Antiplatelet DrugsAntiplatelet Drugs: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines FREE TO VIEW

Platelets are vital components of normal hemostasis and key participants in atherothrombosis by virtue of their capacity to adhere to the injured blood vessel wall; recruit additional platelets to the site of injury; release vasoactive and prothrombotic mediators that trigger vasoconstriction and promote coagulation, respectively; and form aggregates that affect primary hemostasis.¹ Although platelet adhesion, activation, and aggregation can be viewed as a physiologic repair response to the sudden fissuring or rupture of an atherosclerotic plaque, uncontrolled progression of such a process through a series of self-sustaining amplification loops can lead to intraluminal thrombus formation, vascular occlusion, and subsequent ischemia or infarction. Currently available antiplatelet drugs interfere with one or more steps in the process of platelet release and aggregation² and produce a measurable reduction in the risk of thrombosis that cannot be dissociated from an increased risk of bleeding.³

When considering antiplatelet drugs, it is important to appreciate that ∼10¹¹ platelets are produced each day under physiologic circumstances, a level of production that can increase up to 10-fold at times of increased need.⁴ Platelets are anucleated blood cells that form by fragmentation of bone marrow megakaryocyte cytoplasm and have a maximum circulating life span of ∼10 days. Regulation of platelet production is mediated by thrombopoietin, which is produced primarily in the liver as well as in the bone marrow and the kidney and cleared by binding to high-affinity receptors on platelets and megakaryocytes.⁵ In the presence of a high-platelet mass, thrombopoietin levels are reduced, and platelet production falls, whereas in the presence of a low-platelet mass, thrombopoietin levels rise, thereby stimulating thrombopoiesis. Platelets provide a circulating source of chemokines, cytokines, and growth factors, which are preformed and packaged in storage granules. Activated platelets can synthesize prostanoids, primarily thromboxane A₂ (TXA₂), from arachidonic acid released from membrane phospholipids through rapid coordinated activation of phospholipases, cyclooxygenase (COX)-1, and TX synthase (Fig 1). The inducible form of COX (COX-2) not only is found primarily in the vascular endothelium and in monocytes but is also expressed in newly formed platelets, particularly in the setting of accelerated platelet production.⁶ Although activated platelets are incapable of de novo protein synthesis, they can translate constitutive mRNA into protein over the course of several hours.⁷ Thus, platelets may play a part in inflammation, angiogenesis, and wound healing, and antiplatelet therapies may have an impact on these processes by blocking platelet-derived protein signals for inflammatory or proliferative responses.

Negative modulation of platelet adhesion and aggregation is exerted by a variety of physiologic mechanisms, including endothelium-derived prostacyclin (PGI₂), nitric oxide, CD39/ecto-ADPase, and platelet endothelial cell adhesion molecule-1. Some drugs may interfere with these regulatory pathways, as exemplified by the dose-dependent inhibition of PGI₂ production by aspirin and other COX-1 and COX-2 inhibitors.³

The article on antiplatelet therapy in the American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition) reviewed the antiplatelet effects of traditional nonsteroidal antiinflammatory drugs (NSAIDs) and the cardiovascular effects of COX-2-selective NSAIDs. This topic will not be covered here, and interested readers are referred to the previous article.⁸

Aspirin is the most widely studied antiplatelet drug. On the basis of > 100 randomized trials in high-risk patients, aspirin reduces vascular death by ∼15% and nonfatal vascular events by ∼30%.^9,10

The best-characterized mechanism of action of aspirin is related to its capacity to permanently inhibit the COX activity of prostaglandin H-synthase-1 and prostaglandin H-synthase-2 (also referred to as COX-1 and COX-2, respectively).^11-15 COX isozymes catalyze the first committed step in prostanoid biosynthesis, the conversion of arachidonic acid to prostaglandin H₂ (PGH₂) (Fig 1). PGH₂ is the immediate precursor of TXA₂ and PGI₂.

The molecular mechanism of permanent inhibition of COX activity by aspirin is related to blockade of the COX channel as a consequence of acetylation of a strategically located serine residue (Ser529 in COX-1, Ser516 in COX-2), thereby preventing substrate access to the catalytic site of the enzyme.¹⁶ Complete or near-complete inhibition of platelet COX-1 can be achieved with low doses of aspirin (75-150 mg) given once daily. In contrast, inhibition of COX-2-dependent pathophysiologic processes (eg, hyperalgesia, inflammation) requires larger doses of aspirin and a much shorter dosing interval because nucleated cells rapidly resynthesize the enzyme. Thus, 10- to 100-fold higher daily doses of aspirin are required when the drug is used as an antiinflammatory agent rather than as an antiplatelet agent. The benefit/risk profile of aspirin depends on dose because its GI toxicity is dose dependent (discussed in section 1.3 “The Optimal Dose of Aspirin”).

Human platelets and vascular endothelial cells process PGH₂ to produce primarily TXA₂ and PGI₂, respectively.¹² TXA₂ induces platelet aggregation and vasoconstriction, whereas PGI₂ inhibits platelet aggregation and induces vasodilation.¹² Because TXA₂ is largely derived from COX-1 (mostly from platelets), it is highly sensitive to aspirin inhibition. In contrast, although vascular PGI₂ can be derived from COX-1, its major source is COX-2, even under physiologic conditions.¹⁷ COX-1-dependent PGI₂ production occurs transiently in response to agonist stimulation (eg, bradykinin)¹⁸ and is sensitive to aspirin inhibition. COX-2-mediated PGI₂ production occurs long term in response to laminar shear stress¹⁹ and is relatively insensitive to low-dose aspirin, which may explain the substantial residual COX-2-dependent PGI₂ biosynthesis that occurs with daily doses of aspirin in the range of 30 to 100 mg,²⁰ despite transient suppression of COX-1-dependent PGI₂ release.¹⁸ It is not established that the greater suppression of PGI₂ formation produced by higher doses of aspirin is sufficient to initiate or predispose to thrombosis. However, two lines of evidence suggest that PGI₂ is thromboprotective. First, mice lacking the PGI₂ receptor exhibit increased susceptibility to injury-induced thrombosis.²¹ Second, the cardiovascular toxicity associated with COX-2 inhibitors²² also supports the concept that PGI₂ is important for thromboresistance in the setting of inadequate inhibition of platelet TXA₂ biosynthesis.²³

