Italian guidelines for primary headaches: 2012 revised version

Indications They are indicated for the treatment of moderate or severe attacks (level of recommendation I). Efficacy RCT have demonstrated the efficacy of triptans, not only for headache but also for accompanying symptoms and functional disability [8‐19] (Table 2). The consistency of efficacy of triptans in the treatment of multiple attacks and in long-term treatment (no development of tachyphylaxis) has also been shown [8, 9, 20]. They were also effective in menstrual-related migraine attacks [21]. Head-to-head studies did not establish the superiority of one triptan over the others [22]. Preference trials suggest that no ideal triptan exists for all patients, but the treatment must be tailored taking into account the characteristics of each patient and of the attacks [23‐26]. Headache recurrence occurs in about 25–40 % of patients [8, 27, 28]. Observations When a triptan is administered early at the beginning of attack, it has a greater efficacy [29‐31]. About 25–35 % of patients do not respond to a particular triptan, in which case other triptans can be tried [32‐37]. In the case of an unsatisfactory response to a triptan or headache recurrence, a NSAID can be used [38]. Sumatriptan is available in all formulations (subcutaneous, tablet, nasal spray, suppository); subcutaneous sumatriptan is the most effective drug in the class [39‐41]. Rizatriptan and zolmitriptan are available in rapid-dissolving formulations (RPD) which have an effectiveness similar to that of tablet formulations of the same drugs at the same dosages [42, 43]. Pharmacokinetics findings do not show higher blood levels reached at shorter times for RPD formulations. The latter can, however, be useful because they are easier to use without need of water, particularly when moderate or severe nausea is present. Naratriptan is not available in Italy. In some studies oral triptans at lower dosages have not demonstrated to be superior to some NSAIDs, simple or combination analgesics [21]. Oral triptan formulations are superior to oral ergotamine which has a low bioavailability (<1 %). Ergotamine and dihydroergotamine have an efficacy similar to that of triptans but induce more frequent adverse events [44‐47]. The excessive use of triptans (≥10 days a month) exposes the patient to the risk of migraine chronification and should be avoided [48]. The concurrent use of triptans and NSAIDs seems to have a greater efficacy compared with that of triptans alone and is associated to lower headache recurrence [41‐52]. Side effects are mostly mild-to-moderate, of short duration (10–15 min) and include triptan syndrome (chest and neck tightness, chest pain) in 4–5 % in patients treated with s.c. sumatriptan and 2–4 % with the oral formulation, fatigue, somnolence, dizziness and facial flush [6, 204‐207]. Cardio- and cerebrovascular severe adverse events (myocardial infarction, ictus), without an established cause–effect relationship, were rarely reported. ECG modifications are also rarely reported [53]. Dystonic crises, akathisia, euphoria, can also rarely occur. Contraindications to triptans are uncontrolled blood hypertension, coronary artery disease, history of ischemic stroke, pheripheral artery disease, pregnancy and lactation and age >65 years.

Pregnancy and breastfeeding From findings obtained by pregnancy registries, a greater number of preterm newborns or newborns with low birth-weight due to the use of sumatriptan during pregnancy have been described. In the case of repeated administration of sumatriptan in the first trimester there is no increased risk of newborn malformations but the sumatriptan use in the second and third trimester is associated with atonic uterus and bleeding >500 ml at delivery. Information on the safety of triptans during breastfeeding is limited but reassuring, because the minimal quantities secreted with milk are insufficient to induce adverse events to the child [54]. According to Italian Health Ministry Regulatory Agency the use of triptans is not recommended under 18 years of age with the exception of sumatriptan nasal spray 10 mg and zolmitriptan nasal spray 2.5 mg, which may be used in patients over 12 years of age [55, 56]. Also, according to this Agency, the use of triptans after 65 years of age is not recommended. They can be used only with a therapeutic plan approved by an Ethical Committee and with informed consent [57].

SSRI antidepressants A serotoninergic syndrome can occur in the case of contemporary use of triptans and consists in motor incoordination, marked asthenia and hyperreflexia. MAO-A inhibitors They should be suspended at least 2 weeks before starting triptan treatment. Propranolol increases the concentration of rizatriptan. In the case of concomitant administration of propranolol, rizatriptan should be used at the dosage of 5 mg for the single attack, at the maximum daily dosage of 10 mg. After taking propranolol, it is necessary to wait at least 2 h before taking rizatriptan. Drugs which are metabolized via CYP450 Eletriptan, rizatriptan and zolmitriptan may interact with drugs metabolized via CYP450 such as oral contraceptives and antimicotics. The clinical relevance of these observations needs to be clarified [58].

