Oxygen treatment is effective in migraine with autonomic symptoms

Abstract

Introduction

While inhalation of high-flow 100% oxygen is highly effective in cluster headache, studies on its efficacy in migraine are sparse and controversial.

Case

We report the case of a 22-year-old patient with an eight-year history of strictly unilateral migraine without aura but cranial autonomic symptoms. She repeatedly responded completely to inhalation of high-flow pure oxygen within 15 min but suffered from recurrence of attacks within 30 min after discontinuation.

Discussion

In line with experimental animal studies, this case suggests a clinically relevant efficacy of inhaled oxygen in patients with migraine with accompanying cranial autonomic symptoms.

Introduction

Inhalation of normobaric oxygen is well known to be highly effective as an abortive treatment in attacks of cluster headache in several trials, with response rates of up to 80% (1). Although assumed to be ineffective in other primary headache syndromes, surprisingly little is known on its acute effects in migraine. Alvarez and Mason were the first to report efficacy of pure oxygen inhalation at a flow rate of 6–8 l/min in patients with “typical migraine”, reaching pain freedom in 42% and substantial relief in 36% (2). However, in a Cochrane Review from 2008, no placebo-controlled study on normobaric oxygen was found and thus no recommendation was given (3). Three cited small non-placebo controlled studies found no effect of oxygen in acute migraine attacks. A recent randomized and placebo-controlled study on the use of oxygen in acute migraine attacks (n = 57) found oxygen (100%, delivered at a flow-rate of 15 l/min via a non-rebreathing mask) to reduce pain intensity on a visual analog scale (VAS) significantly stronger during a migraine attack than placebo treatment with room air (4). Taken together, little is known on the efficacy of high-flow oxygen in the treatment of acute migraine attacks, and results are ambiguous.

Case report

A 22-year-old female presented with an eight-year history of migraine without aura. She complained of recurrent unilateral and mostly left-sided headache attacks with a constant penetrating fronto-temporal pain and moderate to severe intensity with a frequency of four to five attacks/month lasting for one to three days. These attacks were accompanied by nausea, photo- and phonophobia as well as exacerbation on exertion with transition to a throbbing headache. Additionally, she suffered from ipsilateral lacrimation, rhinorrhea and ptosis during each attack. Symptoms suggestive of migraine aura or an exacerbation during menstruation were denied. There was no family history of headaches. Weather changes, stress and an irregular sleep pattern triggered attacks.

Due to the accompanying autonomic symptoms, the erroneous diagnosis of a cluster headache had been made before referral to our unit and oxygen was prescribed. She reported to have treated at least 10 acute migraine attacks at home and noticed complete pain freedom after inhalation of 100% oxygen with a non-rebreathing mask at a flow rate of 14 l/min within 15 min during each attack. She further reported that a rebound to the original level of pain was observed 30 min after discontinuation in each attack. While non-opioid analgesics (acetylic salicylic acid, paracetamol and ibuprofen) were not effective, oral sumatriptan 100 mg aborted attacks if given early in an attack. As the neurological and psychiatric examination was unremarkable as was the long-lasting history typical of migraine, a cerebral magnetic resonance imaging (MRI) scan was not deemed necessary and the diagnosis of migraine without aura but with autonomic symptoms was made.

Discussion

Migraine may well present with additional autonomic symptoms (5–7), and the misdiagnosis of cluster headache is sometimes made. As presented in this specific case, migraine with cranial autonomic symptoms (CAS) can respond impressively to oxygen. Although no unambiguous data on the abortive efficacy of oxygen inhalation in migraine attacks exists, the commonly accepted (hence not sufficiently tested) theory is that the response to oxygen inhalation is exclusively positive in cluster headache. Given the clinically important note that oxygen inhalation needs high gas flow of more than 10 l/min and a non-rebreathing mask before assuming inefficacy of inhaled oxygen even in cluster headache (8), early clinical observations leading to the assumption that it is not effective in migraine will never have tested this. But even if we assume that the biology of cluster headache and migraine is so distinct that oxygen inhalation is exclusively beneficial in acute cluster headache attacks, the question arises: What is the common link between our migraine case and cluster headache in general that might explain the effect of oxygen inhalation in the above case? The obvious answer is the presence of autonomic activation during an individual attack.

Interestingly, Kaup and colleagues reported a patient with strictly unilateral migraine with aura who suffered from recurrent exacerbations of three to four hours with unilateral CAS during his migraine attacks who could abort these attacks completely with oxygen inhalation (9). Despite some efficacy in migraine patients, the Turkish study found complete response to oxygen in cluster headache patients only (4). For other uncontrolled studies with poor outcomes, no data on autonomic symptoms are given. Based on these observations, a relevant activation of the cranial parasympathetic system, especially the trigemino-parasympathetic reflex, could be a crucial factor or even a prerequisite to benefit from oxygen inhalation. Akerman and colleagues effectively reduced activation of trigeminal afferents in the trigemino-cervical complex and activation of the parasympathetic system in rats after stimulation of the parasympathetic superior salivatory nucleus in the brainstem by inhalation of oxygen but not air. Activation changes due to trigeminal stimulation were not modulated by oxygen (10). These experimental findings would strongly suggest that oxygen exerts its effects on the parasympathetic branch of the trigemino-parasympathetic reflex arc and would support the notion that efficacy of oxygen in migraine patients would require significant parasympathetic activation. A strong argument against this (testable) hypothesis is the fact that oxygen inhalation is not able to abort headache syndromes with autonomic symptoms, such as short-lasting, unilateral, neuralgiform headache attacks with conjunctival injection and tearing (SUNCT) syndrome or paroxysmal hemicranias, which may point to additional pathophysiological factors beyond autonomic activation. Alternatively, the duration of attacks in SUNCT syndrome and paroxysmal hemicrania could be just too short to inhale oxygen sufficiently long enough to achieve therapeutic efficacy. In patients with migraine it is conceivable that the different length of attacks requires prolonged or repetitive inhalation of oxygen to abort attacks (compared to the relatively short-lasting inhalation necessary in cluster headache) or that response to oxygen is differential because of the divergent pathophysiology.

As for migraine features in cluster headache such as photo- and phonophobia, the presence of CAS in migraine patients seems to make diagnosis more difficult despite the high incidence of CAS in these patients. Up to 56% of migraine patients have CAS during attacks (7). Lacrimation occurred in 44% and rhinorrhea in 22% of the patients. In most studies, both headache and CAS were strictly ipsi- and unilateral (6). It is not clear whether migraine attacks with autonomic symptoms may represent a distinct entity in the large spectrum of migraine phenotypes and show, for example, facilitation of the trigemino-autonomic reflex or if autonomic activation occurs as an unspecific mechanism owing to nociceptive input reaching a relevant threshold.

Conclusion

Our case suggests clinically relevant efficacy of inhaled oxygen in patients with migraine and autonomic symptoms; oxygen may be more effective in migraine with CAS than in migraine without CAS based on observations in animal experiments; this theory needs to be tested; and the presence of CAS ipsilateral to the side of pain is not specific for cluster headache and should not delay the diagnosis. The duration of attacks is certainly more specific for the correct classification as migraine (4–72 hours) or cluster headache (15–180 min).