Piatelets and Vascular Occlusion

Selective inhibitors of thromboxane synthase have two theoretical advantages over inhibitors of the cyclooxygenase enzyme as potential antithrombotic compounds. First, they do not prevent formation of prostacyclin, a platelet-inhibitory, vasodilator compound, coincident with inhibiting thromboxane biosynthesis. Second, the prostaglandin endoperoxide substrate that accumulates in the platelet in the presence of thromboxane synthase inhibition may be donated to endothelial prostacyclin synthase at the site of platelet-vascular interactions (endoperoxide "steal"). Selective inhibition of thromboxane biosynthesis coincident with enhanced prostacyclin formation in vivo has been observed after administration of these compounds to man. Despite these attractive features and the efficacy of these compounds in diverse short-term animal preparations of thrombosis, investigations of their efflcacy in human disease have proven disappointing. This may reflect on the importance of thromboxane A2 in the diseases that have been investigated. Alternatively, the lack of drug efficacy may have resulted from either incomplete suppression of thromboxane biosynthesis and/or substitution for the biological effects of thromboxane A2 by prostaglandin endoperoxides during long-term dosing studies. Given that selective inhibition of thromboxane formation can be approached with aspirin, the particular value of these compounds is dependent on enhancing prostacyclin formation. Aspirin inhibits thromboxane-dependent platelet activation, but many platelet agonists are likely to act in concert in vivo and prostacyclin inhibits platelet aggregation induced by both thromboxane-dependent and thromboxaneindependent mechanisms. To test the hypothesis that thromboxane synthase inhibitors are efficacious in human disease, compounds of longer duration of action are required. Combination with antagonists of the prostaglandin/thromboxane A2 receptor may be necessary to reveal their full beneficial action. Circulation 72, No. 6, 1194-1201, 1985. ASPIRIN irreversibly acetylates platelet cyclooxygenase and thereby inhibits production of thromboxane A2 a potent vasoconstrictor and platelet agonist.'1 2 These observations provided a molecular basis for prospective studies of aspirin in syndromes putatively associated with platelet activation. These trials were initially mounted on the basis of aspirin's effects on platelet aggregation3 and retrospective data suggesting a reduced cardiovascular mortality in patients with arthritis who consumed aspirin over the long term.4 In prospective, double-blind, placebo-controlled studies aspirin has since been shown of benefit in the prevention of transient ischemic attack and stroke,5 6 the prevention of death and nonfatal myocardial infarction in patients with unstable angina,7' the prevention of hemodialysis shunt occlusion9 and, with dipyridamole, the prevention of coronary graft occlusionl' 11 and From the Division of Clinical Pharmacology, Vanderbilt University, Nashville. Supported by the National Institute of Health (HL30400). Dr. FitzGerald is an Established Investigator of the American Heart Association. Address for correspondence: Dr. G. A. FitzGerald, Division of Clinical Pharmacology, Vanderbilt University, Nashville, TN 37232. 1194 renal deterioration in patients with membranoproliferative glomerulonephritis.'2 The efficacy of aspirin alone in a recent study of coronary graft occlusion'3 and the similar efficacy of aspirin with or without dipyridamole in the prevention of transient ischemic attacks'4 and coronary graft occlusion'5 suggests that dipyridamole may have contributed little benefit in the last three studies cited.' "2 Residual doubts about the efficacy of aspirin remain, however, particularly fostered by a number of negative studies in the secondary prevention of myocardial infarction. 16-15 Many factors, including delay from the time of presentation to randomization and inadequate sample size may have contributed to these results. However, it has also been suggested that the dosage of aspirin used may have resulted in coincidental depression of formation of prostacyclin, the major cyclooxygenase product of vascular endothelium and a potent inhibitor of platelet aggregation. '9 Consequently, the development of compounds that selectively inhibit thromboxane synthase20, 21 prompted considerable interest. Unlike cyclooxygenase inhibitors, such agents would prevent formation of thromboxane A2 withCIRCULATION by gest on N ovem er 6, 2017 http://ciajournals.org/ D ow nladed from PLATELETS AND VASCULAR OCCLUSION out coincident inhibition of prostacyclin biosynthesis. Furthermore, they might actually enhance vascular prostacyclin formation if platelets could "donate" their accumulated prostaglandin endoperoxide substrate to endothelial prostacyclin synthase.22 Endoperoxide rediversion. A compound that enhanced vascular prostacyclin formation would be of substantial potential benefit as a platelet inhibitor in vivo. Aspirin-like drugs might