whereas with more frequent dosing problems may develop. When the patient has reduced analgesic use substantially a preventive should be introduced. It must be emphasized that preventive therapies most often simply do not work in the presence of analgesic overuse. Thus, the patient must reduce the analgesics or the entire attempt to use the preventive is largely wasted, although this helpful rule must have some limitations that require study. The most common cause of intractability to treatment is the use of a preventive when analgesics continue to be used regularly. For some patients this is very difficult, and often one must be blunt that some degree of pain is inevitable in the first instance if the problem is to be controlled.

Some patients with medication overuse will require admission for detoxification. Broadly this consists of two groups—those who fail outpatient withdrawal or those who have a significant complicating medical indication, such as brittle diabetes mellitus, or complicating medicines, such as opioids, where withdrawal may be problematic as an outpatient. When such patients are admitted acute medications are withdrawn completely on the first day, unless there is some contraindication. Antiemetics, such as domperidone orally or as a suppository, and fluids are administered as required, as well as clonidine for opioid withdrawal symptoms. For acute intolerable pain during the waking hours aspirin (1 g intravenously) is useful and at night chlorpromazine by injection, after ensuring adequate hydration. If the patient does not settle adequately over 3 to 5 days a course of intravenous dihydroergotamine (DHE) can be employed as Raskin described. As time goes by, one feels that DHE is indispensable in this setting. Often 5HT₃ antagonists, such as ondansetron and granisetron, will be required with DHE because it is essential to ensure that the patient does not have significant nausea.

Preventive treatments

The tricylic antidepressants (TCAs), amitriptyline, dosulepin (dothiepin), and nortriptyline, at doses up to 1 mg/kg are very useful in patients with chronic migraine. TCAs are started in low dose (10–25 mg) daily and best given 12 h before the patient wishes to wake up to avoid excess morning sleepiness. The other very useful medications for these patients are the anticonvulsants, such as valproate, topiramate, and gabapentin. For valproate, doses up to 1500 mg daily are used, starting at 200 mg twice daily and increasing to 400 or 600 mg twice daily as tolerated over 2- to 4-week intervals. The blood count and liver enzymes should be checked at baseline and the various side effects explained to patients, especially the fetal abnormalities to women. For topiramate one can start at 25 mg nightly and increase by 25 mg every 10 to 14 days to aim for 50 mg twice daily. For gabapentin the dose is 1800 to 3600 mg daily; it is very well tolerated, although probably less effective from a population viewpoint. For some patients flunarizine can be very effective, as can methysergide or phenelzine. Recently, botulinum toxin type A (onabotulinum toxin) has been shown in a randomized controlled trial to be useful in chronic migraine. One might consider its use, for example, in medically-refractory chronic migraine perhaps after three preventive classes have failed.

New daily persistent headache

New daily persistent headache (NDPH) is a clinically useful concept with a range of important possible causes because some are very treatable (Table 9). From a nosological point of view all that is mentioned here could be placed within various categories of the IHS classification, and indeed the IHS refers to primary NDPH. However, the term as employed here serves both patients and clinicians by highlighting a group of conditions, some of which are curable, and encompasses the IHS term under the primary featureless form of NDPH.

The patient with NDPH presents with a history of headache on most if not all days, starting from one day to the next. The onset of headache is abrupt, often from one moment to the next, but at least in less than a few days with three suggested as an upper limit. The typical history is for the patient to recall the exact day and circumstances, so from one moment to the next a headache develops that never leaves them. This presentation triggers certain key questions about the onset and behaviour of the pain. The pressing issues arise from considering the secondary headache possibilities. Although subarachnoid haemorrhage is listed for some logical consistency, as the headache may certainly come on from one moment to the next, it is not likely to produce diagnostic confusion in this group of patients. Suffice it to say that subarachnoid haemorrhage is so important that it must always be considered if only to be excluded, either by history or by appropriate investigation.

Primary NDPH

Initial descriptions of primary NDPH recognized that it occurs in both males and females. Migrainous features were common, with unilateral headache in about one-third and throbbing pain in about one-third. Nausea was reported in about one-half of the patients, as were photophobia and phonophobia. A number of these patients have a previous history of migraine, but not more than one might expect given the population prevalence of migraine. It is remarkable that the initial report noted that 86% of patients were headache free at 24 months. It is general experience among those interested in headache management that primary NDPH is perhaps the most intractable and least therapeutically rewarding form of headache. In general one can classify the dominant phenotype—migraine or TTH—and treat with preventives according to that subclassification.