Indications They are indicated for the treatment of mild or moderate attacks or when triptans are contraindicated or ineffective [59, 60]. Efficacy The most consistent evidence of efficacy is available for paracetamol, acetylsalicylic acid (ASA), lysine acetylsalicylate, naproxen sodium, ibuprofen, diclofenac sodium and potassium, metamizole and ketorolac, whereas the evidence of efficacy for other NSAIDs is more limited [61‐65]. Head-to-head studies have not shown a clear-cut superiority of a NSAID over another. Few studies have evaluated the efficacy of analgesics and NSAIDs on associated symptoms and functional disability. There is evidence only for ASA, salicylates, ibuprofen and diclofenac sodium. There are no studies supporting the consistency of efficacy and recurrence rates for the majority of analgesics/NSAIDs. The efficacy of ASA and other NSAIDs on migraine aura has never been tested.

Observations ASA is recommended in patients with cardio- and cerebrovascular comorbidities. Paracetamol is first-choice drug for migraine attacks during pregnancy. The excessive use of NSAIDs (≥15 days a month) should be avoided for the risk of migraine chronification. There is evidence of the use of ketorolac i.v. in the emergency department (ED), supporting its efficacy in the treatment of migraine attacks even if the results are less favourable than those obtained with prochlorperazine [66, 67]. In the same setting, ketorolac has been demonstrated to be more effective than sumatriptan nasal spray [68]. Metamizole, both by oral and intravenous route, has been demonstrated to be effective in the treatment of migraine attacks, but the risk of agranulocytosis and hypotension as relevant side effects should be considered [6]. Side effects consist mainly in gastrointestinal adverse events (from gastric pain to gastric or duodenal ulcer). The percentage of adverse events found in clinical trials concerning the use of NSAIDs for migraine attack are lower than those detected in studies regarding their daily use. These adverse events, occasional in migraine patients using sometimes NSAIDs, can occur with higher frequency in the case of daily or almost daily use by chronic migraine patients. Contraindications include severe renal and hepatic failure, hemorrhagic risk, gastric or duodenal ulcer.

Indications Their use should be restricted to low frequency, severe attacks unresponsive to other drugs for their potential risk of abuse [69]. Efficacy Ergotamine tartrate with or without caffeine and dihydroergotamine have been demonstrated to be effective versus placebo or versus an active drug in reducing migraine headache [70, 71]. Ergotamine tartrate is not effective on nausea or vomiting; rather, because of the interaction with dopaminergic receptors, it may itself induce or increase nausea or vomiting accompanying the migraine attack [72]. Studies are lacking for its use for migraine aura. Ergotamine tartrate administration is associated with low incidence of recurrence (<30 %) [70]. Oral ergotamine is inferior to sumatriptan and eletriptan [44, 45].

Observations Dihydroergotamine, the drug of the class with the best risk–benefit ratio, is not available in Italy. Because nausea and vomiting may worsen due to the administration of ergot derivates, the contemporary administration of an antiemetic is generally indicated [72]. Caffeine doubles the rate of absorption of ergotamine and increases its peak blood concentration. This explains the development of combination formulations.

Patients who overuse ergotamine derivatives may develop rebound headaches. The abuse of ergot derivatives may induce an increase in the frequency of attacks and develop into a chronic headache. Therefore, these drugs are recommended for sporadic attacks and cannot be used for more than 10 days/month [48].

Major side effects Nausea, vomiting, diarrhoea and ergotism [73]. Ergotamine has a teratogenic effect [74]. Contraindications Cardiovascular and cerebrovascular diseases, uncontrolled blood hypertension, Raynaud disease, renal failure, pregnancy and lactation. Pharmacological interactions Triptans: an ergot derivate should not be administered within 6 h after the administration of a triptan. Beta-blockers: an increase in the risk of peripheral vasoconstriction has been observed in patients who also used beta-blockers. The majority of patients are able to tolerate such association; caution is necessary for particularly sensitive patients [58].

