Cluster headaches (CH) are arguably the most painful primary headache syndromes known to man, and are often termed ‘suicide headaches’, a term I believe understand better than many having observed my fathers suffering on a daily basis for my entire life. CH has a prevalence around 0.2%1, and presents as a strictly unilateral severe headache with concurrent ipsilateral autonomic symptoms (see Fig.1)2. It is a severely debilitating disease with 96% of patients reporting disease-related lifestyle changes3, therefore effective management of the condition is vital to maintaining a reasonable quality of life for sufferers. 85% of sufferers have episodic cluster headache (ECH)4, with the remainder having chronic cluster headache (CCH) in which attacks occur for more than one year without remission or with less than 1 months remission2. Progression from the episodic to chronic form of the condition is common, suggesting significant similarities in the underlying pathophysiologies, so some studies relating to both conditions will be discussed in this essay. Challenges in CCH management are numerous, stemming mainly from the condition’s severe symptoms, its complex pathophysiology and rarity. Underdiagnosis confounds this rarity and translates to an underrepresentation in research, meaning that the arsenal of evidence-backed medications available to the physician is often inadequate to tackle the disabling symptoms of sufferers of a heterogeneous condition, and more invasive surgical, and unproven pharmacological, treatments are administered on compassionate grounds.
Misdiagnosis and mismanagement
The first challenge for the physician is making the correct diagnosis of CCH so that appropriate treatment can begin, as some of the symptoms (notably the location of pain, nausea, photo- and phonophobia) are non-specific, often leading to incorrect diagnoses by physicians.
Internet surveys have provided considerable insight into the alarmingly common prevalence of misdiagnosis and mismanagement. They are important in facilitating the collection of data from sufficiently large cohorts for uncommon conditions such as CCH allowing appropriately powered study of certain characteristics, such as the diagnosis time, when clinical interviews in the past may not have. Bahra et al.5 conducted an internet survey of CH patients, showing that although mean time to diagnosis has decreased since the 1950’s, the average number of GP’s seen prior to diagnosis remains at 3 (fig. 2). Klapper et al.6 conducted a similar survey finding similar numbers of physicians were consulted before patients were correctly diagnosed, but if the IHS criteria for CH had been strictly adhered to in making the diagnosis, given their overt symptoms, 87% of the respondents should have been correctly diagnosed by the first physician seen.
Apart from the delay in appropriate management this causes, during this time patients are often referred to (incorrectly), or autonomously consult, other specialists due to their symptomology. Treatment may be undertaken aimed at incorrect diagnoses such as sinusitis (ie. sinus washout), or teeth may be extracted for pain in the upper jaw area. The startling prevalence of this mismanagement is illustrated in figure 2 and illustrates the challenge presented to management by a rare condition which physicians are often unfamiliar with and which presents with some non-specific symptoms6. NICE doesn’t detail any guidelines for diagnosis of CCH, which likely contributes to this misdiagnosis, and even if the physician has read the International Headache Society diagnostic criteria, they still may not recognize the condition. This could be overcome by incorporating diagnostic criteria into national guidelines, and perhaps altering the criteria so that they include more of CH’s most distinctive features such as its circadian rhythmicity and the paroxysmal nature of attacks with abrupt cessations.
Pharmacological treatment
When correctly diagnosed, although effective and evidence-backed treatments exist for CCH, it is often the case that migraine therapies such as propranolol and pizotifen are prescribed6 despite the fact that evidence exists that they are ineffective in CCH7. This highlights the need for physicians less familiar with the rare condition to follow treatment guidelines. Such guidelines have been created by the European Federation of Neurological Societies8 but are yet to be adopted by NICE, which could prevent physicians mismanaging CCH. The condition’s rarity also means that few high quality drug trials are carried out, with the most comprehensive recent meta-analysis of pharmacotherapy9 only finding 27 suitable randomized controlled trials (RCTs) of CH treatments. Here I shall outline representative treatments that highlight the challenges faced in CCH management, as well as future treatments and the obstacles they encounter in development.
