Plasma VWF is sufficient for initial adhesion of platelets to collagen, whereas platelet VWF released locally is necessary for forming subsequent occlusive thrombi. Representations of WT and Nbeal2^−/− platelet adhesion and subsequent occlusion during the 8-minute perfusion period. The progressive accumulation of platelets is observed over time. (A) WT mice blood showed rapid and significant platelet adhesion and aggregation leading to occlusion. (B) Nbeal2^−/− mice blood did not form occlusive thrombi during the 8-minute observation period. Occlusive thrombi formed after the addition of WT whole blood (Rescue) to Nbeal2^−/− platelets adherent to collagen. The 1-mm stenosis region is marked with a red line under the image. The ROI appears blank before the adhesion of platelets (green; 3,3′-dihexyloxacarbocyanine iodide [DiOC6]). Flow direction is from left to right. The bar represents 1000 μm. (C) Normalized fluorescent intensities comparing thrombus growth and occlusion in control WT (n = 6), Nbeal2^−/− (n = 6), and Rescue (n = 3) channels. Initial lag phase quickly proceeded to RPA, demonstrated by a rapid increase in thrombus size and intensity at ∼100 seconds for WT whole blood and at 515 ± 47 seconds for Nbeal2^−/−+WT whole blood. Occlusion occurred at the end of the plot for WT and Nbeal2^−/−+WT whole blood. No occlusion was observed with Nbeal2^−/− whole blood alone. The arrow denotes the time point when WT whole blood was added to Nbeal2^−/− channels. All images were postprocessed and averaged to compare the intensity levels in the test sections. (D) Occlusion times in the microfluidic device for WT (326 ± 25 seconds), Nbeal2^−/− (no occlusion), and Nbeal2^−/−+WT (166 ± 48 seconds). All 3 occlusion times were significantly different from each other. ***P < .001.