Indications They have the same indications of simple analgesics and NSAIDs. Few studies have been performed on these combination drugs [75]. Efficacy has been demonstrated only for the association with acetylsalicylic acid, paracetamol and caffeine. Recent trials have demonstrated a significant efficacy on migraine attacks of moderate intensity and moderate disability [76]. This association has been demonstrated to be effective in migraine attacks related to the menstrual cycle [76]. Recent data suggest that the effectiveness of the association of indomethacin + caffeine + prochlorperazine is similar to that of triptans, even though supporting studies are needed [77, 78]. Combination analgesics available in Italy include acetylsalicylic acid + acethaminophen + propyhenazone, acetylsalicylic acid + acethaminophen + indomethacin (with/without caffeine), and acetaminophen + propyphenazone and acetaminophen + codeine. Dosages of active substances in the combinations are different from those tested for migraine attacks. The efficacy of acetylsalicylic acid + acethaminophen + propyphenazone and butalbital + propyphenazone + caffeine has never been investigated in RCT for migraine.

Observations To avoid the risk of abuse, the use of combination analgesics should be limited to ≤10 days/month; abuse can lead to headache chronification [79‐81]. Side effects and contraindications in combination analgesics are the same as those for each component. Caffeine may induce anxiety and insomnia.

Indications Antiemetics are to be considered adjuvants, especially when nausea and vomiting are prominent [82]. Efficacy Most studies have concerned the association of antiemetics with analgesics and NSAIDs (naproxen, paracetamol, tolfenamic acid) or dihydroergotamine [72]. These associations have been proposed to improve the absorption of the symptomatic drugs and to act as adjuvants in reducing nausea or vomiting associated with the attacks. No RCT, however, clearly demonstrated a superiority of this association over NSAIDs alone. There is some evidence that suggests that the use of an antiemetic may improve the efficacy of a triptan [83]. Metoclopramide, prochlorperazine and chlorpromazine have also shown a modest antimigraine effect, besides a clear antiemetic effect [84, 85].

A modest antimigraine effect has been demonstrated for metoclopramide administered intramuscularly or intravenously [86].

Prochlorperazine or chlorpromazine administered intramuscularly or intravenously have been shown to be modestly effective in studies carried out in the ED. Oral prochlorperazine has also shown some partial efficacy [87, 88].

Dated findings on a limited number of patients support some efficacy of domperidone in preventing migraine attacks or reducing head pain intensity [326‐328].

Intramuscular or intravenous formulations can be used in the treatment of attacks of severe intensity in which nausea and vomiting are prevailing and in the case in which other symptomatic drugs are contraindicated or sedation is needed. They can be considered as single drugs for the treatment of migraine in particular clinical settings (i.e. emergency department).

Antiemetics are not recommended in patients with prolactinoma. The use of metoclopramide, chlorpromazine and prochlorperazine must be limited only to cases of extreme necessity in pregnancy and during breast feeding [89].

Simple or combination opioid analgesics Controlled studies have demonstrated the association of paracetamol with codeine, doxilamine or buclizine to be no more effective than paracetamol alone [90]. In a more recent study the combination of paracetamol + codeine has been shown to be more efficacious than ASA [91]. The association of ASA with dextropropoxyphen and phenazone was not more effective than ergotamine [92].

Also, butorphanol (not available in Italy) by intramuscular route has not shown to be more effective than dihydroergotamine administered intravenously in association with metoclopramide.

There are no studies comparing butorphanol nasal spray with other non-opioid symptomatic antimigraine drugs [93, 94]. A double-blind study comparing the efficacy of ketorolac (60 mg) and meperidine (75 mg)/promethazine (25 mg), both administered by intramuscular route, did not show a statistically significant difference between the two drugs [95].

More recently, tramadol administered by intravenous route alone or in combination with paracetamol has been demonstrated effective in the acute treatment of migraine [96‐98].

The Ad Hoc Committee has unanimously decided that this class of drugs does not represent a valid option for the symptomatic treatment of migraine attacks. This is due to the lack of data demonstrating their effectiveness compared with other symptomatic drugs and because of the potential risk of abuse and developing a chronic headache [99].

These drugs are to be preferred in the case of hypertension or tachycardia [116, 117].

Their efficacy as preventive drugs for migraine has been fortuitously demonstrated in migraine patients with concurrent hypertension. The efficacy of atenolol, nadolol, timolol, bisoprolol and nebivolol is supported by few controlled studies [118‐124]. Even though prophylactic treatment is generally not advisable in pregnancy, propranolol may be used with relative safety [125]. The abrupt suspension of these drugs can induce an increase in the frequency of migraine attacks and an increase of blood pressure.

Calcium channel antagonists are particularly recommended for patients with anxiety and insomnia [126].

EfficacyFlunarizine is the most used drug. Cinnarizine has a good antimigraine action although few studies have been carried out to investigate its efficacy in migraine [127, 128]. There are insufficient data supporting the efficacy of nimodipine and ciclandelate in migraine [129]. Therapeutic effects become evident only after some months of treatment. Cinnarizine induces the CNS side effects and accumulation phenomena less frequently.

