Abstract
Introduction
The International Classification of Headache Disorders classifies ophthalmoplegic migraine (OM) under “cranial neuralgias and central causes of facial pain.”
OM is diagnosed when all the following criteria are satisfied:
At least two attacks fulfilling criterion B. Migraine-like headache accompanied or followed within four days of its onset by paresis of one or more of the III, IV and/or VI cranial nerves. Parasellar orbital fissure and posterior fossa lesions ruled out by appropriate investigations.
In children the syndrome is rare and magnetic resonance (MR) shows strongly enhancing thickened nerve at the root entry zone (REZ).
Method
The authors review the literature focusing on pathogenesis theories.
Results
The authors suggest that ischemic reversible breakdown of the blood-nerve barrier is the most probable cause of OM and to include MR findings in the hallmarks of the disease.
Conclusion
OM is the same disease in adulthood and childhood, even if in adults the MR imaging findings are negative. In the authors’ opinion, OM should be classified as migraine.
Introduction
Ophthalmoplegic migraine (OM) is a rare episodic childhood condition characterized by recurrent episodes of headache with unilateral ophthalmoplegia due to paresis or paralysis of the III, IV or VI cranial nerves (1–4). Most often the disease is reversible and is thought to be a self-limited condition (3). OM is recognized as a cranial neuralgia and not as migraine according to the most recent edition of the International Classification of Headache Disorders (ICHD) (5). In ICHD-I, OM was grouped in the migraine variant (6). In 2001, the second edition of ICHD (ICHD-II) classified OM in cranial neuralgias on the basis of magnetic resonance imaging (MRI) and clinical features (7). In the recently published ICHD third edition beta (ICHD-III-Beta), OM is classified among painful cranial neuropathies and more specifically as recurrent painful ophthalmoplegic neuropathy (8).
The onset of the disease typically occurs in childhood and ocular motor nerve palsy usually recovers completely. Paralysis has been reported to recover in four to 84 days (9), even if in some cases, after multiple episodes, the deficit may be permanent with aberrant regeneration of the third nerve (3,10–12). Contrast enhancement MRI shows characteristic reversible focal thickening and enhancement of the cisternal tract of the involved nerve, usually the III cranial nerve, at the root exit zone (REZ) (6,13,14).
Even if in adults the clinical presentation is the same, the MR findings are negative (15).
Many pathogenetic theories have been suggested, but none of them seem to be conclusive (3,15–18). We briefly discuss the previously reported pathogenesis, and emphasize the theory involving a reversible ischemic breakdown of the blood-nerve barrier.
Finally, we discuss whether OM in children and in adults is the same disease and whether it should be classified as migraine or cranial neuralgia.
Discussion
The ICHD classifies OM under “cranial neuralgias and central causes of facial pain,” and OM is diagnosed when all the following criteria are satisfied:
At least two attacks fulfilling criterion B. Migraine-like headache accompanied or followed within its onset by paresis of one or more of the III, IV and/or VI cranial nerves. Parasellar orbital fissure and posterior fossa lesions ruled out by appropriate investigations.
When the criteria are satisfied, the diagnosis is definite OM in patients who have suffered two or more attacks, and probable OM in patients with only a single attack. Laboratory data including erythrocyte sedimentation rate, rheumatoid factor, lupus erythematosus (LE), antinuclear antibody, serum venereal disease research laboratory (VDRL) and cerebrospinal fluid (CSF) pressure and analysis are all normal.
OM in children is a rare condition, characterized by recurrent headaches followed by ocular motor nerve palsy (1–3), with an estimated annual incidence of 0.7 per million (19). The disease most often occurs in infancy or childhood, with average age at the onset under 10 years (3,20). It has rarely been reported, but in such cases patients frequently started their clinical history in infancy (9,15,21,22). Children with recurrent oculomotor paresis without headache have also been reported and diagnosed as having “ophthalmoplegic migraine variant” (9,23–27). Usually the III cranial nerve is affected, but involvement of the IV (28) or VI cranial nerves has rarely been reported (29–32); a case of recurrent OM with simultaneous oculomotor and trochlear nerves palsy with reversible thickening and enhancement of the trochlear nerve has also been reported (28). In OM the clinical type of migraine varies from case to case, and may cover the whole migraine spectrum (32). Computed tomography (CT) is usually reported as negative according to our knowledge; we found very few cases of positive CT (13,21,33). MRI performed during or immediately after the migrainous-like headache attack usually detects enlargement and homogeneous enhancement of the III cranial nerve at the REZ even though few cases of OM in children with normal MR have been reported in the literature (33–35). Enlargement is usually round or trapezoid with the widest area adjacent to the midbrain, but in some rare cases the enlargement and the enhancement of the nerve is linear and spreads along the nerve (9,28). Interestingly, in the patient with OM and right abducens palsy, reported by Lee et al. (30), MRI showed only subtle, linear, oblique reversible enhancement within the right lower pons in an area neuroanatomically corresponding to the intra-axial roots of the nerve.
