Patients with stroke or transient ischemic attack (TIA) are at high risk of poor outcome, recurrence of cardiovascular events or vascular death.Until now, no reliable predictive biological marker could be identified in the acute phase of stroke.We hypothese that, in the acute phase of ischemic stroke or TIA, the increase of cardiac ( brain natriuretic peptide, BNP) or renal markers (albuminuria, cystatin C)might predict recurrence of cardiovascular events or vascular death. We want to assess which one of these markers has the best prognosis value , in a prospective study of 300 stroke patients followed during 3 years.

Patients with ischemic stroke or TIA will be included in the 48 hours following the onset of symptoms, in 3 university hospitals.For each patient will be collected

• 1 blood sample for the BNP measure in pg/ml
• 1 blood sample for the Cystatin C measure in mg/l
• 2 urinary samples to measure the albumine/creatinine ratio, at inclusion and 5 days after the inclusion, and to study the variation of albuminuria in micrograms/min.

The clinical follow-up will be organized during 3 years. The following events will be notified: poor outcome (defined with the Barthel and Rankin scores), recurrence of cardiovascular or cerebrovascular events, vascular death.

After adjustment on the main clinical prognosis factors, we want to determine wich one of these markers has the best prognosis significance and allows to identify in the acute phase the "high-risk" patients, in order to intensify individual treatment and secondary prevention.

Inclusion Criteria:

• Ischemic stroke or TIA in the first 48 hours following the onset of symptoms

Exclusion Criteria:

• Patient unable to give an informed consent
• Urinary sample impossible to collect

Patient with Ischemic stroke or TIA