Abstract
Two patients suffering from ophthalmoplegic migraine had a strictly unilateral headache absolutely responsive to indomethacin, but not to other non-steroidal anti-inflammatory drugs, analgesics or corticosteroids. Such observations raise a therapeutic alternative and suggest that ophthalmoplegic migraine may present with different headache phenotypes.
Ophthalmoplegic migraine (OM) is an enigmatic condition typically characterized by recurrent attacks of headache with features of migraine accompanied or followed by evidence of paresis of one or more of the ocular cranial nerves (commonly the third nerve) that usually recovers over weeks (1). Permanent damage to the involved nerve may occur in some cases of recurrent episodes, suggesting the possibility of an organic process that potentially involves lesioning of the oculomotor nerves (2,3).
OM had previously been considered a migrainous process often involving the oculomotor nerves. However, the International Headache Society reclassified it as a form of cranial neuralgia rather than migraine in 2004 (1). Recent evidence suggests that OM is an acute neuropathy, possibly demyelinating or inflammatory in nature (4). This theory has been substantiated by several magnetic resonance imaging (MRI) studies that have documented both contrast enhancement and thickening of the symptomatic oculomotor nerve (5,6).
Neither the aetiology of OM nor an effective treatment strategy has convincingly been demonstrated. Therapeutic efforts have mostly been targeted at preventing recurrences. However, information regarding therapy in the acute stage, and specifically for the head pain itself, is scarce.
Here, we present the cases of two patients suffering from OM whose headaches responded absolutely to indomethacin but not to treatment with other non-steroidal anti-inflammatory drugs (NSAIDs), analgesics, triptans or corticosteroids. This particular diagnostic and therapeutic combination is an important finding, since the response to indomethacin may link OM to other types of headache that are absolutely responsive to indomethacin.
Case reports
Case 1
A 36-year-old woman, a nurse, had no personal or family remarkable antecedents. She never smoked. Four years ago she had given up contraceptive pills. At age 35 years she began suffering from a left-sided, frontal-orbital, continuous headache that started progressively. One year later, a similar episode occurred. The headache was oppressive in character and severe in intensity, with a continuous, fluctuating, daily pattern. The symptoms neither woke the patient from sleep nor were modified by Valsalva manoeuvres or body position. No precipitating mechanisms were noticed. She denied nausea, vomiting, photo- or phonophobia, feeling of foreign body sensation in the eye of the symptomatic side, conjunctival injection, lacrimation, and rhinorrhoea/nasal congestion. In both episodes, between 12 and 24 h from pain onset, the patient noted a mild left-sided ptosis and horizontal diplopia. Initially she was under the care of her family doctor, who consecutively prescribed paracetamol (1 g every 8 h), naproxen (525 mg every 8 h) and ergotamine plus caffeine (1 mg and 100 mg, respectively, every 8 h) without benefit.
The patient came to the out-patient consultation of our neurological department 7 days after the onset of the first episode's symptoms. Neurological examination revealed paresis of the left third cranial nerve (CN III). No changes in pupil size were noticed. No sensory loss was found in the territory of the trigeminal nerves. Analyses of blood, including blood cell count, erythrocyte sedimentation rate (ESR), biochemical parameters and electrolytes, antinuclear antibodies (ANAs), extractable nuclear antigens (ENAs), anti-neutrophil cytoplasmic antibodies (ANCA), polymerase chain reaction (PCR), rheumatoid factor, immunoglobulin count and electrophoresis were all normal. A spinal tap released crystalline cerebrospinal fluid (CSF) with an opening pressure of 14 cm of water. Biochemical, microbiological and cytological analyses of the CSF were normal. A contrast-enhanced MRI scan of the brain showed enhancement of the cisternal portion of the left CN III.
