Functional Ultrasound Imaging Uncovers Vascular Connectivity and Dynamics in Awake Mice During Hyperacute Stroke Phase

BACKGROUND: Acute ischemic stroke models often rely on anesthesia, which alters neurovascular coupling and limits real-time functional assessment. We tested the hypothesis that an awake mouse model of thromboembolic stroke, combined with multimodal imaging, would reveal very early cerebrovascular dynamics and connectivity changes predictive of final lesion outcomes. METHODS: Male Swiss mice (6–8 weeks old; 30–40 g; n=60) were implanted with a cranial headplate and habituated to restraint and imaging. One microliter pneumatic injection of murine thrombin (1 IU) into the distal middle cerebral artery induced in situ clot formation. Twenty minutes postocclusion, mice received intravenous rtPA (recombinant tissue-type plasminogen activator; Alteplase, 10 mg/kg; 10% bolus/90% infusion over 40 minutes, n=31) or saline (n=29). Primary outcomes included lesion volume (T2-weighted magnetic resonance imaging at 24 hours), brain perfusion (ASL magnetic resonance imaging), cerebral blood volume variations (ultrafast Doppler US imaging), resting-state connectivity, and neurovascular coupling to whisker stimulation (functional ultrasound imaging). Sample size (n=12 per group for imaging) was based on prior variability. Statistical analyses included unpaired t tests, repeated-measures ANOVA with Dunnett or Tukey post hoc, and simple linear regression; significance set at P <0.05. RESULTS: At 24 hours, rtPA reduced lesion volumes by 36.7% (10.97±4.7 versus 17.33±5.92 mm 3 in controls; t 58 =4.624; P <0.0001). ASL magnetic resonance imaging revealed a 66.5±9.9% CBF drop at 1 hour (F=48.63; P <0.0001) and a 51.2±45.1% hyperperfusion at 24 hours compared with baseline (F=11.67; P =0.0024). CBV declined by 59.2±12.9% at 10 minutes and partially recovered with rtPA at 1 hour (+87.3±30.6%; P =0.0165). Early hypoperfused area (10 minutes) predicted final lesion observed at 24 hours ( R ²=0.6465; P =0.0016). Resting-state connectivity shift of 12.0° at 10 minutes was mitigated by rtPA by 1 hour (14.8° versus 3.6°; P <0.01). Whisker-evoked CBV responses were abolished ipsilaterally at 10 minutes (−100.5±3.3%; P =0.0038) and showed partial recovery by 24 hours with rtPA (+37.2% relative to 10 minutes). At this time point, responses no longer differed significantly from baseline ( P =0.093), indicating a modest but functionally meaningful recovery despite marked interindividual variability. CONCLUSIONS: Awake thromboembolic stroke model with early functional ultrasound imaging completed with magnetic resonance imaging uncovers rapid blood flow perturbations and connectivity disruptions that are sensitive to rtPA and predictive of final lesion outcome. This platform enhances translational relevance by enabling hypothesis-testing of novel thrombolytics under physiologically intact conditions.