Cocaine use is associated with sudden cardiac death, cardiac ischemia, and stroke in patients with no additional risk factors and is a frequent cause of these syndromes in patients with sickle cell disease. Pathologic findings include platelet-rich microthrombi and an increased plasma concentration of von Willebrand Factor (VWF). These findings suggest that, in addition to the well-known cocaine effects of vasoconstriction and increased tissue oxygen demand, activation of platelets and/or endothelial cells contributes to cocaine-induced ischemia. However, studies investigating the effect of cocaine on platelet functions have been inconclusive, finding both platelet activation and inhibition depending on the assay used. Further, the ability of cocaine to activate the vascular endothelium has not been examined, in particular, the endothelial secretion of the most adhesive forms of VWF, the ultralarge forms (ULVWF). ULVWF are long VWF multimers that remain tethered to the endothelial surface upon secretion, extend into the blood vessel lumen under laminar flow in long strings measuring up to 0.5 cm in length, and have multiple exposed binding sites for receptors on platelets, erythrocytes, and leukocytes. Elevated levels of ULVWF, due either to enhanced secretion or defective processing, have been implicated in diseases such as thrombotic thrombocytopenic purpura (TTP) and sickle cell anemia. We hypothesized that a major consequence of cocaine exposure is activation of the vascular endothelium to secrete ULVWF, which would provide a platform for blood cell adhesion and subsequent thrombosis or vaso-occlusion. We evaluated the ability of cocaine to stimulate ULVWF from cultured endothelial cells in a parallel-plate flow chamber assay and found that 1 μg/ml cocaine, a level comparable to peak blood levels detectable in cocaine abusers, efficiently induced secretion of ULVWF capable of binding platelets under flow conditions similar to that induced by histamine 6 μg/ml (3.92 ULVWF strings/field with cocaine vs. 4.62 strings/field with histamine). We also assessed the activation of platelets exposed in vitro to cocaine by flow cytometry, using two markers of platelet activation: P-selectin expression (which signals a-granule release) and conformational activation of the platelet integrin aIIbb3, detected with the antibody PAC-1. We found that when platelet-rich plasma was incubated with cocaine at concentrations from 0.1 μg/ml to 10 μg/ml, there was no increase in P- selectin exposure or PAC-1 binding. Furthermore, pretreatment of platelets with cocaine inhibited the ability of platelets to subsequently be activated by ADP in a dose-dependent manner. We did not observe any increase in mean fluorescence above background in ADP stimulated platelets pre-incubated with 1 μg/ml cocaine for P-selectin or PAC-1 binding. However, exposure of platelets in whole blood to 1 μg/ml cocaine resulted in a 3.2-fold increase in P- selectin exposure and a 5.4-fold increase in PAC-1 binding. These results indicate that cocaine directly activates the vascular endothelium to secrete ULVWF, and activates platelets indirectly, involving as yet unknown factors in whole blood, resulting in the formation of microthrombi. These effects of cocaine are likely to have pathogenic roles in cardiovascular syndromes associated with cocaine use, including the triggering of vaso-occlusive crises in sickle cell anemia and may explain the observed association of cocaine use with TTP.

Disclosure: No relevant conflicts of interest to declare.