The role of the endothelium in arterial thrombosis and the influence of antithrombotic therapy

The nonthrombogenicity of vascular endothelium, a single layer of cells separating blood‐borne platelets and thrombogenic components of the underlying subendothelium, results from more than its provision of a physical barrier. These cells synthesize and secrete prostacyclin, plasminogen activator, and inhibitors of the coagulation system. They also catabolize proaggregatory agents such as ADP and serotonin (5‐HT). Damage to or removal of the endothelium is, of course, a potent thrombogenic stimulus. We have employed an acute canine model of coronary arterial thrombosis, which combines vessel narrowing and intimal injury as stimuli for repetitive, regular occlusive thrombus formation. Of all agents tested by us and others in this model, prostacyclin, selective thromboxane (TXA 2 ) synthetase inhibitors, and 5‐HT antagonists consistently inhibit thrombus formation. Available data suggest that the antithrombotic actions of TXA 2 synthetase inhibitors result in part from vascular metabolism of platelet‐derived endoperoxides, as well as endogenous PGI 2 production. Serotonin, alone, exerts weak proaggregatory effects. However, it uniquely amplifies the proaggregatory effects of many physiologic proaggregatory mediators. Serotonin 2 antagonists are more potent and consistently antithrombotic in this canine model than TXA 2 synthetase inhibitors. The discordance between the marked antithrombotic effects of 5‐ht antagonists and txa 2 synthetase inhibitors in vivo and their weak antiaggregatory and proaggregatory effects in vitro, respectively, underscores the danger of relying exclusively upon in vitro platelet aggregatory behavior to assess either the potential importance of a mediator or antithrombotic efficacy of a new drug. On the basis of results obtained from our laboratory and others that employ similar in vivo models of thrombosis, we conclude that (1) since platelet aggregation in vivo is stimulated by many physiologic agonists, in vivo and in vitro, models of aggregation and thrombosis should reflect this; (2) the most potent or efficacious proaggregatory agent in vitro may not necessarily be the most important mediator in vivo and vice versa. Future development of antithrombotic agents should continue; aspirin can be improved upon in several respects. Perhaps future antithrombotic agents will be targeted to specific platelet membrane (g1yco)proteins involved in platelet‐platelet )aggre‐ gation) and platelet‐vessel wall (adhesion) interactions.