Side effects of flunarizine are somnolence, asthenia, weight gain, depression and extrapyramidal symptoms in the long-term treatment and occur more frequently in elderly patients; cinnarizine can induce somnolence and epigastric pain which can be avoided by taking the drug on a full stomach, weight gain, extrapyramidal symptoms in the long-term treatment and occur more frequently in elderly patients [130‐132].

Among tricyclics, there are only two double-blind studies versus placebo supporting the efficacy of amitryptiline [133‐135].

Prophylactic effectiveness is obtained with doses lower than those used for depression (10–20 mg/day) [135].

The most frequent adverse events are antimuscarinic effects such as dry mouth, constipation and sedation. Often an increase in appetite (craving) with a consequent weight gain, or occasionally orthostatic hypotension and impotence, may occur [134].

Promising data have been obtained for venlafaxine but RCT have some methodological limitations [136‐138]. Although paroxetine, escitalopram, fluvoxamine and sertraline are sometimes used for migraine prophylaxis due to their good tolerability profile, available data are limited and contrasting [139‐143]. There are no data available for mirtazapine. There is also some positive evidence for fluoxetine [144, 145].

Venlafaxine can be useful in patients with depression and concomitant anxiety. Fluoxetine can induce insomnia, fatigue, tremor and epigastric pain. SSRI can interfere with 5HT₁ agonists.

Sodium valproate and topiramate are first-choice drugs in the treatment of high-frequency migraine attacks, chronic migraine, with and without symptomatic drug overuse and in the case of comorbid epilepsy [146, 147].

Sodium valproate and topiramate are effective in both migraine with and without aura and in chronic forms. Their long-term efficacy has been shown [148‐153]. The dosages useful for the prophylactic treatment of migraine are lower than those used for epilepsy. Gabapentin has a good tolerability profile [154]. In open studies, lamotrigine (50–200 mg/day) has been shown to be effective in the treatment of high-frequency migraine attack with aura [155‐157]. It is ineffective in migraine without aura [158]. Promising data have been obtained for zonisamide, levetiracetam, pregabalin and need to be confirmed.

Contraindications are hepatitis, pancreatitis, thrombocytemia for sodium valproate; liver and kidney insufficiency, kidney stones, glaucoma for topiramate; hypersensitivity to the drug for gabapentin and lamotrigine. All antiepileptics are contraindicated during pregnancy due to their teratogenic effect [159].

Adverse events caused by sodium valproate include asthenia, dizziness, tremor, alopecia, weight gain, menstrual disorders, hepatopathy and pancreatitis; for topiramate, asthenia, memory problems, anomia, weight loss, paresthesia, dysgeusia and depression; rare adverse events include metabolic acidosis, kidney stones, psychosis and narrow-angle glaucoma; for gabapentin asthenia, somnolence, ataxia, diplopia and constipation; for lamotrigine: skin rashes (which may be warning signs of Stevens–Johnson syndrome; to reduce this risk a slow titration is recommended), somnolence, gastroenteric disturbances and ataxia [160].

5HT-antagonists are the oldest drugs available for migraine prophylaxis and include methysergide, a semisynthetic derivative of ergometrine, which is not available in Italy and pizotifen, a serotonin antagonist with modest antihistaminic and cholinergic effects, which is a second-choice drug due to its side effects [161].

There are a few dated clinical studies supporting the efficacy of pizotifen [162‐165]. This drug has a long half-life (about 23 h) and can be used in a unique dose (0.5–1.5 mg), for cycles of 3 months.

The most frequent side effects are somnolence, increase in appetite, weight gain, xerosthomia and constipation [166].

IndicationsLisinopril and candesartan are second-choice drugs to be considered for patients with a concomitant hypertension [167‐169]. There is only one positive RCT for both drugs [170, 171].

Adverse events for lisinopril are asthenia, hypotension, dry cough, hyperkalemia, gastrointestinal disturbances and impotence; for candesartan asthenia, dizziness, tachycardia and hyperuricemy.

Other drugs Two ergot derivatives, controlled release dihydroergotamine 10 mg/day and dihydroergocryptine 20 mg/day (not available in Italy), can be taken into account as second-line treatments [172, 173]. Riboflavin at high doses (400 mg) showed a certain efficacy in preventing migraine, with few side effects (moderate abdominal pain, diarrhoea), in one RCT [174]. This drug is available as a galenic preparation in Italy. Among herbal remedies, butterbur root extract (Petasites hybridus), at the dosage of 150 mg/day, proved to be effective in two RCTs, while another herbal drug, Tanacetum parthenium, studied in several RCTs, gave more conflicting results [175, 176].