Bullet list 4 Other secondary headaches

Secondary NDPH

The secondary causes of the syndrome of NDPH are worthy of consideration, because they have distinctive clinical pictures that can guide investigation (Bullet list 4).

Low cerebrospinal fluid volume headache

The syndrome of headache as a result of persistent low cerebrospinal fluid volume is an important diagnosis not to miss. The more immediately obvious version of this problem is encountered commonly after lumbar puncture. In that situation the headache usually settles rapidly with bed rest. In the chronic situation the patient typically presents with a history of headache from one day to the next. The pain is generally not present on waking, worsens during the day, and is relieved by lying down. Recumbency usually improves the headache in minutes, and it takes only minutes to an hour for the pain to return when the patient is again upright. The patient may give a history of an index event: lumbar puncture or epidural injection, or a vigorous Valsalva manoeuvre, such as with lifting, straining, coughing, clearing the eustachian tubes in an aeroplane, or multiple orgasms. Patients may volunteer, or a history may be obtained, that soft drinks with caffeine provide temporary respite. Spontaneous leaks are recognized, and the clinician should not be put off the diagnosis if the headache history is typical when there is no obvious index event. As time passes from the index event the postural nature may be less obvious; certainly cases with an index event several years before the eventual diagnosis are recognized. The term ‘low volume’ rather than ‘low pressure’ is used, because there is no clear evidence at which point the pressure can be called low. Although low pressures, such as 0 to 5 are often identified, a pressure of 16 cmCSF has been recorded with a documented leak. One should be aware of the possibility of the development of subdural collections in patients with low cerebrospinal fluid volume headaches, which makes imaging before any invasive studies all the more important.

The investigation of choice is MRI with gadolinium, which produces a striking pattern of diffuse pachymeningeal enhancement, although in about 10% of cases a leak can be documented without enhancement. The finding of diffuse meningeal enhancement is so typical that in the clinical context immediate treatment is appropriate. It is also common to see Chiari malformations on MRI with some degree of descent of the cerebellar tonsils. This is important because surgery in such settings simply worsens the headache problem. It seems appropriate that any patient being considered for such surgery for a headache indication should first be reviewed by a neurologist. To investigate further cerebrospinal fluid pressure may be determined, or preferably a leak sought with indium-111-labelled DPTA (diethylaminetriaminepentaacetic acid) cerebrospinal fluid studies that can demonstrate the site, early emptying of tracer into the bladder, or lack of progression of tracer over the cerebral convexities.

Treatment is bed rest in the first instance. False-positive transient improvement in persistent low cerebrospinal fluid volume headache with chiropractic and other similar therapies is recognized where the treatment necessitates the patient lying down for a prolonged period for the therapy. Intravenous caffeine (500 mg in 500 ml saline administered over 2 h) is a standard and often very efficacious treatment. The ECG should be checked for any arrhythmia before administration. A reasonable practice is to carry out at least two infusions separated by 4 weeks after obtaining the suggestive clinical history and MRI with enhancement. As intravenous caffeine is safe, and can be curative, by an unknown mechanism, it spares many patients the need for further tests. If that is unsuccessful, an abdominal binder may be helpful. If a leak can be identified, either by the radioisotope study or by CT myelogram, or spinal T ₂-weighted MRI or more recently some radiologists using intrathecal Gd with MRI, an autologous blood patch is usually curative. In more intractable situations where as leak is not identified theophylline is a useful alternative that offers outpatient management, although its onset of action is rather slow.

Raised cerebrospinal fluid pressure headache

As is the case for low cerebrospinal fluid pressure states, raised cerebrospinal fluid pressure as a cause of headache is well recognized by neurologists. Brain imaging can often reveal the cause, such as raised pressure due to a space-occupying lesion. The particular setting in which patients enter the spectrum of NDPH are those with idiopathic intracranial hypertension who present with headache without visual problems, particularly with normal fundi. It is recognized that intractable chronic migraine can be triggered by persistently raised intracranial pressure. These patients typically give a history of generalized headache that is present on waking, and gets better as the day goes on. It is generally worse with recumbency. Visual obscurations are frequently reported. Fundal changes on raised intracranial pressure would make the diagnosis relatively straightforward, but it is in those without such changes that the history must drive the investigation. Patients often report a curious whooshing sensation in the occipital region.

