Abstract
There are many case reports of structural brain lesions that appear to cause or provoke headaches that are clinically indistinguishable from primary headache syndromes. ‘Symptomatic migraine’ is the term used under the second edition of the International Classification of Headache Disorders (ICHD-II) to describe such cases, which mimic migraine. However, symptomatic migraine with visual aura (MA) is rare. It has been reported most often in association with parieto-occipital arteriovenous malformations and rarely with cerebral venous thrombosis, extracranial artery dissection and cerebral metastases (1). Migrainous visual aura is believed to be caused by cortical spreading depression (CSD) generated in the visual cortex, and in ‘symptomatic’ cases the anatomical relationship of the lesion to this brain region is presumed to explain the aura. However, we report here the rapid evolution of MA in a patient who proved to have a colloid cyst of the third ventricle, an area remote from the visual cortex. There was no previous or family history of migraine and the MA symptoms resolved immediately and completely following removal of the cyst. On review more than a year later, the patient had experienced no further headache or aura, suggesting that the colloid cyst had provoked the syndrome.
Case history
A 35-year-old man began to experience recurrent stereotyped attacks of visual disturbance that would occur without warning or provocation. He would suddenly develop a coloured, jagged edged obscuration in the centre or to the right side of his visual field, which expanded to encompass the entire visual field over about half an hour to an hour, followed by a very severe, pulsating, holocranial or sometimes right-sided headache. Photo- and phonophobia were not prominent. The aura would be associated with a feeling of disorientation and vague cognitive problems, such as difficulty concentrating. The headache would last about 2 h before gradually fading, although on occasions it could last up to 2 days. He often experienced nausea with the headache and two attacks were associated with vomiting.
The patient is a computer graphics artist and was able to generate images showing precisely what he experienced during the aura phase (Fig. 1). This was not typical of visual auras as commonly reported, being bilateral and comprising repeated ‘positive’ jagged lines with no obvious ‘negative’ (scotoma) component.
Patient's computer graphic of his visual aura. (A) Early stage. (B) Late stage.
The attacks were quite frequent from the outset, occurring several times over 4–5 days followed by several days without symptoms, but the events became progressively more frequent in subsequent months until they were daily. By this point, he had also begun to develop other neurological symptoms. His balance was deteriorating and he had odd sensory symptoms in the legs. He also began to have episodes of memory dysfunction and on one occasion he had a typical attack of transient global amnesia, during which he repeatedly asked the same questions of his partner despite being given the answers consistently. At no time did he experience significant cough headache or headache precipitated by turning the head or neck, visual obscuration on positional changes, loss of consciousness, or vomiting independent of the headache attacks. There was no previous personal or family history of headache or other notable medical problems.
On examination some 4 months after onset of the symptoms, the only abnormality was mild blurring of the margins of the optic discs and loss of retinal venous pulsations.
In view of the rapid evolution of the symptoms, the history of episodic ataxia and amnesia, and the abnormality of the optic discs suggestive of early papilloedema, an urgent computed tomography (CT) brain scan was performed. This revealed a colloid cyst of the third ventricle with obstructive hydrocephalus (Fig. 2). An emergency craniotomy was undertaken via a frontal transcortical approach and the lesion was excised. In the immediate postoperative period, neuropsychometric testing revealed significant verbal and visual memory defects, presumably relating to mild fornix injury from the surgical procedure. However, re-evaluation 6 weeks later showed marked improvement and he had no significant functional problems.
Computed tomography brain scan, showing colloid cyst of the third ventricle and obstructive hydrocephalus.
His MA attacks ceased immediately after operation, and on review in clinic 6 months later he was completely asymptomatic. E-mail enquiry a year after the procedure confirmed that he had experienced no further headache, aura or other neurological symptoms.
Discussion
‘Incidentalomas’—reported incidence of asymptomatic brain tumours on magnetic resonance imaging in 4536 normal healthy subjects*
Excludes congenital vascular malformations and arachnoid cysts. Adapted from Weber F and Knopf H. Incidental findings in magnetic resonance imaging of the brains of healthy young men. J Neurol Sci 2006; 240:81–4, and Vernooij MW, Ikram MA, Tanghe HL, Vincent AJ, Hofman A, Krestin GP et al. Incidental findings on 2000 MRI scans performed in the Rotterdam Study—a prospective population-based cohort study among persons 55 years or older N Engl J Med 2007; 357:1821–8.
Migrainous visual aura is thought to be caused by CSD initiated in the relevant occipital cortex (3). Lesions previously associated with symptomatic migraine with visual aura have involved the cortex either directly or indirectly (1). The thalamus can be a site for spreading depression in experimental animals (4) and a colloid cyst of the third ventricle could, conceivably, trigger a spreading depression in the human thalamus perhaps by a pressure effect. Otherwise, it is not clear how this lesion could induce CSD. This patient's aura was also unusual in that it eventually involved both visual fields rather than being restricted to one hemifield as is typical in spontaneous attacks. This would suggest that CSD in our patient was generated in both occipital cortices. In addition, it comprised only positive elements (jagged lines) without subsequent scotoma.
Colloid cysts are occasional incidental findings on neuroimaging and usually remain asymptomatic (5). Symptoms arise when the cyst blocks the foramina of Monro, causing internal hydrocephalus (6). This can cause sudden death, but more commonly patients develop intermittent and then progressive neurological symptoms. The most common of these is headache, but this is typically paroxysmal and often precipitated by activities such as coughing, straining or head turning (7). Occasionally, the headache may be suggestive of ‘migraine’, but in a review of the published manifestations of symptomatic colloid cysts, Young and Silberstein did not include any instance of migraine with visual aura, although they did describe a patient with a previous history of MA who developed a new type of headache due to colloid cyst (7).
A colloid cyst of the third ventricle would not be expected to generate visual aura. Raskin et al. described several patients who experienced what might be interpreted as fragments of visual aura during stereotactic stimulation of the periaqueductal grey matter for intracj.1468-2982.2009.01978.x pain. Symptoms included blindness, flashes of light, wavy and zig-zag lines, stars and scotomas throughout the visual fields, but they stressed that no patient experienced a slowly evolving fortification spectrum typical of MA (8).
Obstructive hydrocephalus might occasionally be capable of provoking migraine. Schlake et al. reported migraine-like episodic headache without aura in a patient who was found to have obstructive hydrocephalus from an undefined pathology, but the history extended back some 18 years [(9), Case 3]. Donaldson described a 21-year-old woman who suffered attacks of MA between 10 and 15 years old that ceased for 6 years when she was treated with propranolol for essential tremor. She then suddenly developed frequent attacks of an identical nature once again. She was noted to have bilateral papilloedema and CT brain scan revealed a colloid cyst and hydrocephalus. The attacks ceased immediately after the cyst was removed, as in our case [(10), Case 1]. Donaldson also described a second case of MA of 1 year's duration in a patient with obstructive hydrocephalus caused by a tumour in the ventral mid-brain and pons. This patient had no previous history of headaches and the MA resolved completely following shunting and brainstem radiotherapy [(9), Case 2]. We have also previously described severe, paroxysmal headache with transient visual hallucinations in a man who was shown to have a colloid cyst of the third ventricle with obstructive hydrocephalus. His severe MA attacks resolved following removal of the cyst, but he had a previous and family history of migraine without aura, which continued as before following surgery [(11), Case 237].
The only common feature in these cases is obstructive hydrocephalus, and we agree with Donaldson that it would seem that this process alone might rarely be capable of provoking CSD by some as yet undefined mechanism.