One controlled study versus placebo demonstrated the efficacy of Coenzyme Q10 (100 mg × 3 per day) [177]. The superiority of tiotic acid, a drug which also increases brain energetic metabolism, has also been shown compared with placebo [178]. Conflicting findings have been obtained for pidolate magnesium (400–600 mg/day) [179, 180]. The use of non-chelated formulations determines diarrhoea at clinically efficacious doses. It can be used in women with menstrual migraine and premenstrual syndrome, but a precise administration schedule has not been established (646–648). OnabotulinumtoxinA showed some efficacy in various open studies in migraine patients, but contradictory results emerged in double-blind controlled studies versus placebo concerning patients with episodic migraine [181]. These studies used different protocols, individualized or standardized, different inoculation sites and groups with different frequency of attacks (see chronic migraine).

Menstrual migraine In pure menstrual migraine or menstrually related migraine, a short-term prophylactic therapy around menses can be tried if menstrual cycles are regular and migraine attacks are predictable [182]. Two different strategies may be followed: (1) administration of symptomatic drugs, in particular triptans, on a regular basis instead of on demand; (2) administration of estrogens, to avoid the premenstrual oestrogen fall that is thought to be a main cause of menstrual migraine. In the first case, frovatriptan (5 mg/day) or zolmitriptan (5–7.5 mg/day) may be administered [183‐185]. Positive results have also been obtained in one RCT for sumatriptan 100 mg and for naproxen sodium (1,100 mg/day) [186]. The administration should be limited to the period −2 +4/5 days from menstrual onset. If the second option is chosen, estradiol gel (1.5 mg/day) or transdermal estradiol patch (100 μg) are probably the best choices [187, 188].

Chronic migraine Only recently some RCTs have been carried out in patients affected by the most disabling form of migraine, i.e. chronic migraine. Two drugs have so far showed some evidence of efficacy: topiramate and onabotulinumtoxinA. Topiramate (100 mg/day) has been partially effective also in patients with medication overuse; onabotulinumtoxinA (injection every 12 weeks) has given a statistically significant clinical benefit [189‐191].

Data supporting a significant efficacy of onabotulinumtoxinA have been obtained in patients with chronic migraine with or without symptomatic drug overuse [192‐194].

Scientific evidence supports the efficacy of acetylsalicylic acid (ASA), diclofenac, ibuprofen, ketoprofen, metamizol (dipyrone) and naproxen [195‐205].

ASA, ketoprofen and naproxen are first-choice drugs. ASA is especially useful in the case of comorbid cerebrovascular or cardiovascular disease, but is to be avoided in patients with gastric disease. Low doses of NSAIDs are usually sufficient to obtain a therapeutic result. Ibuprofen is reported to have minor gastric adverse effects. Comparative studies were not able to demonstrate the superiority of one drug because the comparisons were not based upon equivalent doses of the different molecules [196, 198‐201, 204, 205]. The assumption of NSAIDs or simple analgesics for 15 days or more per month, for >3 months, can induce a chronification of headache and a chronic medication-overuse headache.

Paracetamol has been tested in TTH patients showing good efficacy and tolerability [196, 200, 204]. It is a first-choice drug both in pregnancy and in patients with gastric disease but it should be used with caution in subjects with liver disease.

Combination analgesics with caffeine, codeine and butalbital share indications with simple analgesics and are more effective than the latter. The combination drugs with caffeine (paracetamol + caffeine, paracetamol + ASA + caffeine and ibuprofen + caffeine) are classified at recommendation level I [206‐210]. Also, the association of indometacin + prochlorperazine + caffeine was demonstrated to be effective in one study [207].

An excessive and frequent use of combination analgesics (for 10 days a month or more) should be avoided because of the high risk of drug abuse, headache chronification and drug-induced headache. Paracetamol + codeine (500 + 30 mg) and butalbital + propyphenazone + caffeine (50–150 + 125–175 + 25–75 mg) are recommended at the level III because of the addiction potential [211, 212]. Patients taking opiates or barbiturates should be informed about the risks deriving from an abrupt discontinuation and undergo a hospitalized drug discontinuation schedule. Combination analgesics share the same contraindications, drug interactions and side effects of the single components (for details see “Migraine”).

Antiemetics are generally not necessary in the usual management of TTH acute therapy. Intravenous metoclopramide, however, was found to be effective in the emergency treatment of TTH episodes [213]. Also, intravenous administration of chlorpromazine was effective in the ER TTH management [214].

