Symptomatic intracranial hemorrhage following intravenous thrombolysis for acute ischemic stroke: a critical review of case definitions.

BACKGROUND Symptomatic intracranial hemorrhage (SICH) is a devastating complication of intravenous thrombolysis treatment that is associated with high mortality. Clinical trials, stroke registries and cohort studies employ different case definitions to identify stroke patients with SICH following intravenous thrombolysis. We systematically reviewed the reported rates of SICH following intravenous thrombolysis and compared their consistency with mortality outcomes. METHODS Studies were identified from the PubMed and Embase databases from January 1994 to July 2011 by cross-referencing the following MeSH terms: 'thrombolysis', 'recombinant tissue plasminogen activator', 'rtPA', 'hemorrhagic stroke', 'cerebral hemorrhage', 'hematoma' and 'ischemic stroke'. Demographic information, baseline National Institute of Health Stroke Scale (NIHSS) scores, time from stroke onset to intravenous thrombolysis, SICH and mortality rates were derived from published data in 7 randomized controlled trials, 7 stroke registries and 10 cohort studies (4 multicenter and 6 single center) with more than 200 consecutively recruited patients. Mortality rates were considered as the percentage of patients treated with intravenous thrombolysis who died within 90 days after stroke. RESULTS The mean age of patients included in this analysis was 68.8 years (standard deviation, SD 2.9, range 63-75), of whom 56.3% (SD 4.5, range 45-63) were men. They presented with a mean baseline NIHSS of 12.5 (SD 1.4, range 9-15) and received intravenous thrombolysis 175 min (SD 62, range 120-328) from stroke onset. The overall mean SICH and mortality rates of patients treated with intravenous thrombolysis were 5.6% (SD 2.3) and 14.7% (SD 4.8), respectively. A moderate correlation was observed between the incidence of SICH and mortality in patients treated with intravenous thrombolysis (r = 0.401, p = 0.050). The variation in SICH rates was highest across studies that reported SICH rates using the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST) criteria compared with the European Cooperative Acute Stroke Study and National Institute of Neurological Disorders and Stroke (NINDS) criteria. Studies that defined SICH as parenchymal hemorrhage with a neurological decline NIHSS ≥4 occurring within 36 h of intravenous thrombolysis reported a higher consistency between SICH and mortality rates (correlation coefficient 0.631). CONCLUSIONS SICH rates vary considerably between studies and these differences may relate to the differences in the criteria used to define SICH. Until a case definition with high interrater agreement and good correlation with stroke outcomes becomes available, detailed information on the type of bleeding, the extent of NIHSS deterioration, neuroimaging features and the time from thrombolysis to diagnosis of hemorrhage should be reported to permit a correct interpretation of SICH rates.