Article Text


Prolonged hemiplegic migraine associated with unilateral hyperperfusion on perfusion weighted magnetic resonance imaging
  1. A J Lindahl1,
  2. S Allder1,
  3. D Jefferson1,
  4. S Allder2,
  5. A Moody2,
  6. A Martel2
  1. 1Department of Clinical Neurology, University Hospital, Nottingham, UK
  2. 2Department of Academic Radiology, University Hospital, Nottingham
  1. Correspondence to:
 Dr A J Lindahl, Frenchay Hospital, Bristol BS16 1LE, UK;

Statistics from

Hemiplegic migraine (HM) is an unusual subset of migraine with aura, in which headache is associated with unilateral motor deficits, thought to be attributable to an underlying calcium channelopathy.1 In some cases the neurological dysfunction may outlast the headache and persist for many days. In the initial stages, hemiplegic migraine may mimic cerebral infarction. Within the first few hours after stroke, both computed tomography and magnetic resonance imaging (MRI) are often normal. However, more information can be gained using diffusion (DWI) and perfusion weighted imaging (PWI), which are much more sensitive to acute events in cerebral ischaemia. A recently reported post-processing technique (factor analysis of dynamic studies FADS) can be applied to PWI to generate images representing arterial (“early”) and venous (“late”) contributions to signal intensity.2 This report outlines the findings arising from the application of multimodal MRI techniques to a patient with prolonged hemiplegic migraine.

Case report

A 21 year old woman, with a long history of familial HM, presented with a six hour history of headache, nausea, right sided weakness, and expressive dysphasia. Her maternal aunt had also suffered with the condition. Genetic testing for CACNA 1A had not been performed. In previous episodes, her symptoms had resolved within four hours, so a clinical diagnosis of migraine related stroke was made and she underwent urgent T2 weighted MRI and MR angiography with DWI and PWI. FADS was applied to the perfusion weighted images. These sequences revealed no large vessel obstruction and normal DWI, thus excluding infarction, but there was gross hyperperfusion of the left hemisphere on the early FADS images (fig 1A). The late FADS images, representing venous signal were unremarkable (data not shown).

Figure 1

Multimodal MRI findings during the course of an episode of hemiplegic migraine. (A) Day 1, 6 hours post-onset. Dense hemiparesis, aphasia, and severe headache associated with normal diffusion weighted imaging (DWI), magnetic resonance angiography (MRA), and gross hyperperfusion of the left hemisphere on factor analysis of dynamic studies (FADS). (B) Day 4. Weakness improving, dysphasic, persistent headache. Hyperperfusion of inferior frontal lobe on FADS. (C) Three months. Asymptomatic. Full resolution of perfusion changes. (DWI not repeated at three months).

Her headache persisted for the next 10 days. However, the right limb weakness improved over the next four days, but the dysphasia persisted. Repeat PWI on day four showed focal hyperperfusion of the inferior frontal lobe (fig 1B). A follow up scan at three months, at which time she was totally asymptomatic, showed complete resolution of these appearances (fig 1C).


This report demonstrates multimodal MRI findings in a case of HM and that by applying the techniques of DWI, PWI, and FADS, it was shown that hyperperfusion existed in the phase of neurological deficit thus excluding acute stroke.

There have been conflicting views regarding whether migraine is primarily attributable to vascular or neuronal dysfunction. One of the earliest models of its pathogenesis proposed that symptoms of migraine aura were related to hypoperfusion and ischaemia and that the subsequent headache was related to reactive hyperaemia. Initial studies of regional cerebral blood flow initially seemed to support this theory, demonstrating hypoperfusion in the aura phase and hyperperfusion in the headache phase. However, the observation by some authors that hypoperfusion may outlast the aura symptoms and extend into the headache phase makes a simple vascular pathogenesis less likely.

Our findings of normal DWI appearances associated with hyperperfusion in the context of neurological deficit and headache also argue against an ischaemic mechanism. The presence of hyperperfusion affecting the entire hemisphere and therefore not respecting anatomical boundaries is difficult to explain on a simple vascular model and may therefore be a secondary phenomenon resulting from underlying neuronal dysfunction. The concept of a primarily neuronal dysfunction in migraine has been drawn from parallels with the experimental phenomenon of “cortical spreading depression” (CSD) described by Leao.3 The rate of propagation of CSD is similar to Lashley's estimate of the rate of progression of symptoms during the human migrainous visual aura.

Subsequent studies have shown that in migraine with and without aura there is a similar spreading hypoperfusion across the cerebral hemispheres. This has also been demonstrated in HM.4 However, hemispheric hyperperfusion during HM, correlating with contralateral hemiplegia has also been previously reported.5 On the balance of available evidence, it seems likely that in the early phase of migraine there is spreading cortical oligaemia and that this is linked to a neuronal dysfunction analogous to that reported in CSD. Oligaemia was not found in our patient although this may be because she was not scanned at the very onset of her attack.

HM has been linked to mutations in the CACNA 1A gene, which encodes a P/Q type calcium channel.1 This channel is located on presynaptic membranes and is tightly coupled to neurotransmitter release. Conceivably, CACNA 1A mutations could cause neuronal membrane instability resulting in episodic loss of neural control of cerebrovascular tone and that, as neuronal membrane function recovers, neurogenic control of rCBF is regained. It is of interest, in this respect, that the recovery of neurological function in our patient was closely mirrored by the restoration of normal cerebral perfusion.

In conclusion, we show that the use of multimodal MR in a case of acute hemiparesis in a patient with HM was effective in excluding cerebral infarction. FADS analysis demonstrated reversible focal hyperperfusion of the cerebral cortex correlated with the clinical distribution of neurological deficit. Recovery was mirrored by the restoration of a normal vascular pattern.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.