Complementary or alternative drugs Promising results have been obtained with a topical preparation of peppermint oil 10 g and ethanol (90 %) to 100 [215, 216]. Tiger balm has also shown to have a modest but significant effect in inducing headache relief [217].

Among tricyclicsamitriptyline is a first-choice drug. It is also recommended in case of comorbidity with anxiety, insomnia, depression or migraine [219, 220]. Amitriptyline is utilized at much lower dose for TTH than that necessary to obtain an antidepressive effect. A slow titration of the drug is recommended to increase its tolerability and avoid adverse events.

Contraindications include cardiac arrhythmias, prostatic hypertrophy, glaucoma and epilepsy. It should be used with caution in the elderly because of its anticholinergic action. For side effects see paragraph on migraine. Also, clomipramine (150 mg/die) and desipramine (75 mg/die) may be useful [221, 222].

SSRI are less effective than amitriptyline but better tolerated [223, 224]. Fluvoxamine was the most studied molecule in this class. Also fluoxetin and paroxetin showed some evidence of efficacy.

Mirtazapine, a specific noradrenergic and serotoninergic antidepressant, is another first-choice drug which is especially indicated in the case of comorbid anxiety, insomnia and depression [225, 226]. It should be taken at low initial doses because it can induce sedation and sleepiness. Promising results come from venlafaxin, a noradrenalin and serotonin selective reuptake inhibitor (SNRI), which is able to reduce headache frequency independently of the association with anxiety or depression [227, 228]. Maprotiline and mianserine are other antidepressants that have shown to be effective in TTH preventive therapy [229, 230]. The latter was compared to both clomipramine and fluvoxamine with similar efficacy [221]. No conclusive results have been obtained for nefazodone [231] and ritanserine [232, 233], not available in Italy; sulpiride has been used at the dosage of 30 mg/day and has been shown to be more effective than paroxetine [234].

One RCT supports the use of tizanidine in chronic headache including TTH. The drug was slowly titrated over 4 weeks to 24 mg or the maximum tolerated dose showing moderate adverse effects (somnolence, dizziness, dry mouth, asthenia) [235, 236]. In one study cyclobenzaprine has demonstrated to be effective in the improvement or complete relief of headache in half of 20 TTH patients, whereas only partial relief was obtained in one-third of patients treated with placebo [237].

In dated studies diazepam (5 mg/day) seemed to be more effective than placebo in reducing headache frequency [238, 239]. It can be useful in the case of comorbid anxiety. Also, alprazolam (0.75 mg/day) has been demonstrated to be effective in TTH preventive therapy with a lower level of rating (level of recommendation III) [240].

Positive results have been obtained with topiramate (25–100 mg/die), which was tested at the initial dose of 25 mg/die titrated to 100 mg/die in chronic TTH patients in a recent open study [241]. Buspirone, a non-benzodiazepine anxyolitic drug, has been compared to amitriptyline, showing moderate results in a small sample of patients [220]. Further investigation is needed (level of recommendation III). L-5-hydroxytryptophan at the daily dose of 300 mg showed mild efficacy in the prophylaxis of chronic TTH in one RCT (level of recommendation III) [242]. Conflicting results have been found with onabotulinumtoxinA which may be attributed, at least in part, to variability of doses, study protocols, inoculation sites and headache frequency of enrolled patients. Further studies are needed (level of recommendation IV) [243, 244].

Its efficacy has been demonstrated for the subcutaneous (s.c.) and intranasal formulations. Sumatriptans.c. (6 mg) has been shown to be effective both on pain and associated symptoms [248‐250]. When administered for long periods, it maintains its efficacy without tachyphylaxis and with a good safety profile [251, 252]. The efficacy of sumatriptan nasal spray (20 mg in a nostril) has been shown within 30 min after administration in a double-blind randomized study [253]. The greater latency of action suggests its use for attacks lasting at least 45 min. Sumatriptan nasal spray does not have the indication for cluster headache in Italy. The most common adverse events are reactions in the site of injection, dizziness, paraesthesia, sensation of cold or warm and irritation of the nostril for the intranasal formulation. In 90 % of cases they are mild or moderate [254].

Its efficacy has been demonstrated for the oral and intranasal formulations. Oral formulation (5 and 10 mg) has been demonstrated to be effective at 30 min (reduction of 2 points of pain intensity in a 5-point scale) in one RCT [255]. In Italy the only dosage available is 5 mg (to be reached only once within 24 h if needed). Zolmitriptannasal spray (5 and 10 mg) has also been shown to induce CH relief in two RCTs, with a higher efficacy for the 10-mg formulation and with few adverse events for both dosages [256‐258].

