023. The proportion of favorable functional outcome across studies were heterogeneous, I2: 60%, 95% CI: 22-80%. Rates of good functional outcome at study level are presented find more as a Forest plot in Fig 2. The direction of association did not change after excluding the 2 studies (one from each group) where the proportion of patients with mRS of 0 or 1 at last available follow-up was not

provided. The magnitude of association decreased from 1.6 to 1.4 and significance could not be detected because of the small sample size. Assessment for publication bias for favorable outcome revealed no publication bias for .9 mg/kg and suggested 2 missing studies for .6 mg/kg yielding an estimate of 35%. Partial or complete recanalization was observed in 179 (56%) of patients in the .6 mg/kg group compared with 94 (67%) of patients in the .9 mg/kg group, OR 1.57 (95% CI 1.03-2.37, P= .03). There was only borderline significance in the difference of the

rates between the 2 treatment groups using the random effects model (P= .07). Heterogeneity across studies regarding angiographic recanalization rates was high I2: 72% (50-84%). Rates of angiographic recanalization in the studies included in the analysis are shown as a Forest plot in Fig 3. Clinical and angiographic outcomes are summarized in Table 5. Assessment for publication bias for partial or complete recanalization revealed no publication buy Romidepsin bias. We found no significant difference in sICH rates between the .6 mg/kg (8%) and the .9 mg/kg (7%) groups. In the

.9 mg/kg group, rates of angiographic recanalization and favorable functional outcome appeared to be higher (OR 1.60, 95% CI 1.07-2.40 and OR 1.57, CI 1.03-2.37, respectively) when compared using a logistic regression model with events/trial syntax. Using the more stringent random effects model, the results were similar with the exception of recanalization, which achieved only borderline significance. The .9 mg/kg dose for IV rt-PA was established following the 2 NINDS dose-finding studies.13,14 Escalating rt-PA doses were administered to patients, within 90 minutes from stroke onset in Part I13 and between 91 and 180 minutes from onset in Part II.14 No sICH was noted in the 58 patients who received Nabilone .85 mg/kg of IV rt-PA or less in Part I versus 3/26 patients who had received a dose of .95 mg/kg or greater. Higher doses of rt-PA were significantly related to the risk of developing sICH (P= .045). There was no clear correlation between early neurological improvement and rt-PA dose administered. Based on these findings, an intermediate dose between .85 and .95 mg/kg was selected for the NINDS efficacy trial.1 Subsequent studies combining IV thrombolysis and endovascular treatment were designed to avoid exceeding a total dose of .9 mg/kg rt-PA by administering a partial IV dose (.6 mg/kg) followed by IA administration of up to .3 mg/kg. Our findings suggest that .