The tale of two COXs

In this issue of Blood, Dragani and colleagues provide evidence in ET for increased reticulated platelets and for a role of COX-2 in the enhanced thromboxane A₂ production. These findings highlight the need to rethink the optimum antithrombotic regimens in ET and other states with accelerated platelet generation.

A key feature of platelet biology in essential thrombocythemia (ET) patients is increased thromboxane production, as shown by enhanced excretion of the urinary metabolite of thromboxane A₂. These patients have an increased incidence of arterial thrombotic events. The most common therapy has been low-dose aspirin in addition to cytoreduction. The platelet mechanisms involved in the thrombotic events and the optimum therapeutic approaches remain unknown. The studies of Dragani et al shed new insights on these areas.¹ 

Cyclooxygenase-1 (COX-1) and COX-2 isozymes catalyze the conversion of arachidonic acid to prostaglandin (PG)H₂, the initial step in prostanoid synthesis leading to the formation of several products including thromboxane A₂, PGE₂, and PGI₂. Aspirin acetylates Ser529 in human COX-1 and Ser516 in human COX-2. When given once daily at low doses, aspirin almost completely inhibits COX-1 in platelets, which lack the ability to replenish the irreversibly inactivated enzyme. Much higher doses are needed to inhibit COX-2, an inducible enzyme; this is influenced also by the ability of nucleated cells to rapidly regenerate the enzyme.

Under physiologic conditions, platelet thromboxane production is primarily driven by COX-1 and endogenous PGI₂ by COX-2. Less than 10% of normal platelets have COX-2.²  Recent studies show that COX-2 is present in megakaryocytes, up-regulated during megakaryopoiesis, and expressed in young platelets.²  These findings have drawn attention to the potential contribution of COX-2 to thromboxane production in diverse clinical conditions. Dragani et al show convincingly that younger thiazole-staining reticulated platelets are increased, indicating enhanced platelet generation, and that platelet COX-2 expression is up-regulated in ET patients.¹  This is also described in other states associated with increased platelet generation.^2-4  Dragani et al show that aspirin at doses that almost completely suppress thromboxane production in healthy subjects is not as effective in ET. In an open-label randomized study in ET patients who are taking daily aspirin, the authors show that the addition of a COX-2 inhibitor to aspirin decreases by approximately 30% both urinary excretion of a thromboxane metabolite (marker of endogenous production) and ex vivo thromboxane production in serum (measure of total synthetic capacity). These studies corroborate that COX-2 contributes to endogenous thromboxane production in ET. Of note, the addition of the COX-2 inhibitor did not abolish thromboxane production, suggesting the role of additional factors.

The presence of increased numbers of reticulated platelets in ET draws attention to another major mechanism—namely, COX-1 activity in newly released platelets not inhibited by the once-daily low-dose aspirin. This is because of the relatively short half-life of aspirin in plasma. In states with rapid platelet generation, one daily dose of low-dose aspirin may be insufficient because of the immense capacity of bone marrow to accelerate platelet production (even 10-fold of basal rate). Dragani et al go on to show that addition of aspirin (50μM) in vitro abolishes thromboxane production in patients already receiving aspirin, consistent with an effect on COX-1 not acetylated by the single dose of aspirin administered 24 hours earlier. In a study of 60 healthy subjects, Guthikonda et al showed increased thromboxane production in healthy subjects with higher reticulated platelets, while ex vivo addition of COX-1 or COX-2 inhibitor attenuated the thromboxane production; the effect was greater with a COX-1 inhibitor.⁴  They also showed decreased inhibitory effects of aspirin and clopidogrel in patients with coronary artery disease with higher numbers of reticulated platelets.³ 

The findings of Dragani et al raise important issues regarding optimum regimen to suppress thromboxane production. What is clear is that low-dose aspirin administered once daily is inadequate to fully suppress thromboxane production in ET and likely in other disorders with increased platelet production. The addition of a COX-2 inhibitor may not be the answer, because of the persistent thromboxane production following a 7-day dual drug therapy with a COX-2 antagonist and aspirin. The short plasma half-life of aspirin argues against added benefits of higher doses at the same dosing interval. The studies of Dragani et al advance strong support for a dosing interval shorter than the traditional once-daily aspirin—and this remains to be tested with clinical endpoints of thrombotic events.