Aspirin is rapidly absorbed in the stomach and upper intestine. Plasma levels peak 30 to 40 min after aspirin ingestion, and inhibition of platelet function is evident within 1 h. In contrast, it can take 3 to 4 h to reach peak plasma levels after administration of enteric-coated aspirin. Therefore, if a rapid effect is required and only enteric-coated tablets are available, the tablets should be chewed instead of swallowed intact. The oral bioavailability of regular aspirin tablets is ∼40% to 50% over a wide range of doses.²⁴ A considerably lower bioavailability has been reported for enteric-coated tablets and for sustained-release, microencapsulated preparations.²⁴ The lower bioavailability of some enteric-coated preparations and their poor absorption from the higher pH environment of the small intestine may result in inadequate platelet inhibition when these preparations are used at low doses, particularly in heavier subjects.²⁵ Both a controlled-release formulation¹⁸ and a transdermal patch²⁶ with negligible systemic bioavailability have been developed in an attempt to achieve selective inhibition of platelet TXA₂ production without suppressing systemic PGI₂ synthesis. The former was used successfully in the Thrombosis Prevention Trial,²⁷ but it remains unknown whether the controlled-release formulation has any advantages over plain aspirin.

The plasma concentration of aspirin decays with a half-life of 15 to 20 min. Despite the rapid clearance of aspirin from the circulation, the platelet-inhibitory effects last the life span of the platelet²⁸ because aspirin irreversibly inactivates platelet COX-1.^13,14 Aspirin also acetylates megakaryocyte COX-1, thereby inhibiting thromboxane production in newly released platelets as well as those already in the circulation.^11,29-31 The mean life span of human platelets is ∼10 days, which means that ∼10% to 12% of circulating platelets are replaced every day.^32,33 The recovery of thromboxane production and thromboxane-dependent platelet aggregation after prolonged aspirin treatment is stopped is faster than what would be predicted based on the rate of platelet turnover²⁰ possibly because of the nonlinear relationship between inhibition of platelet COX-1 activity and TXA₂ biosynthesis^34,35 (Fig 2) and the capacity of small amounts of thromboxane produced by nonaspirinated platelets to sustain thromboxane-dependent platelet aggregation.

Placebo-controlled randomized trials have shown that aspirin is an effective antithrombotic agent when used long term in doses ranging between 50 and 100 mg/d, and there are data to suggest that it is effective in doses as low as 30 mg/d.^9,10 At a dose of 75 mg/d, aspirin was shown to (1) reduce the risk of acute myocardial infarction (MI) or death in patients with unstable angina³⁶ or chronic stable angina,³⁷ (2) reduce the risk of stroke or death in patients with transient cerebral ischemia,³⁸ and (3) reduce the risk of stroke after carotid endarterectomy.³⁹ In the European Stroke Prevention Study (ESPS)-2, aspirin 25 mg bid reduced the risk of stroke and the composite outcome of stroke or death in patients with prior stroke or transient ischemic attack (TIA).⁴⁰ In the European Collaboration on Low-dose Aspirin in Polycythemia vera trial,⁴¹ aspirin (100 mg/d) was effective in preventing thrombotic complications in patients with polycythemia vera, even in the face of higher-than-normal platelet counts. The lowest doses of aspirin demonstrated to be effective for these various indications are shown in Table 1.

The clinical effects of different doses of aspirin have been directly compared in randomized controlled trials.^42-48 In the United Kingdom-Transient Ischaemic Attack (UK-TIA) study that randomized 2,435 patients after a TIA or minor ischemic stroke to receive one of two doses of aspirin or placebo, aspirin doses of 300 and 1,200 mg/d were associated with a similar rate of MI, major stroke, or vascular death (20% and 20%, respectively; OR, 1.03; 95% CI, 0.83-1.29).⁴² In the Dutch TIA trial, which randomized 3,131 patients after a TIA or minor ischemic stroke, aspirin doses of 30 and 283 mg/d were associated with a similar rate of MI, stroke, or cardiovascular death (14.7% and 15.2%, respectively; hazard ratio [HR], 0.91; 95% CI, 0.76-1.09).⁴⁷ The ASA and Carotid Endarterectomy (ACE) trial reported that the risk of stroke, MI, or death within 3 months of carotid endarterectomy was significantly lower for patients taking aspirin 81 or 325 mg/d than for those taking 650 or 1,300 mg/d (6.2% vs 8.4%, P = .03).⁴⁶ The Clopidogrel Optimal Loading Dose Usage to Reduce Recurrent Events/Organization to Assess Strategies for Ischemic Syndromes (CURRRENT-OASIS 7) trial, which included 25,086 patients with acute coronary syndromes (ACSs), found that 30 days of treatment with aspirin 300 to 325 mg/d was no more effective than aspirin 75 to 100 mg/d for the prevention of stroke, MI, or cardiovascular death (4.2% and 4.4%, respectively; HR, 0.97; 95% CI, 0.86-1.09).⁴⁸ Thus, on the basis of results from randomized studies comparing different doses of aspirin, there is no convincing evidence that higher doses are any more effective at reducing the risk of serious vascular events than lower doses. In fact, the indirect comparisons reported in the overview of the Antithrombotic Trialists’ Collaboration (Table 2) and the results of several direct randomized comparisons are compatible with the reverse; that is, there is blunting of the antithrombotic effect with higher doses of aspirin, a finding consistent with dose-dependent inhibition of PGI₂.