The patient was treated on this first episode with 6 mg subcutaneous sumatriptan, and later with prednisone (60 mg daily, during three consecutive days) without any benefit. Subsequent treatment with 75 mg of indomethacin provided complete relief from the pain within the first day of treatment. The oculomotor nerve paresis subsided in 2 weeks, recovery being complete, and treatment with indomethacin was discontinued. The headache of the second episode again showed an absolute response to indomethacin, which in this case was taken early in the course of symptoms.
Case 2
A 30-year-old woman, a housewife, had neither a personal nor family medical history of any note. In particular, she denied toxic habits or any current treatment, including hormonal contraception. At the age of 28 years, she experienced an acute, right-sided, hemicranial headache that later predominantly involved the right orbit and forehead. The headache was rather intense and continuous, oppressive and pulsatile in character, and was accompanied by nausea without vomiting. Neither conjunctival injection, lacrimation, rhinorrhoea nor nasal stuffiness was noted. The patient also denied photophobia and feeling of foreign body sensation in the eye of the symptomatic side. There were no precipitating mechanisms. No substantial changes were observed upon exercise, Valsalva manoeuvres or body position. The headache persisted with a daily pattern but did not disturb sleep. Forty-eight hours after the onset of pain, she noticed a progressive horizontal diplopia and right-sided ptosis that reached a maximum within a week and was then associated with slight mydriasis of the eye on the symptomatic side. One day after the onset of headache she consulted her family doctor, who prescribed treatment with paracetamol (1 g every 8 h) and ibuprofen (600 mg every 8 h), which were of no avail. Two days later she was referred to the emergency department of our hospital, where intravenous (i.v.) treatment with both dexketoprofen (50 mg every 8 h) and metamizol (2 g every 8 h) were of no avail. Sumatriptan (6 mg) was administered subcutaneously and also failed to provide a substantial benefit. From the Emergency Department she was admitted to our neurological ward. Neurological examination demonstrated paralysis of the right CN III with mild parasympathetic involvement. No sensory loss was found in the territory of the trigeminal nerves. Analyses of blood, including blood cell count, ESR, biochemical and electrolytes, ANA, ENA, ANCA, PCR, rheumatoid factor, immunoglobulin count and serum electrophoresis were all normal. A spinal tap released crystalline CSF with an opening pressure of 12 cm of water. Biochemical, microbiological and cytological analyses of the CSF were normal. An MRI scan of the brain and orbits demonstrated a thickened right oculomotor nerve that was contrast-enhanced in its cisternal portion.
At the onset of i.v. treatment with 90 mg/day of prednisone, the patient experienced significant worsening of the pain and the treatment was discontinued. She was then treated with 150 mg/day of indomethacin, which resulted in rapid (within 24 h) and complete improvement of the pain. Over the next 2 weeks, the oculomotor palsy recovered. Indomethacin treatment was maintained for 2 weeks at a dose of 150 mg/day, and another week at a dose of 75 mg daily.
Two months later, the patient developed another episode similar to the previous one and involving the same side, which again proved to be absolutely responsive to indomethacin.
Discussion
According to the prevailing diagnostic criteria (1), our patients suffered from OM given the presence of a documented lesion of an oculomotor nerve that seemed to produce a strictly unilateral ipsilateral headache. However, the headaches were absolutely responsive to indomethacin, whereas treatment with other NSAIDs, triptans and corticosteroids failed to provide any improvement. In addition to such a differential pharmacological response, the clinical features of the headaches were at variance with the acknowledged concept of migraine. The nature of the accompanying headache, its response to indomethacin, and the possible relationship with an acute oculomotor palsy all warrant further comment.
Admittedly, noxious stimulation of an exclusively motor nerve, such as an oculomotor nerve, can produce pain, but with a different distribution and quality than a typical neuropathic pain. Motor nerve pain is a projected-type of pain and is perceived as a deep and boring pain localized within the muscles innervated by such a nerve. Activation of nociceptors in the oculomotor nerve probably accounts for the pain. However, the fact that OM may present with migrainous features suggests a possible triggering of the migraine pathophysiological process, either by co-stimulation of adjacent trigeminal fibres or indirectly by the release of neurogenic inflammatory mediators from the involved oculomotor nerve.
