Background and Purpose-In a model of mechanical focal ischemia, we investigated the involvement of thrombolysis products (TLP) in recombinant tissue plasminogen activator (rtPA)-induced intracerebral complications and the effects on infarct volume and postischemic endothelial function. Methods-Hemorrhage incidence and severity were evaluated by histomorphometric analysis in male spontaneously hypertensive rats (SHR) subjected to 60-minute intraluminal middle cerebral artery (MCA) occlusion and receiving intravenously 5 hours later either saline, rtPA (3, 10, or 30 mg/kg), or rtPA (10 mg/kg) associated with TLP (rtPAϩTLP). In addition, MCA reactivity was assessed in rtPA-or rtPAϩTLP-treated SHR versus control Wistar-Kyoto rats or SHR. Results-No hemorrhage was observed visually in SHR receiving saline. In contrast, rtPA administration induced hemorrhagic complications in infarcted areas in a dose-independent manner. Administration of rtPAϩTLP solution, containing a high concentration of plasmin, did not affect hemorrhage incidence but significantly increased hemorrhage severity (8.8Ϯ2.3 petechiae versus 3.0Ϯ1.0 petechiae in rtPA group; PϽ0.001). This increased severity was associated with a significant increase of both infarct volume (182Ϯ10 versus 144Ϯ15 mm 3 in rtPA group; PϽ0.01) and postischemic impairment of MCA endothelium-dependent relaxation (9Ϯ0.5% versus 13Ϯ1% in rtPA group; PϽ0.05). Key Words: hemorrhage Ⅲ ischemia, focal Ⅲ plasmin Ⅲ thrombolysis Ⅲ tissue plasminogen activator T hrombolytic therapy with recombinant tissue plasminogen activator (rtPA) has demonstrated improvement in clinical outcome in acute ischemic stroke. 1,2 However, the threat of intracerebral hemorrhage (ICH) is an important barrier to widespread administration of thrombolytic agents. While many factors, such as elevated blood pressure, diabetes, age, and rtPA administration time, contribute to the occurrence of ICH, the underlying mechanisms remain elusive. [3][4][5] ICH are preferentially located in the infarct area, suggesting a role of ischemia per se. 6,7 In response to ischemic and inflammatory stimuli, vascular modifications may also be involved in ICH pathophysiology. 8 In particular, endothelium-dependent relaxation, a marker of ischemia-induced impairment of cerebral vasculature, 9,10 is worsened in vitro by the application of rtPA. 11 It remains unknown whether rtPA directly induces ICH or if the interaction between rtPA and thrombus, via the resulting thrombolysis products (TLP), is involved. Existence of rtPA-induced ICH in thromboembolic models has been well established, 7,12-15 whereas it remains more uncertain in mechanical models of ischemia. 6,16 The difference between the 2 models suggests that thrombus may contribute to the occurrence of ICH. To test this hypothesis, we compared the effects of rtPA or rtPA-induced TLP in a model of mechanical middle cerebral artery (MCA) occlusion on ICH incidence and severity. We also studied infarct volume and MCA endothelial function to investigate the mechanism of ICH occur...