Repeat MRI in the quiescent period shows reduction or complete resolution of the enhancement, whereas the thickening of the nerve may last for weeks, months, years, or it may even be permanent (3). In 1997 Wong and Wong (16) suggested including these MRI findings associated with reversible ophthalmoplegia as supportive diagnostic criteria in OM. As we previously emphasized (13), the triad of migraine, III cranial nerve palsy, and enhancement of an enlarged III cranial nerve at the REZ during the acute phase, and subtle or no enhancement during the quiescent phase, should be considered as pathognomonic of probable OM, after the first attack, and of definite OM after the second or more attacks; further investigation should be avoided. Because of the self-limiting nature of the disease, there are few post-mortem studies of patients fulfilling the general semiology of OM (36,37); none of them has been conclusive and the pathogenesis of this syndrome remains questionable (36,37). Enhancement of the cisternal portion of the oculomotor nerve can occur in a variety of neoplastic and infiltrative conditions (3), benign such as schwannomas, and hemangiomas, or malignant such as lymphoma, leukemia, and carcinomatous meningitis (38), but none of these conditions could resolve spontaneously.
Enhancement of the cisternal portion of the oculomotor nerve also occurs in a variety of infectious and noninfectious inflammatory conditions, such as neuroborelliosis (28,29,39,40), syphilis (41,42) coccidioidomicosis, and human immunodeficiency virus (HIV) infection (17). In these cases as well, it is highly unlikely that these diseases resolved spontaneously without specific treatment.
An arteriovenous malformation (AVM) may have a clinical picture mimicking OM (43), but a true AVM unlikely regresses spontaneously and MRI shows evidence of flow (3). In a reported case of oculomotor nerve schwannoma at the REZ mimicking OM (44), the paper has too many speculative statements and does not seem to be convincing (45). Recently Akimoto et al. (46) presented a case that really did mimic OM clinically and radiologically. The pathological specimen showed neuromuscular hamartoma involving the third (III) cranial nerve at the REZ (46). This case is of particular importance because it shows that an intracranial lesion could have typical clinical and radiological findings like OM. Tolosa-Hunt syndrome, which is a condition characterized by retro-orbital pain and a variable degree of ophthalmoplegia, may also clinically mimic OM. The characteristic clinical presentation of Tolosa-Hunt is a cavernous sinus or an orbital apex syndrome, and neurological involvement is usually more extensive than in OM, including involvement of the optic, oculomotor, trochear, ophthalmic division of the trigeminal, abducens, and ocular sympathetic nerves. Isolated third cranial nerve palsy has occasionally been seen (17). In this syndrome, MRI often shows abnormal soft tissue enhancing in the ipsilateral cavernous sinus, usually resolving after steroid treatment. Myasthenia is not associated with headache and is easily ruled out by the Tensilon test.
Different mechanisms for the etiology of OM have been suggested.
Previous pathogenetic theories of OM include oculomotor nerve compression due to swelling of the posterior cerebral artery (PCA), pituitary swelling, and vascular anomaly with compression of the III cranial nerve, but none of these theories has been conclusive (18,22,24). Ischemia with microvascular infarct of the III cranial nerve, such as in diabetic oculomotor nerve palsy, has been suggested as a cause of OM (47), but, even though it is often associated with severe unilateral headache, it would neither explain the enlargement of the thickened nerve nor the complete recovery of the patient.
In 1998 Mark et al. (3) suggested a benign viral infection as being the cause of OM because in the authors’ opinion it can explain imaging, clinical findings, and spontaneous recovery over two or three weeks. Moreover, they found a striking similarity with the enhancement of the VII cranial nerve shown by MRI in patients with Bell’s palsy, which is considered a form of benign viral neuritis, that may recur, and sometimes recovers spontaneously (47), although in our opinion, this hypothesis does not seem to be convincing. The association between migraine and VII cranial nerve palsy is not reported in Bell’s palsy; moreover in idiopathic facial palsy, MRI, when performed, shows that the most frequently enhancing segments of the nerve are the geniculate ganglion and the canalicular segment, and not the REZ, as occurs in OM. Finally, the enhancing VII nerve does not enlarge during the episode (3).
It has recently been suggested that OM is better characterized as a recurrent demyelinating neuropathy, in which the enlargement and the enhancement of the oculomotor nerve during symptomatic periods is reminiscent of the swelling and the enhancement detected in the nerve roots and plexi reported in patients affected by chronic inflammatory demyelinating polyradiculo-neuropathy (CIDP) (20,48). The pathological substrate of CIDP is thought to be recurrent demyelination and remyelination with nodular hypertrophy of nerves and plexi (47). This pathogenetic theory is interesting enough, but there are some main differences between CIDP and OM: In CIDP the enhancement is not focal (49), it also lasts into the remitting period, and the enlargement of the nerve increases at every relapse of the disease.