Brain imaging is mandatory if raised pressure is suspected, and it is most simple in the long run to obtain an MRI scan, and include MR venography (MRV). The cerebrospinal fluid pressure should be measured by lumbar puncture, taking care to do so when the patient is symptomatic, so that both the pressure and response to removal of cerebrospinal fluid can be determined. A raised pressure and improvement in headache with removal of cerebrospinal fluid are diagnostic of the problem. The fields should be formally documented even in the absence of overt ophthalmic involvement. Initial treatment can be with acetazolamide (250–500 mg twice daily). The patient may respond in weeks with improvement in headache. If this is not effective topiramate has many actions that may be useful in this setting: carbonic anhydrase inhibition, weight loss, and neuronal membrane stabilization probably through actions on phosphorylation pathways. A small number of severely disabled patients who do not respond to medical treatment will come to intracranial pressure monitoring and even shunting. This is exceptional and not undertaken without careful work-up.

Post-traumatic headache

The issue of post-traumatic headache is vexed. The Headache Classification Committee accepts the existence of such a syndrome. Much of the scientific discussion becomes marred by the often-quoted medicolegal morass concerning delayed effects of head injury. The term is used here to indicate trauma in a very broad way. NDPH may be seen after a blow to the head but more commonly after an infective episode, typically viral, or even malarial meningitis. A recent series identified that one-third of all patients with NDPH reported the headache starting after an influenza-like illness. The patient may note a period in which they had a significant infection: fever, neck stiffness, photophobia, and marked malaise. The headache starts during that period and never stops. Investigation reveals no current cause for the headache. It has been suggested that some patients with this syndrome have a persistent Epstein–Barr virus infection, but this syndrome is anything but clearly delineated. A complicating factor will often be that the patient had a lumbar puncture during that illness, so a persistent low cerebrospinal fluid volume headache needs to be considered first. Post-traumatic headache may be seen after carotid artery dissection, subarachnoid haemorrhage, and following intracranial surgery for a benign mass. The underlying theme seems to be that a traumatic event involving the dura mater can trigger a headache process that lasts for many years after that event.

The treatment of this form of NDPH is substantially empirical. TCAs, notably amitriptyline, and anticonvulsants, valproate, topiramate, and gabapentin, have been used with good effects. The monoamine oxidase A inhibitor phenelzine may also be useful in carefully selected patients. On the positive side, the headache seems to run a limited course of 3 to 5 years in most patients, so will eventually settle. It can certainly be very disabling in that period.

Other important forms of secondary headache

Giant cell arteritis (temporal arteritis)

This is an important cause of headache because delay in steroid treatment may result in blindness due to retinal artery ischaemia. It is also known as temporal arteritis or cranial arteritis. Patients are usually older with focal tenderness of the scalp, which may be provoked markedly by resting the head on the pillow. Jaw claudication provoked by chewing is a characteristic, but relatively uncommon, feature. Constitutional symptoms are common, particularly weight loss, malaise, or polymyalgia rheumatica. An elevated erythrocyte sedimentation rate (ESR) is a strong pointer to the diagnosis. The temporal artery may be tenderly inflamed, swollen, or pulseless. On suspicion of this diagnosis, steroid treatment should be started pending the result of temporal artery biopsy. Treatment is very often long term and requires careful monitoring for reactivation and the side effects of corticosteroids.

Cervicogenic headache

It is a time-honoured concept that the neck is responsible for many headaches. Unfortunately, as with much of history, the good story is often ruined by the facts. Although there is little doubt that there is a rich overlap between the innervation of intracranial pain-producing structures by the ophthalmic division of the trigeminal nerve, and the posterior fossa and high cervical innervation by branches especially of the C2 dorsal root, causality is another issue. The Headache Classification Committee recognizes that head pain can arise from the neck and labels this ‘cervicogenic headache’. The term has been used by others to define a syndrome that is so poorly described as to be useless in practice. Most patients with neck discomfort and headache referred to speciality practice have migraine. They will have neck stiffness or discomfort as a premonitory symptom that can clearly persist in all stages of the attack. They may respond to local therapies, such as greater occipital nerve injection; however, this implies no more than triggering, and is to be expected. The pursuit of neck pathology and the treatment of patients with migraine by manipulative or physical means have no support in the controlled literature, and are rarely of long-lasting value.

Back to top

Further reading