be expected to prevent thromboxane-dependent platelet activation. However, many endogenous mediators of platelet activation, such as thrombin and collagen, could stimulate aggregation despite the presence of aspirin. Prostacyclin, by increasing platelet cyclic AMP, prevents platelet aggregation induced by all known agonists and has been shown to disaggregate aggregated platelets in vitro.23 Multiple agonists are likely to act in concert in vivo, so even in a syndrome of platelet-mediated vascular occlusion, inhibition of thromboxane formation alone would not be expected to be completely effective. Enhanced production of prostacyclin, particularly at the site of platelet-vessel wall interaction, as envisaged by the endoperoxide rediversion or "steal" hypothesis,24 might represent an important advantage over aspirin. Initial attempts to demonstrate endoperoxide rediversion were unsuccessful.25 However, Marcus et al.26 demonstrated that in a perturbed mixture of endothelial cells and platelets, approximately half of the prostacyclin produced by the endothelial cells originated from platelet endoperoxides. This phenomenon was amplified by thromboxane synthase inhibitors. These observations, which were subsequently confirmed by others,27 provide important experimental support for the concept of endoperoxide transfer. However, the model lacks important components such as flow, shear stress, and albumin (which avidly binds eicosenoids28), which operate in vivo. Early reports that were purported to demonstrate an increase in plasma 6-keto-prostaglandin (PG)Fi, after administration of a thromboxane synthase inhibitor to human volunteers29 were uninterpretable due to methodologic shortcomings.30 More commonly, inhibitors were screened for evidence of endoperoxide rediversion by measurement of immunoreactive thromboxane B2 and 6-keto-PGFI, formation in serum in vitro or ex vivo.31 32 However, although 6-keto-PGF,1 is formed in small amounts in whole blood,33 immunoreactive 6-keto-PGFi, may greatly overestimate actual 6-keto-PGFIa formation in serum in the presence of a thromboxane synthase inhibitor.34 PGE2, PGF21, and PGD2 are formed in much greater amounts than 6-keto-PGF,a in serum in the presence of a synthase inhibitor.33 Thus, although cross reactivity of a 6-keto-PGF,a antibody with these compounds may be small in percentage terms, increased PGE2, PGF2a, or PGD2 production would be likely to introduce a substantial quantitative artifact into estimates of 6-keto-PGFla production in the presence of a synthase inhibitor. Confirmation of this hypothesis was obtained by measurement of both thromboxane B2 and 6-keto-PGFla by radioimmunoassay and gas chromatography-mass spectrometry in serum with and without addition of the thromboxane synthase inhibitor dazoxiben.34 An alternative approach to the detection of enhanced prostacyclin formation after the administration of these compounds has been the measurement of 2,3-dinor-6keto-PGFla (PGI-M), a major urinary prostacyclin metabolite in man.35 We have conducted three doubleblind, short-term studies with structurally distinct thromboxane synthase inhibitors .3638 PGI-M excretion increased approximately twoto threefold after two of these compounds, dazoxiben and CGS 13080, while the increase after dazmagrel was less marked. Assuming that the "basal" level of 6-keto-PGFIa, is 2 to 3 pg/ml or less,39, 40 a threefold increment would result in levels that are still below the limits of sensitivity of most assay systems. This is consistent with the marginal increase in plasma 6-keto-PGFi reported by Patrignani et al.4' after administration of dazoxiben to volunteers. Obviously, such an increment would be insufficient to result in circulating concentrations of prostacyclin of biological significance. However, it might reflect production at sites of platelet-vessel wall interaction that could be relevant to local thromboresistance. It is also of interest that in patients with severe peripheral vascular disease, prostacyclin metabolite excretion is increased threeto fourfold.42 Finally, although these data are consistent with the vascular steal hypothesis, it is important to remember that one cannot definitively attribute a tissue of origin to metabolites measured in urine. It is possible that the increment in PGI-M reflects its production by a tissue other than vascular endothelium. In summary, information has been provided in constrained circumstances in vitro that demonstrates enhanced transfer of platelet endoperoxides to vascular prostacyclin synthase in the presence of thromboxane synthase inhibitors. Data compatible with this mechanism has been obtained after short-term administration of thromboxane synthase inhibitors in vivo. "Nonresponse" to thromboxane synthase inhibition. Bertele et al.43 reported interindividual differences in the platelet-inhibitory response to thromboxane synthase inhibitors in vitro. Despite efficient inhibition of 1195 Vol. 72, No. 6, December 1985 by gest on N ovem er 6, 2017 http://ciajournals.org/ D ow nladed from