The antithrombotic effects of a range of doses of aspirin have been compared with an untreated control group in a number of thrombotic vascular disorders. The aspirin doses have ranged from 50 to 1,500 mg/d. Aspirin has been shown to be effective in the following conditions: unstable angina where the incidence of acute MI or death was significantly reduced in four separate studies using daily doses of 75,³⁶ 325,⁴⁹ 650,⁵⁰ or 1,300 mg⁵¹; stable angina where a dose of 75 mg/d reduced the incidence of acute MI or sudden death³⁷; aortocoronary bypass surgery where the incidence of early graft occlusion was similarly reduced with daily doses of 100,⁵² 325,⁵³ 975,⁵⁴ or 1,200 mg⁵⁴; thromboprophylaxis in patients with prosthetic heart valves who also received warfarin where the incidence of systemic embolism was reduced with daily doses of 100,⁵⁵ 500,⁵⁶ or 1,500 mg^57,58; thromboprophylaxis in long-term hemodialysis patients with arterial venous shunts where a dose of 160 mg/d was shown to be effective⁵⁹; acute MI in which a dose of 162.5 mg/d reduced 35-day mortality as well as nonfatal reinfarction and stroke⁶⁰; transient cerebral ischemia in which doses between 50 and 1,200 mg/d were effective^{38,40,42,61-63}; acute ischemic stroke where doses of 160 to 300 mg/d were effective in reducing early mortality and stroke recurrence^64,65; and polycythemia vera in which 100 mg/d,⁴¹ but not 900 mg/d,⁶⁶ was effective in reducing fatal and nonfatal vascular events. Thus, aspirin is an effective antithrombotic agent at doses between 50 and 1,500 mg/d. Based on the results of the Dutch TIA study, it is also possible that 30 mg/d is effective.⁴⁷

There is evidence that the GI side effects of aspirin are dose dependent. Thus, aspirin doses of ∼300 mg/d are associated with fewer GI side effects than doses of ∼1,200 mg/d.⁴² There is also evidence that aspirin doses ≤ 100 mg/d are associated with fewer side effects than 300 mg/d.⁴⁷ In an observational analysis in patients with ACS, the Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) investigators demonstrated that aspirin ≤ 100 mg/d alone or in combination with clopidogrel was associated with lower rates of major or life-threatening bleeding complications than aspirin alone at a dose of ≥ 200 mg/d.⁶⁷ In the randomized CURRENT-OASIS 7 trial, aspirin given at a dose of 75 to 100 mg/d produced less GI bleeding than a dose of 300 to 325 mg/d. The incidence of other types of major bleeding was not different between the two groups.⁴⁸ In summary, the lack of a dose-response relationship for the efficacy of aspirin in clinical studies and the dose dependence of GI bleeding support the use of lowest proven effective doses of aspirin as the most appropriate strategy to maximize efficacy and minimize toxicity (Table 1).⁹

High platelet reactivity in patients prescribed aspirin has been associated with an increased risk of thrombotic events. Therefore, this phenomenon has been used as one definition of aspirin resistance. Observational studies demonstrate that about one-third of patients treated with aspirin demonstrate less-than-expected inhibition of agonist-induced platelet aggregation and increased levels of urinary thromboxane.^8-10 Estimates of the prevalence of high on-treatment platelet reactivity are affected by differences among the studies in patient characteristics (eg, age, female sex, diabetes), concomitant therapies (particularly NSAIDs, eg, ibuprofen), the laboratory test used to measure the antiplatelet effects of aspirin (light transmission or whole-blood aggregometry, shear stress-induced platelet activation, expression of activation markers on the platelet surface as measured by flow cytometry, inhibition of thromboxane production), the cutoff used to define high on-treatment reactivity, and patient compliance with aspirin therapy.²⁵ Despite these differences, high platelet reactivity in patients prescribed aspirin has been consistently associated with a twofold to fourfold higher risk of MI, stroke, or death.^9,10,26,27

If thrombotic events in aspirin-treated patients with high on-treatment platelet reactivity were solely attributable to reduced responsiveness to aspirin, strategies aimed at improving the response would be expected to reduce this risk. Observational studies suggest that aspirin inhibits platelet function^68-72 and coagulation^73-79 in a dose-dependent manner; a finding confirmed in a randomized dose comparison.⁸⁰ Thus, in a randomized, double-blind, crossover trial that included 125 patients with stable coronary artery disease, Gurbel and colleagues⁸⁰ demonstrated that at doses of 81, 162, or 325 mg/d, aspirin inhibited adenosine diphosphate (ADP) and collagen-induced platelet aggregation, blocked shear-dependent platelet aggregation measured by the PFA-100 device, and reduced urinary thromboxane concentrations in a dose-dependent manner. Most of the patients included in this study demonstrated near-complete inhibition of arachidonic acid-induced platelet aggregation with the 81-mg/d dose of aspirin, but higher aspirin doses reduced the proportion of patients in whom arachidonic acid-induced aggregation exceeded a cutoff value of 20%.⁸⁰ However, the dose-dependent inhibition of platelet function and thromboxane production by aspirin observed in this study does not fit with the results of the CURRENT-OASIS 7 trial, which failed to demonstrate a reduction in the risk of thrombotic events with higher doses of aspirin (See section 1.3 “The Optimal Dose of Aspirin”). How do we explain this apparent paradox?

The most likely explanation is that the relationship between platelet reactivity and thrombotic risk is confounded by comorbidities, such as smoking or diabetes, that affect both platelet function and cardiovascular risk. It is also possible that the laboratory tests used to measure platelet reactivity fail to monitor the mechanism by which aspirin reduces the risk of thrombotic events. Many of these tests use nonphysiologic stimuli to induce platelet aggregation, and none assess platelet interaction with the vessel wall or the effect of aspirin on COX-2-dependent PGI₂ production. Thromboxane production appears to be the most specific measure of the inhibitory effects of aspirin because thromboxane is the major biochemical end product of the platelet COX-1 biosynthetic pathway that is targeted by aspirin. However, serum thromboxane levels reflect the maximum capacity of platelets to produce thromboxane; urinary thromboxane is also produced from nonplatelet sources, and measures of thromboxane concentration do not capture the effect of aspirin on PGI₂ production.