Lance and Zagami have suggested that the migrainous pain in patients with OM is due to irritation of the trigeminal sensory fibers because of the post-viral inflammatory process affecting the oculomotor nerve. This activates the trigemino-vascular system in patients who are migraneurs, thereby triggering the associated migraine headache (20). However, the universal absence of any antecedent or concomitant viral prodrome and a completely normal CSF examination strongly argues against this theory in patients with OM. The activation of the trigemino-vascular system leads to neurogenic inflammation releasing neuropeptides into the vessel wall of vasa nervorum, which in turn leads to the opening up of the blood- nerve barrier. Moreover, this inflammation of the wall of the vasa nervorum may lead to vasospasm and aggravate nerve edema due to neurogenic inflammation. These events subsequently cause enhancement and dysfunction of the nerve (4,15,50). However, this theory cannot explain the pain associated with involvement of the abducens and the trochlear cranial nerves (15).
In our opinion, the most likely pathogenetic theory in OM is ischemic reversible breakdown of the blood-nerve barrier due to vasospasm during the migraine attack (13).
The cisternal portion of the III, IV and VI cranial nerves is supplied by branches arising from the cavernous tract of the carotid artery, and by a perforating vessel arising from proximal PCA (51). This region has a high intracranial trigeminal receptor density, making the oculomotor nerves vulnerable because of the vessel wall edema and ischemia to occlusion of the ostia. Shin et al. have well documented reversible ipsilateral ischemia in the territories of perforating branches of the PCA by brain single-photon emission-computed tomograpy (SPECT) in patients with OM during a headache attack and ophthalmoplegia due to III cranial nerve involvement (52). They concluded that reversible ischemia in the territories of the perforating branches of the PCA may accompany OM and possibly bear some relationship to the clinical features, even in the absence of enlargement and enhancement of the nerve at REZ (52). In his report on a patient with OM with pupillary sparing, Walsh and O’Doherty (36) suggested a possible ischemic mechanism. The authors also stated that most of the clinical evidence suggests that the ophthalmoplegia in OM is due to a delayed ischemic neuropathy, but ischemic neuropathy alone does not account for the enhancement of the thickened nerve. It is well known that the myelin of the cranial nerves at the REZ is of the central type, and the blood-nerve barrier is deficient at this level (53). During migraine attack the vasospasm of the vasa nervorum leads to failure of the blood-nerve barrier at the REZ, where it is less efficient, thus causing vasogenic edema accounting both for the focal enlargement and enhancement of the ocular motor nerves. During the resolving migraine attack, the regional cerebral blood flow turns to normal, the blood-brain barrier is restored, the vasogenic edema resolves, and the focal enlargement and the enhancement of the nerve is reduced or disappears
For all these reasons, we suggest that ischemic reversible breakdown of the blood-nerve barrier is the most probable pathogenesis in OM.
In 2009, in their excellent research involving 62 adult cases, Lal et al. wondered whether OM in children and in adults is the same disease (15). In accordance with Lal et al. (15), we believe that in adults as well the etiology is transitory reversible ischemia of the involved ocular motor nerve, but vasogenic edema seems to be less involved in this group, considering that in adults with OM the MRI is always normal.
We speculate that the blood-nerve barrier in adults is more mature and effective than in children, and we suggest that in adult patients the breakdown occurs but the barrier is more mature, hence edema is minimal, which is enough to cause malfunction of the nerve even when enhancement is absent. During the attack the different behavior of the pupillary fibers, most frequently spared in adults and usually involved in children, may support the theory of reversible ischemic neuropathy. It is well known that parasympathetic fibers travel on the superficial aspect of the cisternal portion of the oculomotor nerve, and are most susceptible to extrinsic compression by extraneural masses such as posterior communicating artery aneurisms. In children with OM the frequent pupillary involvement could be due to thickening and enlargement of the nerve causing a sort of compression of the fibers from inside, while in adults the absence of the enlargement of the nerve could account for sparing of papillary fibers.
Conclusions
We suggest including MR findings, which show enlarged strongly enhancing ocular motor nerves at the REZ, in the hallmarks of the disease in children, as we previously reported (13).
If the hypothesis of ischemic reversible breakdown of the blood-nerve barrier due to vasospasm during the migraine attack is correct, OM could in the future be classified under migraine.
We agree with Lal et al. (15) that OM in children and migraine with ophthalmoplegia in adults are the same disease.
Clinical implications
The finding of an enlarged strongly enhancing ocular motor nerves at the root exit zone (REZ) in magnetic resonance imaging (MRI) should be considered one of the hallmarks of ophthalmoplegic migraine (OM) in children. We believe that OM could in the future be classified under migraine. OM in children and migraine with ophthalmoplegia in adults should be considered the same disease.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflicts of interest
None declared.