Acute therapy
Given the paroxysmal nature of the debilitating CCH attacks, acute therapies are aimed at aborting attacks as quickly as possible. This limits the number of administration routes suitable and this presents certain challenges to management of its own. Most current acute treatments are aimed at the trigeminovascular system, as this was generally considered central to the pathology of the condition (pre-2000). The first clues for vascular involvement were provided in the late 60’s by Ekbom’s group10, who discovered that nitric oxide could trigger CH attacks, and later conducted a carotid angiography showing venous and arterial dilation occurred during CH attacks11. In the mid 90’s Goadsby’s group showed that during spontaneous attacks there is a release of the cranial neuropeptide CGRP indicating activation of the trigeminovascular system12.
Sumatriptan, the abortive ‘drug of choice’ according to the British National Formulary (BNF), is a 5-HT1D receptor agonist, and was initially trialed for treatment of CH due to its success in treating migraine13. It is thought to have its effects through triggering the constriction of cranial arteries, as well as possibly decreasing the activity in the trigeminal nerve14. A recent meta-analysis of two RCTs gave a summary odds ratio of 6.22 in favor of 6mg sumatriptan given subcutaneously (s.c) over placebo (p<0.00001) at reducing attack severity from moderate/severe/very-severe intensities to mild/pain free states within 30 minutes of administration9. For those with an aversion to needles, the nasal spray has previously been the only option, but this has a much lower efficacy than subcutaneous administration. Recently, a nitrogen powered needle-less injection system has been developed, with a randomized crossover trial showing it to have similar bioavailability as the current needle-based auto-injector system15. Another trial showed that sufferers who had used both systems reported a significant increase in tolerability and functionality of the needle-less system over the needle-based system, and no difference in efficacy16. It is available in the US, and should be available in the EU in the near future.
Treatment (average daily dose) / Type / Yearly cost/CCH patient34Sumatriptan s.c
(1.8 * 6mg) / Abortive / £23651
Oxygen
(3.6 * inhalation) / Abortive / £872
Zolmitriptan nasal spray (1.5 * 5mg) / Abortive / £7064
Verapamil (474mg) / Preventative / £286
Lithium (746mg) / Preventative / £131
Topimarate (178mg) / Preventative / £371
Abortive therapies such as sumatriptan (and commonly ergotamines) are often self administered by sufferers on a daily basis to control their pain. Physicians must monitor their patients for the development of a medication overuse headache (MOH). This occurs in 5-10% of CCH sufferers and manifests as an enhancement of CH attack frequency or a chronic daily headache with dull and generalized pain17. Withdrawal of the offending treatment is the only effective cure for the headache17 although this only practical if an alternative effective acute treatment is available or preventative treatment possible, as often patients will prefer MOH and shorter or less severe CH attacks if another treatment is not as effective. The limited arsenal of effective treatments makes this a considerable management problem for physicians, and it’s hoped that in the future the development of more alternative pharmacological/surgical treatments shall circumvent this.
High flow (6-12l/min) oxygen therapy is an effective acute therapy that does not cause MOH. Although it’s mechanism was not known prior to the discovery of its efficacy, it is now believed that some of its effect may be achieved through inhibiting neuronal activation in the trigeminocervical complex following studies in the rat18. Cohen et al.19 conducted a crossover RCT (control: air inhalation) showed that 78% of patients treated with 12l/min 100% oxygen for 15 minutes were pain free by the end of the treatment, compared to 20% of those treated with air (p<0.001). The results of these trials however could be misleading, as oxygen sometimes acts to simply delay the attack rather than aborting it5. Physicians therefore need encourage patients to maintain a headache diary so as to track the actual effects of the treatment on pain load rather than on isolated attacks.
Oxygen therapy causes few side effects and therefore it is surprising that it is rarely prescribed by GP’s5. This could in part be due to the practicalities of administering the treatment, as obviously it is not always practical for patients to bring an oxygen canister with them to work, for example, to abort attacks. It is for this reason that management of CCH requires a logical approach by the physician, possibly utilizing combinations of the acute therapies for different situations ie. at work and in the house, balancing cost (table 1), side effects, and compatibility with daily life.
Prophylactic medications
Prophylactic medications are aimed at decreasing the occurrence of attacks. Verapamil and Lithium are the most commonly prescribed preventatives. A crossover RCT comparing the two drugs showed that 50% of those treated daily with 360mg Verapamil and 37% of those treated with 900mg Lithium daily showed a significant decrease in headache index over 2 weeks of treatment compared to the run-in period20.
