Ischaemic strokes with reversible vasoconstriction and without thunderclap headache: a variant of the reversible cerebral vasoconstriction syndrome?

BACKGROUND: Reversible vasoconstriction (RV) may cause ischaemic stroke (IS) in the absence of any other defined stroke aetiology. The three objectives of our study were to evaluate the frequency of RV in a prospective series of young IS patients, to describe the detailed clinical-radiological features in the patients with RV and IS, and to compare these characteristics with those of reversible cerebral vasoconstriction syndrome (RCVS). METHODS: We identified between October 2005 and December 2010, 159 consecutive young patients (<45 years) hospitalized for an acute IS confirmed by cerebral magnetic resonance imaging. An extensive diagnostic work-up was performed including toxicological urinary screening for cannabis, cocaine and amphetamines, and the usual biological, cardiac and vascular investigations for an IS in the young. We specifically studied patients with IS and RV, which was defined as multifocal intracranial arterial stenoses confirmed by intracranial arterial imaging that resolved within 3-6 months. RESULTS: Out of 159 patients with IS, 21 (13%, 12 males, 9 females; mean age 32 years) had multifocal cerebral arterial stenoses that were fully reversible at 3-6 months, and no other cause for stroke. IS were located on posterior territory in 71% of cases, and vasoconstriction predominated on posterior cerebral and superior cerebellar arteries. Precipitating factors of IS and RV were the use of cannabis resin (n = 14), nasal decongestants (n = 2) and triptan (n = 1). Most cases (74%) had unusual severe headache, but none had thunderclap headache. None of 21 cases had reversible posterior leukoencephalopathy, cortical subarachnoid or intracerebral haemorrhage. CONCLUSION: RV was the sole identified cause of IS in 13% of our cohort. These young patients with IS and RV may have a variant of RCVS, related to an increased susceptibility to vasoactive agents in some individuals. RV in our patients differs from the classical characteristics of RCVS by the absence of thunderclap headache, reversible brain oedema and subarachnoid or intracranial haemorrhage. Intracranial arteries should be looked for, by appropriate vascular imaging, in young patients with IS at the acute stage and during the follow-up period.