A Cochrane analysis of six randomized studies, controlled versus placebo, demonstrated that sumatriptan and zolmitriptan are superior to placebo [254]. Adverse events of triptans include paraesthesia, asthenia, nausea, dizziness and irritation of the nostril for the intranasal formulation.

The efficacy of oxygen inhalation at the flow of 7 l/min for 15 min has been shown in dated open studies and in a more recent controlled crossover study versus room air [259]. In a further randomized study using an oxygen mask and oxygen flow of 12 l/min a complete remission of cluster headache attacks both in the episodic and chronic forms was obtained within 15 min in 78 % of cases versus 20 % for room air [260]. In the case of no response to usual recommended flow it may be increased to 14–15 l/min [261].

Dated studies have shown the efficacy of ergotamine tartrate 250 μg i.m. (only 1 study vs. placebo), ergotamine tartrate (1 mg) plus caffeine (100 or 200 mg) tablets (in a study also in association with bellafoline 100 mg) and ergotamine tartrate (2 mg) plus caffeine (100 mg) and dihydroergotamine nasal spray for the acute treatment both in patients with episodic and chronic forms [262‐266].

Contrasting results have been obtained for lidocaine intranasal (4 %) in open studies [101, 267, 268]. Better results were achieved in a study using 10 % lidocaine solution with respect to saline [269].

The application of a solution of cocaine 10 % in both nostrils has been shown to block the attack in a controlled study versus placebo (saline solution) in patients with episodic and chronic cluster headache [269, 270]. No significant adverse events were recorded with the exception of a mild state of arousal in a patient who had abused the drug. Cocaine is not available in Italy as medication.

Two randomized controlled trials are available, one for somatostatin i.v. (25 μg in 50 ml saline) and one for octreotide i.v. (an analogue with longer half-life: 1.5 h, 100 μg in 1 ml vehicle) demonstrating a significant reduction of pain in 20 and 30 min. respectively [271, 272]. Most common side effects are nausea, diarrhoea and meteorism.

In a double-blind controlled study versus placebo this drug at the dosage of 120 mg × 3 per day has demonstrated to be effective in patients with episodic CH. In these patients it is indicated as first-choice drug [274]. In the chronic form, according to two open studies and one head-to-head study versus lithium carbonate, it was effective in 50–55 % of the patients [275‐277]. Dosages used were higher than those used for the episodic form (up to 960–1,200 mg). In the comparison study verapamil was effective more rapidly with fewer side effects [276]. The most relevant adverse events include arrhythmia (19 % of cases) and bradycardia (36 %); ECG monitoring of patients is recommended to avoid an atrioventricular block and symptomatic bradycardia [274, 278, 279]. In Italy, verapamil is not indicated for cluster headache.

Results are available from some dated open studies and two RCTs demonstrating the effectiveness of this drug in the preventive treatment of chronic CH. In the two RCTs, dosages were 300 mg × 3 per day and 800 mg/day, respectively [277, 280, 281]. Blood lithium levels should range from 0.4 to 0.6 mmol/l. The most frequent adverse events include tremor, gastrointestinal disturbances, dizziness and polyuria [281, 282].

In a retrospective study, prednisone (10–80 mg/day) induced a significant reduction (72 % of cases) or a complete remission (58 % of cases) of attacks within 3–10 days in a small sample of CH patients with episodic or chronic form, with the best results obtained for dosages ≥40 mg [283]. When dosage was gradually reduced, (<20 mg) the attacks reappeared. The drug is recommended for short-term usage (i.e., at the start of other recommended preventive treatments).

Methylprednisolone i.v. (250 mg in 100 ml saline) followed by prednisone per os (10 mg/day) induced a further benefit in patients treated with optimal doses of verapamil [284]. The i.v. administration of methylprednisolone (30 mg/kg in 500 ml of saline in 3 h) induced the interruption of the cluster period for 2–3 days, with a subsequent prompt reappearance of attacks [285]. No significant side effects were observed for both drugs at the used dosages.