Given the multifactorial triggers of atherothrombosis and the likelihood that platelet activation and subsequent aggregation are not the sole mediators of vascular events, it is not surprising that only a fraction (usually one-fourth to one-third) of all vascular complications can be prevented by aspirin alone. There is no evidence that patients who experience a thrombotic event despite aspirin therapy benefit from treatment with higher-dose aspirin. Concomitant administration of nonselective NSAIDs, such as ibuprofen, should be avoided because, as outlined previously, these drugs can interfere with the antiplatelet effect of aspirin.¹⁶ A pharmacodynamic interaction between naproxen and aspirin has also been described,⁸¹ but this does not appear to occur with rofecoxib,⁸² celecoxib,⁸³ or diclofenac,⁸² drugs endowed with variable COX-2 selectivity.⁸⁴ The US Food and Drug Administration (FDA) has issued a statement informing patients and health-care professionals that ibuprofen can interfere with the antiplatelet effect of low-dose aspirin (81 mg/d), potentially rendering aspirin less effective when used for cardioprotection or stroke prevention.⁸⁵

The efficacy and safety of aspirin are documented from analyses of > 100 randomized controlled trials that have included thousands of patients representing the entire spectrum of atherosclerosis, ranging from apparently healthy low-risk individuals to patients presenting with an acute MI or acute ischemic stroke. Trials have evaluated aspirin therapy of only a few weeks duration or as long as 10 years.^9,10 Although aspirin has consistently been shown to be effective in preventing fatal and nonfatal vascular events in these trials, the absolute benefits depend on the clinical setting.

In Second International Study of Infarct Survival (ISIS-2),⁵⁹ a single 162.5-mg tablet of aspirin started within 24 h of the onset of symptoms of a suspected MI and continued at the same dose daily for 5 weeks produced highly significant reductions in vascular mortality, nonfatal reinfarction, and nonfatal stroke (23%, 49%, and 46%, respectively). There was no associated increase in hemorrhagic stroke or GI bleeding with aspirin, although there was a small increase in minor bleeding.⁵⁹ Based on the results of this study, a 5-week course of aspirin treatment in 1,000 patients with suspected acute MI will prevent ∼40 vascular events,¹⁰ a proportional odds reduction of 30%.

Two separate trials with a similar protocol tested the efficacy and safety of early aspirin use in acute ischemic stroke. The Chinese Acute Stroke Trial⁶⁴ and the International Stroke Trial⁶⁵ collectively randomized ∼40,000 patients within 48 h of the onset of stroke symptoms to 2 to 4 weeks of daily aspirin therapy (at doses of 160 and 300 mg/d, respectively) or to placebo. An overview analysis of the results of both trials indicated an absolute benefit of nine fewer deaths or nonfatal strokes per 1,000 patients in the first month of aspirin therapy.¹⁰ The proportional odds reduction in the risk of fatal or nonfatal vascular events was only 10% in this setting. Although the background risk of hemorrhagic stroke was threefold higher in the Chinese Acute Stroke Trial than in the International Stroke Trial, the absolute increase in this risk was similar in the two studies (an excess of 2/1,000 aspirin-treated patients).^64,65

Long-term aspirin therapy confers a conclusive net benefit on the risk of subsequent MI, stroke, or vascular death among subjects with a high risk of vascular complications. These include patients with chronic stable angina,³⁷ prior MI,¹⁰ unstable angina,^36,49-51 history of TIA or minor stroke,^{37,40,42,62,63,86} and other high-risk categories.¹⁰ The proportional reduction in vascular events with long-term aspirin therapy in these various clinical settings ranges from 20% to 25% based on an overview of all of the randomized trials.¹⁰ Estimates of relative benefits based on the results of individual trials vary from no statistically significant benefit in patients with peripheral arterial disease to an ∼50% risk reduction in patients with unstable angina.¹⁰ In terms of absolute benefit, the protective effects of aspirin translate into avoidance of a major vascular event in 50 of 1,000 patients with unstable angina treated with aspirin for 6 months to 36 of 1,000 patients with prior MI, stroke, or TIA treated with aspirin for ∼30 months.¹⁰

For patients with various manifestations of ischemic cardiac or neurologic disease, there is consensus that the optimal dose of aspirin for prevention of MI, stroke, or vascular death lies within the narrow range of 75 to 160 mg/d. This concept is supported by an overview of all antiplatelet trials that showed no obvious aspirin dose dependence for the protective effects of aspirin based on direct and indirect comparisons¹⁰ as well as by the results of individual trials that randomized patients to treatment with low-dose aspirin or placebo or to two different doses of aspirin (Table 3). As discussed earlier, there is no convincing evidence that the dose requirement for the antithrombotic effect of aspirin varies in different clinical settings.

Among most high-risk patient groups, the expected number of individuals avoiding a serious vascular event by using aspirin substantially exceeds the number experiencing a major bleed. It is less certain, however, whether aspirin is of benefit in apparently healthy people who are at intermediate risk for serious vascular events because they have well-established cardiovascular risk factors. The Antithrombotic Trialists’ Collaboration addressed this issue in an individual participant data meta-analysis of the results of large randomized trials of aspirin for primary prevention of vascular events.⁸⁷ The analysis was based on the results of six primary prevention trials that included 95,456 subjects^27,88-92 with a mean follow-up of 6.9 years and a median follow-up among survivors of 5.5 years, reflecting the fact that the Women’s Health Study,⁹¹ which accounted for almost one-half of the participants, had a mean follow-up of ∼10 years (Table 4). The effects of aspirin for primary prevention were compared with its effects in high-risk settings using the results of six trials among patients with a history of MI, nine trials among patients with a history of TIA or stroke, and one trial in patients with moderately severe diabetic retinopathy.¹⁰

The results of the Antithrombotic Trialists’ Collaboration individual patient meta-analysis indicated a 12% proportional reduction in the incidence of serious vascular events (rate ratio [RR], 0.88; 95% CI, 0.82-0.94; P = .0001) and an 18% proportional reduction in the incidence of major coronary events (RR, 0.82; 95% CI, 0.75-0.90; P < .0001). Most of the benefit of aspirin was due to a 23% reduction in nonfatal MI (RR, 0.77; 95% CI, 0.67-0.89; P < .0001); there was no apparent reduction in cardiovascular death (RR, 0.95; 95% CI, 0.78-1.15; P = .50). Aspirin was associated with a nonsignificant 10% reduction in nonhemorrhagic stroke (RR, 0.90; 95% CI, 0.80-1.01; P = .08).