Nerve trunk-associated pain has been usually ascribed to increased activity in sensitized nociceptors within the nervi nervorum or vasa nervorum (7,8). In such cases, the resulting pain is nociceptive type and the common mechanisms might be compression, ischaemia, or inflammation (9). Neurological symptoms and signs appear when the nerve injury produces demyelination or involves the axons; such circumstances, depending on the aetiology, tend to occur with some delay after the beginning of the heralding of nociceptive pain (9). Such a sequence of events may explain why episodes of OM usually start with head pain and why some days later the ophthalmoparesis ensues. Permanent nerve damage may occur when the process is either intense enough or recurring with cumulative damage up to the severance of axons. The oculomotor nerve damage/dysfunction may explain the topographically consistent pain (i.e. unilateral headache), but not how the headache may have migrainous features.
Noxious stimulation of the nerve endings of the nervi nervorum can lead to the increased synthesis and release of algogenic substances, resulting in neurogenic inflammation (10,11). Otherwise, sensory fibres from the first (ophthalmic) division of the trigeminal nerve (i.e. the V1 nerve) pass through the oculomotor nerve on their way to the spinal trigeminal nucleus (12), thereby rendering them susceptible to a given pathogenic process. Indeed, the V1 nerve receives anastomoses from the oculomotor nerves, so the V1 nerve might represent the sensitive root of the oculomotor muscles (13). Admittedly, activation of these V1 afferents may trigger the trigeminovascular system, investing the associated headache with migrainous features. OM has even been considered to be a migrainous process of a vascular nature, possibly an acute ischaemic neuropathy due to migrainous vasospasm. This is a reasonable hypothesis, as ischaemic nerves certainly produce pain. However, the vascular hypothesis has not gained wide acceptance.
Since OM is no longer classified as a migraine variant but rather now as a cranial neuralgia with ipsilateral head pain (1), the associated headache could theoretically be at variance with migraine and responsive to different treatments. This may explain why reports of abortive treatment of the headache with aspirin, ergotamine derivatives, triptans and corticosteroids describe unconvincing or contradictory results. Accordingly, it would have been more accurate to name the syndrome ‘ophthalmoplegic hemicrania’, which may encompass headaches with features of migraine as well as other phenotypes (e.g. indomethacin-responsive headache). In fact, headaches associated with intracranial lesions/disturbances may exhibit a clinical phenotype reminiscent of any of the primary headaches. Indeed, intracranial lesions with similar topography may produce a varied assortment of headache phenotypes.
The pain produced by a possible acute inflammatory neuropathy might respond to treatment with NSAIDs. Yet interestingly, our patients rapidly and absolutely responded to indomethacin, whereas treatment with other NSAIDs, analgesics, triptans and corticosteroids was of no benefit. With indomethacin treatment, the pain was antagonized in an antidote-like fashion; however, restoration of nerve function occurred with a considerable latency, probably because healing of the nerve is a time-consuming process. Although indomethacin treatment convincingly blocked the production of pain, it is unclear whether indomethacin could have influenced the course of the neuropathy itself. Nevertheless, it is likely that both the ophthalmoplegia and the pain were dependent on the same pathogenic process that was counteracted by the indomethacin, with the recovery of all of the signs and symptoms being asynchronous.
The superb response to indomethacin in our patients with OM is reminiscent of other types of headaches characterized by an absolute response to indomethacin, such as paroxysmal hemicrania (PH) and hemicrania continua (HC). In fact, PH, HC and OM also have in common a female preponderance, and a strict unilaterality of symptoms. Theoretically, it could well be that at least a subset of OM is indomethacin sensitive. If our observations were to be confirmed by further reports, great importance would have to be attached to cases of OM with an absolute response to indomethacin. Such a phenomenon would inherently bear important therapeutic, diagnostic and nosological implications.