Intravenous thrombolysis without thrombectomy in the 4.5–24–hour time window guided by perfusion imaging: a systematic review and meta-analysis

Recent clinical trials have demonstrated improved neurological outcomes with intravenous thrombolysis (IVT) compared to standard care (SC) in patients presenting beyond 4.5 h. This review provides updated outcomes of IVT versus SC in the 4.5–24–hour time window for non–thrombectomy patients selected by perfusion imaging. Systematic review and meta-analysis of published, phase III randomized clinical trials through April 2026 of anterior and posterior circulation ischemic stroke patients comparing IVT with SC at 4.5–24 h, selected by perfusion imaging, when thrombectomy was not planned. The primary and secondary outcomes were 90–day modified Rankin Scale (mRS) scores 0–1 and 0–2, respectively. Safety outcomes included symptomatic intracranial hemorrhage (sICH) and 90-day mortality. Common-effects models via the Mantel-Haenszel method were used to calculate risk ratios (RR) and 95% confidence intervals (CIs) for the outcomes. Three studies and a combined 1,454 patients (IVT group: n = 732; SC group: n = 722) were included in the pooled analysis. IVT significantly increased the rate of 90-day mRS scores 0–1 (RR 1.36, 95% CI 1.18–1.58; p < 0.01) and 0–2 (RR 1.20, 95% CI 1.08–1.33; p < 0.01), though was also associated with higher sICH rates (RR 6.56, 95% CI 2.15–20.00; p < 0.01). No statistically significant difference in mortality was observed between the groups (RR 1.09, 95% CI 0.79–1.50; p = 0.61). Although the exploratory subgroup analysis showed that internal carotid artery occlusion patients had similar rates of excellent 90-day outcomes with IVT and SC, interaction testing did not demonstrate a heterogeneous treatment effect of IVT by occlusion site. Our findings suggest that non-thrombectomy patients selected by perfusion imaging in the 4.5–24–hour time window have better clinical outcomes with IVT, despite higher rates of sICH and no statistically significant difference in mortality. This metaanalysis was registered with PROSPERO (CRD420261388845).