Aspirin had no significant effect on the aggregate of all vascular causes of death (RR, 0.98; 95% CI, 0.87-1.10; P not significant), and there was no evidence of a protective effect on the two-thirds of vascular deaths due to coronary heart disease (RR, 0.95; 95% CI, 0.78-1.15; P = .5), the one-sixth due to stroke (RR, 1.21; 95% CI, 0.84-1.74; P = .2), or the remaining vascular causes of death (RR, 0.90; 95% CI, 0.65-1.26; P = .4). Aspirin had no significant effect on nonvascular mortality (RR, 0.93; 95% CI, 0.85-1.02; P = .1) or unknown causes of death (RR, 0.95; 95% CI, 0.75-1.21; P not significant) but increased the risk of major extracranial bleeds (RR, 1.55; 95% CI, 1.31-1.83; P < .0001).

The absolute benefits of aspirin were summarized under the term ”occlusive vascular events,” that is, vascular events other than hemorrhagic stroke or fatal extracranial bleeds. Among low-risk individuals (annual risk of coronary heart disease ≤ 1%), allocation to aspirin resulted in the avoidance of four (95% CI, 2-6) occlusive vascular events per 1,000 after 5 years, which was chiefly attributable to three (95% CI, 1-5) fewer major coronary events. This benefit was offset by two (95% CI, 1-3) additional major extracranial bleeds per 1,000 over the same 5-year period, but there was no significant excess risk of hemorrhagic stroke (0.1 [95% CI, −0.5 to 0.7] excess; P not significant). Among the much smaller number of moderate-risk participants (yearly coronary heart disease risk > 1%), the net effect of aspirin over 5 years on occlusive vascular events was statistically uncertain (13 [95% CI, −3-29] fewer; P = .1) because despite a definite reduction in major coronary events (18 [95% CI, 5-30] fewer per 1,000), there was no significant reduction in presumed ischemic stroke (one fewer per 1,000 [95% CI, nine more to 11 fewer]), and there was an excess of both hemorrhagic stroke (five [95% CI, 1-9] more per 1,000) and major bleeds (4 more per 1,000 [95% CI, one fewer to 10 more]). About one-half of the hemorrhagic strokes were fatal (and the remainder would be expected to result in moderate or severe disability), so this hazard substantially offsets any cardiac benefits in moderate-risk individuals.

The results of three subsequently completed primary prevention trials involving 7,165 patients with diabetes, peripheral arterial disease, or both (Table 4^27,88-95) are consistent with those of the individual patient meta-analysis performed by the Antithrombotic Trialists’ Collaboration.^93-95 A meta-analysis of all nine primary prevention trials demonstrated a borderline significant reduction in all-cause mortality with aspirin but no reduction in cardiovascular mortality,⁹⁶ a finding that raises the possibility that aspirin also prevents nonvascular (cancer) mortality (Discussed in “Cancer Incidence and Mortality”).

Previous meta-analyses of the effects of antiplatelet therapy in persons at high risk of occlusive vascular disease⁷ have shown that the benefits of aspirin far exceed the bleeding risks. By contrast, the majority (92%) of participants in the primary prevention trials was at low absolute risk of coronary disease; on average, the annual risk of a vascular event in the primary prevention trials was only about one-tenth of that in the high-risk trials. Although the proportional benefits of aspirin appeared broadly similar when used for primary or secondary prevention, the absolute benefits of aspirin in the primary prevention trials were very small. When used for primary prevention, fewer than one person of every 1,000 treated with aspirin would avoid an occlusive vascular event, whereas a comparably small number would experience a major extracranial bleed. Until the benefits of aspirin can be defined more precisely, therefore, the possibility of a benefit for vascular prevention does not seem to justify the potential for harms. However, these estimates do not take into account the benefits of aspirin for the prevention of cancer and cancer-related mortality, which might tip the balance in favor of aspirin use for primary prevention. Additional trials evaluating the utility of aspirin for primary prevention of cardiovascular disease in elderly patients^97,98 and in patients with diabetes⁹⁹ are ongoing. Readers are referred to the prevention of cardiovascular disease article in this supplement by Vandvik et al.¹⁰⁰

Anticoagulant therapy with dose-adjusted warfarin (international normalized ratio, 2.0-3.0),¹⁰¹ the direct thrombin inhibitor dabigatran etexilate,¹⁰² or the direct factor Xa inhibitors rivaroxaban and apixaban¹⁰³ is very effective in reducing the risk of stroke in patients with nonvalvular atrial fibrillation. The efficacy of aspirin (in doses ranging from 75-1,200 mg/d) has been compared with that of placebo or no antiplatelet treatment in seven randomized trials that included 3,990 patients with nonvalvular atrial fibrillation.¹⁰¹ A pooled analysis revealed a relative risk reduction of ∼19% with aspirin compared with placebo or no treatment (95% CI, −1%-35%), which is consistent with the 22% (95% CI, 6%-35%) relative risk reduction obtained when comparing any antiplatelet therapy with placebo or no antiplatelet therapy for stroke prevention in patients with nonvalvular atrial fibrillation.¹⁰¹ Pooled analysis of 10 trials involving 4,620 patients with nonvalvular atrial fibrillation revealed that dose-adjusted vitamin K antagonist therapy was significantly more effective than aspirin, with a 39% relative risk reduction (95% CI, 19%-53%).¹⁰⁴ Warfarin is also more effective than the combination of aspirin and clopidogrel.¹⁰⁵

The efficacy of antiplatelet therapy for stroke prevention in atrial fibrillation has been confirmed by the results of the Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events (ACTIVE) A trial, which compared the combination of aspirin plus clopidogrel with aspirin alone in 7,554 patients deemed ineligible for warfarin.¹⁰⁶ Aspirin plus clopidogrel reduced the risk of major vascular events, comprising the composite of stroke, MI, non-CNS embolism, or death from vascular causes, by 11% compared with aspirin (95% CI, 2%-19%) primarily because of a 28% reduction in stroke (95% CI, 17%-38%). However, the combination of clopidogrel plus aspirin is less effective than warfarin and is associated with a similar risk of bleeding.¹⁰⁵ Readers are referred to You et al¹⁰⁷ in this guideline for further discussion of the use of antiplatelet therapy for stroke prevention in patients with atrial fibrillation.

