This paper reviews the importance of blood flow phenomena in models of
experimental thrombosis used for measuring antithrombotic drug effica
cy. The characteristics of these systems and their application for stu
dies with human blood and in animal models are considered. Central to
these investigations has been the development of various types of perf
usion chambers in which a thrombogenic test surface is exposed to flow
ing blood under well-defined conditions of blood flow and device geome
try. Such perfusion chambers, which have been used in vitro, ex vivo,
and in vivo by insertion into arteriovenous shunts in various animal s
pecies, have allowed reproducible testing of both conventional and exp
erimental agents. Sheer-dependent antithrombotic effects have been obs
erved with anticoagulants such as heparin and with selective inhibitor
s of thrombin, Factor Xa, and factor VIIa. However, the degree of shee
r dependency depends on the chemical composition of the thrombogenic s
urface; for example, anticoagulant effects may be more pronounced on a
tissue factor-rich surface than on a collagen-rich surface, particula
rly at venous or low arterial sheer rates. Platelet inhibitors such as
aspirin, thromboxane antagonists, or inhibitors of van Willebrand fac
tor platelet interactions are also shear dependent, being more efficie
nt at high shear rates. In contrast, inhibitors of adenosine diphospha
te-dependent platelet reactions or antagonists of the platelet membran
e glycoprotein IIb/IIIa complex ore sheer rate independent. At very hi
gh sheer rates characteristic of severely stenotic arteries, aspirin l
oses its direct antithrombotic effect, whereas adenosine diphosphate p
athway inhibitors and GP IIb/IIIa antagonists are still beneficial. In
general, results obtained with many of these models have predicted an
tithrombotic efficacy in human beings when comparisons were possible.
Therefore shear-dependent models of experimental thrombosis are routin
ely used in the evaluation of antithrombotic pharmacologic agents, bot
h preclinically and clinically.