Methysergide has been tested only in dated open studies which demonstrated an efficacy in 76–77 % of cases [286, 287]. The recommended dose is 8 mg (although the dosage of 16 mg has also been tested). The drug should be started with the dose of 2 mg and should be increased gradually every 3–7 days. Side effects occur in 20–45 % of patients. The most frequent include nausea, dizziness, stomach pain, restlessness, somnolence and cramps. The drug should not be used for more than 6 months due to the possible development of retroperitoneal and lung fibrosis [288]. It is not available in Italy. Pizotifen (1–4 mg/day), in a dated double-blind study, showed a certain efficacy in a limited number of patients with episodic CH (reduction >50 % of attack frequency in 36 % of the cases and interruption of cluster period in 21 %) [289]. Lisuride has been administered at variable dosages from 0.075 to 0.400 mg/day to patients with episodic and chronic cluster headache in an open study. In all patients the drug induced a benefit without relevant side effects [290]. Lisuride is available in Italy at the dosages of 0.2–0.5 mg and does not have the indication for CH.

A positive effect of sodium valproate (1–2 g/day) had been suggested by an open study but was not confirmed by a more recent RCT [291, 292]. In the latter study, drug/placebo was administered for 2 weeks and 50 % of responders were identified in the valproate group versus 62 % of placebo group [292]. Two open studies showed an effect of topiramate at different dosages (50 and 125 mg/day and 25–200 mg/day, respectively) with a remission within about 3 weeks from the beginning and a reduction of duration of the cluster period; a RCT (50–250 mg/day) did not show a superiority of topiramate compared with placebo [293‐295]. After obtaining promising results in a single case, gabapentin has been tested in three open trials involving a limited number of patients at doses ranging from 800 to 3,600 mg/day. The drug, administered for 4–6 months, reduced significantly the frequency and intensity of attacks or interrupted the cluster period in at least 50 % of cases and was well tolerated [296‐299]. Most of the antiepileptics listed above do not have the indication for CH in Italy.

Ergotamine derivatives i.m. ergotamine has been tested in an open study involving a limited number of patients during the active phase at dosages ranging from 0.25 to 0.50 mg/day to 0.25 mg four times per day. It induced a complete disappearance of attacks in all the patients without a reduction of cluster period. Side effects were somnolence, anorexia, bad taste and daze [300]. Dihydroergotamine i.v. was evaluated in two retrospective studies involving hospitalized patients. In the first study dihydroergotamine at the dosage of 0.5 mg three times per day interrupted the cluster period in patients with both episodic and chronic forms, refractory to other preventive treatments. In the second study dihydroergotamine i.v. (0.5 + 1 mg) and dihydroergotamine nasal spray 1 mg or dihydroergotamine s.c. at the dosage of 0.5–1 mg induced the disappearance of attacks or a reduction >50 % of attacks in 88 % of patients with episodic and 57 % of patients with chronic CH [301, 302]. Side effects were mild and only a few patients had to discontinue the drug.

After the first observations of Horton (838), one dated open study involving few patients demonstrated that histamine sulphate i.v. (2.75 mg in 250 ml the first day, followed by 11 mg in 500 ml saline in the subsequent 9 days) induced a reduction of 75–100 % in one-third of cases, of 10–49 % in an another one-third of cases and no effect in the remaining cases [303].

They were proposed for the short-term prophylaxis at the beginning of the cluster period instead of steroids (defined also as transitional therapy) or for the short-term prophylaxis of mini cluster. A reduction of the attacks was observed for frovatriptan (5 mg/day) in studies involving episodic CH patients treated with verapamil (retrospective study) and chronic CH patients resistant to preventive treatments (open study) [304]. In an open study eletriptan (40 mg two times per day for 6 days) induced 50 % reduction of attacks in one-third of the cases [305]. Responders were all treated with verapamil, whereas non-responders did not use preventive drugs. In a randomized double-blind versus placebo study sumatriptan 100 mg administered three times per day for 7 days did not induce a significant reduction of attack frequency [306].

A first study demonstrated that the bilateral application of 300 μg per nostril, repeated to reach a complete desensitization, induced a significant reduction and disappearance of the attacks, while a second study demonstrated that the application in the ipsilateral nostril is equally efficacious, whereas the application in the contralateral nostril was ineffective [307, 308]. This effect was observed in the majority of patients with both episodic and chronic CH after 10 days of treatment. In the chronic patients after 25–40 days from the last application of capsaicin the attacks reappeared. Capsaicin is not available in Italy.

Its efficacy at the dosage of 10 mg per os has been tested in a limited number of patients with episodic form demonstrating the ability of the drug to reduce significantly the attack frequency. In responders this effect was obtained in 3–5 days [309, 310]. These promising results were not confirmed in a further pilot study involving patients with episodic CH [311].

Levels of recommendation for symptomatic and preventive drugs are reported in Table 13.