Abstract
Introduction: Stroke-like migraine attacks after radiation therapy, or SMART syndrome, is characterised by migraine-like headache with or without aura, transient neurological dysfunction, including seizures, and gyriform enhancement on magnetic resonance imaging (MRI) which resolves over a period of weeks. Detailed neuropsychological characterisation in SMART syndrome is lacking and there are no published data on the course and pattern of neurocognitive recovery.
Results: The acute clinical presentation was one of unstable, fluctuating neurocognitive disturbances, complicated by seizure activity, followed by a longer term lag in the recovery of focal neuropsychological deficits which, we believe, was due to the more slowly resolving cerebral effects of SMART.
Conclusions: To our knowledge, this is the first case of SMART syndrome in which neuropsychological functioning has been comprehensively and serially examined. This case is also unique due to the development of complex partial seizures. We suggest that epileptiform activity during clinical seizures should not be regarded as inconsistent with a diagnosis of SMART, provided that the seizures do not explain the onset of the other clinical and radiological features.
Introduction
Proposed diagnostic criteria for SMART syndrome (2)
Neurological symptoms and neuro-imaging abnormalities typically resolve over a period of weeks and complete recovery eventually occurs (1–3). Cognitive manifestations have also been described (1–8). The most commonly reported neurocognitive deficits include confusion/disorientation, aphasia, and visuospatial deficits, such as neglect. Detailed neuropsychological characterisation in cases with SMART syndrome is lacking and there are no published data on the course and pattern of neurocognitive recovery.
We describe the findings from a detailed neuropsychological examination undertaken in a patient with SMART syndrome, both during an attack and upon serial neuropsychological review in the days and weeks that followed.
Case history
A 40-year-old, right-handed man presented with 3 days of severe left posterior headache, visual disturbance and language impairment. Nine years earlier, a right occipital metastatic melanoma had been detected, and treated with a right parieto-occipital craniotomy and resection, followed by a 6-week period of cranial radiation therapy (36 cGy to the right parieto-occipital region). Postoperative sequelae included a partial left upper quadrantanopia and mild subjective memory symptoms. Neurological examination on admission revealed a right homonymous hemianopia and mild expressive dysphasia. Cerebral magnetic resonance imaging (MRI; day 2) demonstrated diffuse cortical signal abnormalities involving the left occipital, temporal, and parietal lobes, with no convincing diffusion restriction, as well as gyriform and leptomeningeal enhancement in these regions (Figure 1A,B). Headache persisted and he subsequently developed recurrent complex partial seizures (CPS), the semiology of which consisted of dystonic posturing of the right arm, eye–head deviation to the right, and nystagmus. The immediate post-ictal phase was characterised by disorientation, dysphasia, visual field loss, and difficulties with visual recognition. An initial EEG (day 2) revealed left occipital slowing, but no frank epileptiform activity was seen. Ictal EEG during a witnessed CPS (day 8) showed epileptiform discharges arising from the left posterior quadrant (Figure 2). Anti-epileptic drug therapy was commenced, first with carbamezepine then topiramate as he could not tolerate the former. One week later, neurological signs had largely resolved, proceeding to complete neurological recovery within 2 weeks of admission. Repeat neuro-imaging 5 months later showed that the cortical signal abnormalities and leptomeningeal enhancement had resolved, with no evidence of regional brain atrophy (Figure 1C,D).
MRI at presentation (A,B) and 5 months (C,D) in a case of SMART syndrome. (A) Axial FLAIR image on presentation demonstrates high FLAIR signal posteriorly on the left, particularly involving cortex. There is also high signal in the right occipital region consistent with previous resection. (B) Coronal post-contrast T1 weighted image on presentation demonstrates gyral enhancement in the left parieto-occipital region best appreciated medially in the left occipital lobe. (C) Axial FLAIR image at 5 months demonstrating resolution of the left posterior high FLAIR signal (D) Coronal post-contrast T1 weighted image at 5 months demonstrates complete resolution of the gyral enhancement and there is no evidence of regional brain atrophy on the left. The pre-existing resection site on the right remains unchanged in appearance. Ictal EEG in bipolar longitudinal montage. There is rhythmic, high amplitude, sharply contoured 2 Hz delta activity arising from the left posterior quadrant, with some spread to the right. The patient had impaired awareness with dystonic posturing of the right arm and eye-head deviation to the right.
Initial neuropsychological examination
Subset of neuropsychological data obtained during the acute and late recovery phase in a patient with SMART syndrome
Performances that remain below estimated premorbid level.
