Microfluidic Modeling of Thrombolysis

Abstract

Objective— Despite the high clinical relevance of thrombolysis, models for its study in human flowing blood are lacking. Our objective was to develop a microfluidic model for comparative evaluation of thrombolytic therapeutic strategies. Approach and Results— Citrated human blood was supplemented with 3,3′-dihexyloxacarbocyanine iodide and Alexa Fluor 647 fibrinogen conjugate, recalcified, and perfused for 3 to 4 minutes at venous or arterial wall shear rate in microfluidic flow chambers coated with collagen and tissue factor to generate nonocclusive fluorescent thrombi. A second perfusion was performed for 10 minutes with rhodamine-6G-labeled citrated whole blood, supplemented or not with r-tPA (recombinant tissue-type plasminogen activator), fluorescein isothiocyanate-conjugated r-tPA, and Alexa Fluor 568 plasminogen conjugate. Plasminogen and r-tPA bound to preformed thrombi and r-tPA caused a concentration-dependent decrease in thrombus fibrin content (up to 50% reduction at 15 µg/mL r-tPA) as assessed by fluorescence microscopy. Fibrinolysis was confirmed by measurement of D-dimers in the output flow. Remarkably, despite ongoing fibrinolysis, new platelets continued to be recruited to the thrombus under lysis. Under the arterial condition, combining r-tPA with hirudin enhanced fibrinolysis but did not prevent the recruitment of new platelets, which was, however, prevented by antiplatelet agents (ticagrelor or the GPVI [glycoprotein VI]-blocking antigen-binding fragment 9O12). Conclusions— Our microfluidic thrombolysis model is suitable for studying thrombolysis and testing the efficacy of drugs used in combination with r-tPA. Real-time analysis of fibrin and platelets during r-tPA-mediated fibrinolysis at arterial or venous flow conditions showed that platelets continue to accumulate during fibrinolysis. Such platelet accumulation may impair r-tPA-mediated recanalization.