The Pulmonary Embolism Prevention (PEP) trial results demonstrated that aspirin is effective in preventing VTE after major orthopedic surgery.¹⁰⁸ This double-blind, multicenter study included 13,356 patients undergoing surgery for hip fracture and an additional 4,088 patients undergoing elective hip or knee arthroplasty. Patients were randomized to receive aspirin (160 mg/d) or placebo for 5 weeks, with the first dose administered prior to surgery. Other forms of prophylaxis were allowed, and either heparin or low-molecular-weight heparin was used in ∼40% of the patients. Among the 13,356 patients undergoing surgery for hip fracture, aspirin produced a 36% reduction in symptomatic DVT or pulmonary embolism (absolute risk reduction, 0.9%; P = .0003). A similar relative risk reduction was observed in aspirin-treated patients who did or did not receive concomitant heparin or low-molecular-weight heparin. These results are consistent with those of meta-analyses performed by the Antiplatelet Trialists’ Collaboration¹⁰⁹ and by Sandercock and colleagues¹¹⁰ of antiplatelet trials in patients with stroke.

When compared with warfarin, heparin, low-molecular-weight heparin, or danaparoid, aspirin was associated with similar or higher rates of DVT detected by screening ultrasound or venography; however, the frequency of symptomatic events was low.¹¹¹ A large randomized controlled trial is required to compare the effectiveness of aspirin with that of anticoagulants for the prevention of fatal or symptomatic VTE events. Readers are referred to Falck-Ytter et al¹¹² in this guideline for further discussion of the use of antiplatelet therapy for prevention of VTE.

Preeclampsia and fetal growth restriction are believed to be related to reduced placental blood flow, which is believed to be caused by constriction, thrombosis, or both of small placental arteries.¹¹³ The initial reports that low-dose aspirin therapy reduces the risk of severe low birth weight among newborns¹¹⁴ and lowers the need for cesarean section in mothers with pregnancy-induced hypertension¹¹⁴ led to the widespread use of prophylactic aspirin for prevention of preeclampsia. Subsequently, several larger trials reported no beneficial effects of aspirin.^115-121 However, a systematic review of data from 59 trials in 37,560 women confirmed that antiplatelet therapy (mostly aspirin 60 mg/d) is beneficial. Aspirin was associated with a 17% decrease in the risk of preeclampsia, an 8% reduction in the risk of preterm birth, a 14% reduction in the risk of fetal or neonatal death, and a 10% reduction in small-for-gestational age babies.¹²² An individual patient meta-analysis of 31 trials involving 32,217 patients who received antiplatelet therapy for primary prevention of preeclampsia revealed a consistent benefit of aspirin for the prevention of eclampsia (overall 10% relative risk reduction) in all of the subgroups studied (first pregnancy with or without any high risk factor; second pregnancy with or without high risk factors or history of hypertensive disorder of pregnancy; preexisting renal disease, diabetes, hypertension, or previous infant small for gestational age; maternal age; singleton or multiple pregnancy; and timing of starting of treatment or intended aspirin dose ≤ 75 mg/d or ≥ 75 mg/d).¹²³ Aspirin given in doses ranging from 50 to 150 mg/d accounted for 98% of women included in this data set. Readers are referred to Bates et al¹²⁴ in this guideline for further discussion of the use of antithrombotic therapy during pregnancy.

There is compelling evidence from randomized controlled trials that aspirin reduces the incidence of colorectal cancer and cancer mortality.^125,126 An individual patient meta-analysis of eight randomized controlled trials that included 25,570 subjects demonstrated that compared with no aspirin, daily aspirin for a scheduled mean treatment duration of at least 4 years reduced the odds of cancer deaths by 21% (95% CI, 8%-32%). The mortality benefit appeared to be unrelated to aspirin dose, only became apparent after 5 years of follow-up, and the absolute benefit increased over time. The greatest mortality benefit was seen with adenocarcinoma. Among patients aged ≥ 65 years at the start of the trials, the absolute reduction in cancer deaths over 20 years was 7.1% (95% CI, 2.4%-11.7%).¹²⁶ Separate analyses based on individual patient data from four trials of 14,033 patients followed for a median of 18.3 years demonstrated that aspirin (at doses of 75-300 mg/d) also reduced the incidence of colorectal cancer. Furthermore, an analysis from the Dutch TIA trial suggested that the risk of fatal colorectal cancer was higher with aspirin doses of 30 mg/d than it was with a dose of 283 mg/d.¹²⁶

Aspirin-induced impairment of primary hemostasis cannot be separated from its antithrombotic effect and appears to be similar with all doses ≥ 75 mg/d.⁹ The balance between preventing thrombotic events and causing bleeding with aspirin critically depends on the absolute thrombotic vs hemorrhagic risk of the patient. Thus, in individuals at low risk for vascular occlusion (eg, ≤ 1% per year), the very small reduction of vascular events is probably offset by bleeding complications. In contrast, in patients at high risk of cardiovascular or cerebrovascular complications (eg, > 3% per year), the substantial absolute benefit of aspirin prophylaxis clearly outweighs the harm (Table 5). For example, the absolute excess of major bleeds (ie, those requiring transfusion) in patients with acute MI is ∼1/100th the absolute number of major vascular events avoided by aspirin therapy.¹⁰

The overall risk of major extracranial and intracranial hemorrhage associated with antiplatelet drugs is difficult to assess in individual trials because their incidence is < 1% per year. This makes detection of even a 50% to 60% relative increase in risk unrealistic in most trials of a few thousand patients.

