Treatment of acute cluster headache with 100 % oxygen was first suggested in the early 1950s [8]; interestingly it is not clear what the basis for the suggestion was [9]. Janks [10] set the use out very clearly as a sufferer himself, contributing the need for a rate of 10 L/min. Kudrow [11] reported the first series using 100 % oxygen at 7 L/min noting three-quarters of patients responded. He also reported oxygen to be better than sublingual ergotamine.

The scientific study of oxygen in cluster headache began with Fogan [12] who reported 19 patients in whom he compared air and oxygen, at 6 L/min, in a crossover study. Eleven patients completed the crossover. Oxygen was more effective than air.

The only substantial, well-powered study of oxygen in cluster headache used a double-blind, placebo (air)-controlled crossover design to treat four attacks in a balanced allocation. Oxygen 100 % at 12 L/min for 15 min rendered 78 % of patients pain free at 15 min while air rendered 20 % of patients pain free; the difference was highly significant with 78 subjects treating 298 attacks [13]. Oxygen reduced associated features and was extremely well tolerated. A proportion of patients have a rebound effect after the oxygen is stopped with the attacks returning; there is no large database to estimate this with although experience suggests it happens in up to one-third of patients. It can be responsible for a perception that attack frequency has been increased [14].

Hyperbaric oxygen offers no advantages [15–17] and very considerable logistical challenges such that it cannot be recommended.

Triptans were primarily developed by Humphrey and colleagues [18] for the acute treatment of migraine. The broad pharmacology of the triptans is similar [19], while there are some differences in pharmacokinetics and thus far only sumatriptan is available as an injection.

Parenteral dihydroergotamine (DHE) has been considered to be an effective abortive agent for cluster headaches for some time [8, 29]. There are no controlled trials of injectable DHE; however, clinical experience has demonstrated that intravenous administration can be helpful even when attacks have been resistant to other treatments [30]. DHE nasal spray 1 mg in a double-blind, placebo-controlled, crossover trial in 25 patients demonstrated a significant reduction in pain intensity in acute cluster headache [31]. The development of new, inhaled formulation thus far studied in migraine may be very useful and should certainly be tested in the cluster headache [32].

Lidocaine (10 %) was reported to be effective at aborting nitroglycerin-induced cluster attacks in a double-blind, placebo-controlled study in nine patients [33]. Lidocaine or saline was applied for 5 min using a cotton swab in the area corresponding to the pterygopalatine fossa, under anterior rhinoscopy. The onset of effect was slow and the method not applicable to most patients. Based on this evidence, and our experience of a modest adjunctive effect, lidocaine solution 20–60 mg, given as nasal drops (4–6 % lidocaine solution) is prescribed, and applied bilaterally in the region of the pterygopalatine fossa. To ensure that the solution reaches the pterygopalatine foramen, the patient should be instructed to lie down horizontally as early as possible during an attack, with the head extending out of the bed, bent downwards 30–45° and rotated 20–30° towards the side of the headache. The tip of the dropper is inserted above the rostral end of the inferior turbinate and pushed inwards as deep as possible before dripping. The patient should be asked to maintain the position for about 2–5 min. It is certainly a cumbersome approach so that few patients find it useful in the medium term.