Neuropsychological examination during the early and late recovery phase
Neuropsychological review on day 14 revealed a marked improvement. Moderate improvements in attention, new learning and verbal recent memory function were noted, but performances remained below expectation (Table 2). Marked improvement in visuoconstructional function was evident, but he continued to approach the reproduction of a complex figure in a piecemeal and segmented fashion and was unable to extract numbers from Ishihara plates despite testing in the relatively intact visual field and normal visual acuity. These findings raised the question of a residual simultanagnosia (9).
Language was fluent with no apparent word-finding difficulties. The right homonymous hemianopia and object recognition deficits had resolved. He was able to direct his gaze voluntarily and displayed normal hand–eye co-ordination. The dysnomia had largely resolved (Table 2). He could read fluently and writing was fully legible, with only minor graphemic errors persisting.
At 6-week out-patient review, the patient performed at estimated premorbid levels on measures of attention, working memory, verbal memory, calculation, reading, and writing. Language and visuoconstructional function had largely returned to normal, although subtle high level deficits persisted on sensitive tasks (Table 2).
Discussion
SMART is characterised by migraine-like headache with or without aura, associated transient neurological dysfunction, including seizures, and gyriform enhancement on MRI which resolves over a period of weeks. We describe the neuropsychological profile and pattern of recovery in a case of SMART syndrome with witnessed CPS, pronounced ictal EEG abnormalities, and compatible MRI abnormality resolution. To our knowledge, this is the first case of SMART syndrome in which CPS are described and neuropsychological functioning is comprehensively and serially examined.
Examination of neurocognitive function during the acute, early, and late recovery phase revealed a complex pattern of focal cortical deficits, referable to involvement of the left temporal and parietal cortices, and the occipital cortex bilaterally. During the acute stage of recovery, our patient demonstrated a generalised dampening of cognitive function, associated with pronounced attentional impairment, likely to have been compounded by seizure activity. Nevertheless, several focal cortical neurocognitive deficits, predominately affecting language, visuoperceptual, and visuoconstructional function were observed, into which the patient retained insight.
The patient also experienced CPS. Generalised seizures can occur in SMART syndrome (2), but CPS, as observed in our case, have not been previously described.
There are several points that can be made in relation to this finding and its relevance to differential diagnosis. First, the onset of neurological symptoms and cognitive impairments preceded the development of CPS. Second, MRI abnormalities characteristic of SMART syndrome were reported on the same day as the initial EEG study, which did not reveal any epileptiform activity. Third, focal epileptiform activity was seen for the first time during witnessed complex partial seizures that emerged well after the onset of the neurological, cognitive, and MRI abnormalities mentioned above (but without interval recovery of these features).
The diagnostic criteria proposed for SMART syndrome include clinical seizures as a component of the symptom complex, but the case reports accompanying these criteria imply that epileptiform abnormalities are typically absent. Our case suggests that epileptiform activity during clinical seizures should not be regarded as inconsistent with a diagnosis of SMART, provided that the seizures do not explain the onset of the other clinical and radiological features.
Partial status epilepticus is also an important differential diagnosis since it can produce SMART-like MRI abnormalities (10,11), such as transient leptomeningeal enhancement with cortical hyperintensity. Lansberg and colleagues (11) reported, in a series of three patients with complex partial status epilepticus, that subtle regional atrophy is present on follow-up neuro-imaging. In our case, the left cortical abnormalities had resolved on 5-month follow-up MRI, without evidence of residual atrophy (Figure 1).
Significant residual neurocognitive deficits were observed on serial review 2 weeks after symptom onset, and 6 days after the last clinical seizure. The likelihood that there was a residual post-ictal effect on cognitive function is believed to have been minimised by this time. These deficits were not apparent on gross neurological examination, but could be detected readily with sensitive neuropsychological measures (Table 2). Disruption of language, visuoperceptual, and constructional functions, referable to the left temporoparietal and occipital cortices bilaterally, were documented. These impairments were functionally disabling. By 6 weeks, fundamental neuropsychological functions had recovered to the extent that the patient had resumed many of his previous activities of daily living. There was evidence, however, of subtle persisting neuropsychological dysfunction on measures of higher level language and visuoconstructional function.
In terms of the overall pattern of neurocognitive recovery, the acute clinical presentation was one of unstable, fluctuating neurocognitive disturbances, complicated by seizure activity, which stabilised following anticonvulsant therapy. Over the ensuing days to weeks there was a longer term lag in further neurocognitive recovery which, we believe, was due to the more slowly resolving cerebral effects of SMART. This highlights the need for careful neuropsychological characterisation and serial review of neurocognitive function to ensure appropriate patient management in cases of SMART syndrome.