Both treatments present management issues however, as a result of their side-effects. For example, verapamil’s L-type calcium channel blocking effects cause drops in blood pressure and heart rate at the high doses used to treat CCH (up to 900mg/day c.f 100-200mg/day for cardiac indications). Therefore physicians must titrate the dose of verapamil so that a balance of CCH preventative effects and tolerable side effects is reached, as well as carefully monitoring the ECG at regular intervals for prolonged Q-T intervals indicative of atrio-ventricular block.
Many other prophylactic pharmacological interventions, such as sodium valproate and various triptans, have been tested in open label or non placebo-controlled trials, however such studies provide poor quality evidence of efficacy that may be misleading. In the future, more evidence-backed effective prophylactic treatments are desperately needed, however creation of feasible RCTs testing the efficacy of such drugs in CCH is very challenging. One attempt to create a double blind RCT to test the efficacy of frovatriptan based on positive open label trial results had to terminate the trial prematurely on two counts; firstly the enrollment was very slow, owing to the rarity of the disease, and secondly all 13 of the participants who took part in the trial breached major protocols – mainly the prohibition of prophylactic therapy and acute treatment with triptans21. This highlights the difficulty of creating RCTs of CCH prophylactics; other prophylactics need be stopped and, in this case, triptans couldn’t be administered so as to isolate the lone effect of frovatriptan. Such is the pain of CCH that preventing use of arguably the most effective drugs (sumatriptan and verapamil) leads to an unacceptable level of suffering for the subject. This makes the design of future pharmacological therapies very difficult in CCH unless treatments are of an equal level of efficacy to current treatments, and/or attacks can be managed with other therapies that don’t interact with the results obtained from testing the new treatment.
Pharmacogenetics
The varying responses of sufferers to treatments highlights the extent of heterogeneity within the population of sufferers. Treatment involves prescription of medications sequencially, meaning that, combined with the time taken to reach correct diagnosis, optimal pharmacological management of the attacks may take years to be correctly assigned. Evidence provided by a clinic based sample observing patient depression and CH characteristics showed that depressed patients had undertaken twice as many prophylactic medications than those who weren’t depressed, highlighting the effect of multiple ineffective treatments22. This adds another complicating factor to the management of CCH patients, and emphasizes the importance of prompt diagnosis and correct management. In the future it’s hoped pharmacogenetic study will predict the treatments a person with a specific genotype will likely respond best to, reducing the time the condition exists poorly managed and hence decrease the suffering experienced by the patient and consequently reduce the development of depression. Already such trials are taking place, one study showing that carriers of the 825 C<T polymorphism for G-protein Beta-3 sub-unit are three times more likely to respond to treatment with triptans compared with carriers of the 825CC genotype (p=0.037)23. Genetic studies are also providing key clues to the underlying pathophysiology of CCH and are providing targets for potential treatments in the future.
Medically intractable cases
Nevertheless, 20% of CCH patients obtain a less than 50% reduction in CH symptoms from any of the conventional therapies and are classed as drug resistant (drCCH)24. The search for an effective treatment outside of guidelines by the physician on compassionate grounds provides many challenges of its own; unproven treatments, pharmacological and surgical may be employed with varying degrees of risk, and patients often resort to alternative therapies with little supporting evidence and occasionally recreational drugs.
Recreational drugs
CH sufferers, particularly those who are drug resistant, are often desperate for a treatment and a thriving internet community of sufferers exists (www.clusterheadache.com). Patients are often referred to such forums by physicians, and this can have the effect of pushing the field of research forward through chance findings, but also can lead to the patient putting themselves at risk as potentially dangerous advice is exchanged. One such startling, but nevertheless interesting, finding was uncovered by an internet survey of 53 forum users who’d self-administered lysergic acid diethylamide (LSD) and/or the related indolalkylamine psilosybin (found in ‘magic mushrooms’)25. 25/48 psilocybin users and 7/8 LSD users reported considerable cluster attack remission after extremely low numbers of doses (1-3) at sub-hallucinogenic levels. The findings of such an unblinded and non placebo-controlled retrospective survey should be viewed with caution, especially seeing as the chemical composition of the drugs would have varied considerably. This aside, to my knowledge there has never been a report of a drug which can terminate a cluster period with a single dose, and hence this is an important finding.