Aspirin-induced GI toxicity, as detected in randomized clinical trials, appears to be dose-related, with doses in the range of 30 to 1,300 mg/d.¹²⁷ This conclusion is based on both indirect comparisons of different trials and direct randomized comparisons of different aspirin doses, as reviewed previously in this article. The dose-response relationship for GI toxicity is believed to reflect at least two COX-1-dependent components: dose-dependent inhibition of COX-1 in the GI mucosa and dose-independent (within the range of examined doses) inhibition of COX-1 in platelets.⁶ Thus, it is not surprising that the antithrombotic effect of aspirin can be dissociated, at least in part, from GI bleeding. Even at low doses, aspirin causes serious GI bleeding.^40,47 Because of the underlying prevalence of gastric mucosal erosions related to concurrent use of other NSAIDs and Helicobacter pylori infection in the general population, it should be expected that any dose of aspirin will cause more GI bleeding from preexisting lesions than placebo. Consistent with this mechanistic interpretation, the relative risk of hospitalization due to upper-GI bleeding and perforation associated with low-dose aspirin therapy (mostly 100-300 mg/d) is comparable to that with other antiplatelet drugs and anticoagulants (ie, 2.3 [95% CI, 1.7-3.2], 2.0 [95% CI, 1.4-2.7], and 2.2 [95% CI, 1.4-3.4], respectively, in a large population-based observational study¹²⁸).

In the 2002 overview of randomized trials of aspirin for secondary vascular prevention performed by the Antithrombotic Trialists’ Collaboration,¹⁰ information was available on 787 major extracranial hemorrhages in 60 trials recording at least one such hemorrhage. These were generally defined as hemorrhages that were fatal or required transfusion; among them, 159 (20%) caused death. Overall, the proportional increase in risk of a major extracranial bleed with antiplatelet therapy was about one-half (OR, 1.6; 95% CI, 1.4-1.8), with no significant difference between the proportional increases observed in each of the five high-risk categories of patients. A similar proportional increase in extracranial bleeding was obtained in the 2009 individual patient meta-analysis by the Antithrombotic Trialists’ Collaboration of six trials of aspirin for primary prevention that included 554 extracranial hemorrhages (OR, 1.5; 95% CI, 1.3-1.8).⁸⁷

A case-control study with hospital and community controls examined the risks of hospitalization for bleeding peptic ulcer associated with three different regimens of aspirin prophylaxis.¹²⁹ ORs were calculated for different doses of aspirin: 75 mg (2.3; 95% CI, 1.2-4.4), 150 mg (3.2; 95% CI, 1.7-6.5), and 300 mg (3.9; 95% CI, 2.5-6.3). Additional epidemiologic studies have found a dose-response relationship between aspirin prescription and upper-GI complications, as reviewed by García Rodríguez et al.¹³⁰ It has been calculated that ∼900 of the 10,000 episodes of ulcer bleeding occurring in persons aged > 60 years each year in England and Wales could be associated with and ascribed to prophylactic aspirin use.¹²⁹ If the assumptions from indirect comparisons are correct, a general change to lower doses (75 mg/d) of aspirin would not eliminate the risk but would reduce it by ∼40% compared with a 300-mg dose and by 30% compared with a 150-mg dose.¹²⁹ The mortality rate among patients who are hospitalized for NSAID-induced upper-GI bleeding is 5% to 10%.^131,132

The widely held belief that enteric-coated and buffered aspirin preparations are less likely to cause major upper-GI bleeding than plain tablets was evaluated in a multicenter case-control study.¹³³ The relative risks of upper-GI bleeding for plain, enteric-coated, and buffered aspirin at average daily doses of ≤ 325 mg were 2.6, 2.7, and 3.1, respectively. At doses > 325 mg, the relative risk was 5.8 for plain and 7.0 for buffered aspirin; there were insufficient data to evaluate enteric-coated aspirin at this dose level.¹³³ Similar conclusions were reached by a case-control study using data from the UK General Practice Research Database.¹³⁴

Suppressing acid secretion is believed to reduce the risk of ulcers associated with the regular use of NSAIDs. In patients who required continuous treatment with NSAIDs and who had ulcers or > 10 erosions in their stomach or duodenum, omeprazole healed and prevented ulcers more effectively than did ranitidine.¹³⁵ In these patients, maintenance therapy with omeprazole was associated with a lower rate of relapse and was better tolerated than misoprostol.¹³⁶ In patients with a history of previous ulcer bleeding who took low-dose aspirin for 6 months, omeprazole and H pylori eradication were associated with similar rates of recurrent bleeding (0.9% and 1.9%, respectively),¹³⁷ although clinically important differences between the two preventive strategies could not be excluded owing to the small sample size (n = 250).

Two relatively small studies^138,139 have challenged earlier guidelines that recommended the use of clopidogrel for patients who have major GI contraindications to aspirin, principally recent significant bleeding from a peptic ulcer or gastritis. Both studies enrolled patients who developed ulcer bleeding after the use of low-dose aspirin. In a study by Chan et al,¹³⁸ after healing of ulcers and eradication of H pylori, if present, 320 patients were randomly assigned to receive either clopidogrel 75 mg/d or aspirin 80 mg/d plus esomeprazole 20 mg bid for 12 months. The cumulative incidence of recurrent bleeding was 8.6% (95% CI, 4.1%-13.1%) among patients who received clopidogrel and 0.7% (95% CI, 0%-2.0%) among those who received aspirin plus esomeprazole (P = .001).¹³⁸ In a study by Lai et al,¹³⁹ 170 patients with prior ulcer bleeding were randomly assigned to treatment with clopidogrel 75 mg/d or aspirin 100 mg/d and esomeprazole 20 mg/d for 1 year. The cumulative incidence of recurrent ulcer complications was 13.6% and 0%, respectively (95% CI for the difference, 6.3%-20.9%; P = .0019).¹³⁹ The combination of esomeprazole and low-dose aspirin is superior to clopidogrel for the prevention of recurrent GI bleeding as is now recommended by the 2008 guidelines of the American College of Cardiology/American College of Gastroenterology/American Heart Association.¹⁴⁰

Substantially less information is available about the risk of intracranial hemorrhage associated with aspirin use. In the Nurses’ Health Study cohort of ∼79,000 women aged 34 to 59 years, infrequent use of aspirin (1-6 tablets per week) was associated with a reduced risk of ischemic stroke, whereas high frequency use (≥ 15 tablets per week) was associated with an increased risk of subarachnoid hemorrhage, particularly among older or hypertensive women.¹⁴¹ In the 2002 overview of the Antithrombotic Trialists’ Collaboration,¹⁰ the overall absolute excess of intracranial hemorrhage due to aspirin therapy was less than one per 1,000 patients per year in trials that involved patients at high risk for cardiovascular events, with a somewhat higher risk in patients with cerebrovascular disease. The 2009 individual patient meta-analysis of primary prevention trials indicated that aspirin was associated with five additional hemorrhagic strokes per 1,000 among moderate-risk participants (risk of coronary event, > 1% per year) over 5 years (ie, ∼1/1,000 per year) but substantially fewer events in low-risk participants.⁸⁷

