Abstract 3788: Mechanical Thrombectomy in a Model System of Cerebrovascular Occlusion: Device-Embolus Interactions, Flow Restoration, and Risk of Distal Embolization

Abstract

Background: Previous clinical studies have shown that the performance of current FDA-cleared mechanical endovascular thrombectomy (MET) devices still can be improved. We hypothesize the efficacy and safety of MET are in part affected by proximal flow arrest, mechanical properties of the embolus, device-embolus interactions, geometry of the vasculature, cerebral hemodynamics, and MET mechanism. To attest our hypothesis, a model system of cerebrovascular occlusion is built for evaluation of different MET technologies. Materials: The proposed model system was composed of an anatomically representative internal carotid artery/ middle cerebral artery silicone replica, a clinically validated embolus analogue (EA), and physiological hemodynamic conditions. Different MET mechanisms were employed under X-ray guidance, namely, mechanical retrieval, aspiration and ultrasonic cavitation. The self-expanding stents (Solitaire FR and Enterprise stent) were partially deployed to ensnare the EA, followed by EA retrieval. Balloon guide catheter was used to temporarily arrest the antegrade flow when MERCI retriever or Solitaire FR was used. The efficacy endpoints included the amount of blood flow restored and the ability to achieve recanalization. The safety endpoints were an analysis of EA fragments and their size distribution by Coulter principle. A minimum sample size for each device was 5. Results: Both the MERCI retriever and self-expanding stents were able to engage the EA after deployment. However, in some cases, it was observed that the EA was not retained in the stent (or retriever) during retrieval. The composite flow restoration (recanalization rate × flow restoration rate x 100%) was: MERCI retriever 67%, Solitaire FR 92%, Penumbra aspiration system 77%, Enterprise stent 15%, and the ultrasound waveguide 0%. More large (>200μ m) EA fragments were generated during Penumbra aspiration. No EA fragments had dimension greater than 1000μ m, and mean size of the small and large EA fragments were between 23 to 37 and 215 to 285 μ m, respectively. Conclusion: The model system of cerebrovascular occlusion allows visualization of device-EA interactions and quantitative measurement of flow restoration. It also offers a reproducible testing environment to evaluate the existing devices and/or newly developed prototypes. The recanalization rate of the retriever device is related to the ability of the device to capture the EA and temporary blockage of antegrade flow during device removal. The tortuosity of the cerebrovasculature may impede the ultrasonic wave propagation and in part accounts for the failure to restore flow. The risk of the embolic shower is influenced by the mechanism of action for the MET device.