[Cryptogenic cerebral infarction: from classification to concept].

Since the foundations laid by Sacco and Mohr in 1989, from the Stroke data bank, cryptogenic infarctions have had a predominant place among the causes of ischemic strokes. In that study, they accounted for approximately 40% of the stroke causes. Cryptogenic infarctions are infarctions without a defined cause, despite a complete work-up; they differ from infarctions of undetermined causes, which may involve overlapping causes or an incomplete investigation. The size of this group will probably shrink as knowledge advances. Patent foramen ovale (PFO), with or without a septal aneurysm, is more frequent in patients with a cryptogenic infarction. Transesophageal echocardiography is the reference examination for screening for these abnormalities. A meta-analysis of several case-control studies showed a significant association between PFO and stroke in subjects younger than 55 years. For now, these septal abnormalities constitute a risk factor but not a cause. Complex aortic atheroma affecting area upstream of the left subclavian artery may be a source of cerebral embolisms in some conditions. The prevalence of this disease increases with age. It is identified most frequently in patients older than 60 years with a cryptogenic infarction. The thickness of the atheromatous plaque determines whether it is a risk factor or a cause. Recent stroke classifications do not consider carotid atheromatous lesions less than 50% to be a source of ischemic stroke. Nonetheless some studies identify moderate stenosis of the carotid artery more frequently in infarctions of unknown causes than in other categories. The increased risk of cerebral infarction when parents and homozygous twins have a history of stroke suggests that there may be genetic causes that have not yet been detected. An unknown genetic cause would thus be included in the infarctions of unknown causes. A recent study tested for Fabry disease in young patients with a cryptogenic infarction: 4.9% of the men and 2.4% of the women had a functional mutation of the alpha-galactosidase gene. These findings must be confirmed. Some studies suggest an association between cryptogenic infarction and hereditary thrombophilias. Nonetheless the risk attributable to these thrombophilic disorders is slight and the discovery may be only a coincidence. The work described above shows the importance of stratification in the identification of stroke causes: age older or younger than 55/60 years, type of interatrial abnormality (PFO and aneurysms of the interatrial septum), type of atheroma of the aortic arch (more or less than 4mm). They also show the difficulty involved in attributing cause to an identified abnormality: is carotid stenosis of less than 50% a marker of atherosclerosis or also a cause of stroke? To continue improving our understanding of the mechanisms of strokes, new investigational techniques are under evaluation. They include magnetic resonance imaging (MRI), computed tomographic angiography (CT), positron emission tomography (PET) of carotid plaque and of the aortic arch, transcranial Doppler, cardiac recording by telemetry, and even new biological assays.