Low-dose aspirin therapy has not been reported to affect renal function or BP control,¹⁴² consistent with its lack of effect on renal prostaglandins¹⁴³ primarily derived from constitutively expressed COX-2 in the kidney.⁸⁴ Moreover, aspirin 75 mg/d did not affect BP or the need for antihypertensive therapy in intensively treated patients with hypertension.⁸⁹ The suggestion that the use of aspirin and other antiplatelet agents is associated with reduced benefit in enalapril-treated patients with left ventricular systolic dysfunction¹⁴⁴ is not supported by the results of a large meta-analysis of MI trials.¹⁴⁵ Similarly, no negative interaction occurred between angiotensin-converting enzyme (ACE) inhibition and the cardiovascular benefits of low-dose aspirin in intensively treated patients with hypertension.¹⁴⁶ The ACE Inhibitors Collaborative Group performed a systematic overview of data from 22,060 patients included in six long-term randomized trials of ACE inhibitors to assess whether aspirin altered the effects of ACE inhibitor therapy on major clinical outcomes.¹⁴⁷ Even though the results from these analyses cannot rule out the possibility of an interaction, they show unequivocally that even if aspirin is given, the addition of ACE inhibitor therapy produces substantial additional benefit in all major vascular outcomes. Therefore, in the absence of clear contraindications, concomitant use of aspirin and ACE inhibitors should be considered in all patients who are at high risk of major vascular events.¹⁴⁷

Dipyridamole is a pyrimidopyrimidine derivative with vasodilator and antiplatelet properties. The mechanism of action of dipyridamole as an antiplatelet agent is controversial.¹⁴⁸ Both inhibition of cyclic nucleotide phosphodiesterase (the enzyme that degrades cyclic adenosine monophosphate [AMP] to 5^′-AMP, resulting in the intraplatelet accumulation of cyclic AMP, an inhibitor of platelet aggregation) and blockade of the uptake of adenosine (which binds to A₂ receptors, stimulates platelet adenyl cyclase, and increases cyclic AMP) have been suggested. Moreover, direct stimulation of PGI₂ synthesis and protection against its degradation have been reported, although the dipyridamole concentrations required to produce these effects far exceed the low-micromolar plasma levels achieved after oral administration of conventional doses (100-400 mg/d).¹⁴⁸ Dipyridamole also differentially inhibits the expression of critical inflammatory genes by platelet-leukocyte aggregates.¹⁴⁹

The absorption of dipyridamole from conventional formulations is quite variable and may result in low systemic bioavailability of the drug. A modified-release formulation of dipyridamole with improved bioavailability has been developed in a combination pill with low-dose aspirin.¹⁵⁰ Dipyridamole is highly protein bound to albumin, eliminated primarily by biliary excretion as a glucuronide conjugate, and is subject to enterohepatic recirculation. A terminal half-life of 10 h has been reported. This is consistent with the bid regimen used in recent clinical studies.

The clinical efficacy of immediate-release dipyridamole, alone or in combination with aspirin, was questioned on the basis of earlier randomized trials.^3,151 As a result, the reformulated extended-release preparation was evaluated in the more recent ESPS-2, Aspirin Plus Dipyridamole Versus Aspirin Alone After Cerebral Ischaemia of Arterial Origin (ESPRIT), and Prevention Regimen for Effectively Avoiding Second Strokes (PRoFESS) randomized trials. In ESPS-2, the new preparation of dipyridamole was evaluated in 6,602 patients with prior stroke or TIA.⁴⁰ This study showed that the addition of modified-release dipyridamole (200 mg bid) to aspirin (25 mg bid) was associated with a 22% relative risk reduction in major vascular events compared with aspirin alone. Headache was the most common adverse effect of dipyridamole.

In the ESPRIT trial,¹⁵² 2,739 patients within 6 months of a TIA or minor stroke of presumed arterial origin were randomized to receive aspirin (30-325 mg/d) with or without dipyridamole (200 mg bid). Compared with aspirin alone, the primary outcome (a composite of major vascular events or major bleeding complications) was reduced by 20% with the combined treatment. Patients on aspirin plus dipyridamole discontinued trial medication almost three times more often than those on aspirin alone mainly because of headache.¹⁵²

In the PRoFESS trial,¹⁵³ 20,332 patients with recent ischemic stroke were randomized to receive the combination of aspirin (25 mg bid) plus extended-release dipyridamole (200 mg bid) or clopidogrel (75 mg once daily) for a mean of 2.5 years. The HR for the primary efficacy outcome (1.01; 95% CI, 0.92-1.11) failed to reach the prespecified noninferiority margin. Major hemorrhages occurred more frequently in those given aspirin plus extended-release dipyridamole than in patients treated with clopidogrel (HR, 1.15; 95% CI, 1.00-1.32), and the combination was associated with an excess of intracranial hemorrhage (HR, 1.42; 95% CI, 1.11-1.83).

A meta-analysis of six randomized trials involving 7,648 patients with a history of TIA or stroke in which stroke was reported as an outcome demonstrated that compared with aspirin alone (dose range, 50-1,300 mg/d), the combination of aspirin (dose range, 50-1,300 mg/d) plus dipyridamole reduced stroke by 23% (RR, 0.77; 95% CI, 0.67-0.89), with no statistical evidence of heterogeneity.¹⁵⁴ Consistent estimates were obtained from trials that used the immediate-release preparation of dipyridamole (four trials) and those that used the extended-release preparation. A Cochrane review of 29 randomized trials involving 23,019 patients confirmed the superiority of the combination of aspirin plus dipyridamole over aspirin alone for prevention of vascular events in patients with a history of TIA or stroke but found no evidence of a benefit of the combination in studies involving patients with a history of coronary or peripheral arterial disease or in other high-risk patients.¹⁵⁵