Aspirin has long been regarded as one of the essential drugs in the management of atherothrombotic disease and is widely used in primary and secondary prevention, being regarded as cost effective and widely available. Aspirin is clearly recommended as an antiplatelet agent for secondary prevention for the reduction of cardiovascular events and mortality, as endorsed in current British, European, and American guidelines. Because of the increasing prevalence of conditions predisposing to cardiovascular disease, primary prevention with aspirin is potentially applicable to a huge population. Understandably, the absolute risk reduction is much smaller in primary prevention because event rates are lower, although the proportional reduction may potentially be significant at the population level. Importantly, clinical trial data supporting aspirin use in primary prevention are equivocal at best, and low event rates have often rendered these studies underpowered.

For example, aspirin failed to show beneficial reductions in cardiovascular events in patients with diabetes with evidence of peripheral arterial disease and in the largest group of asymptomatic patients with diabetes studied for primary prevention. A primary prevention study in women also failed to achieve the primary end point. Furthermore, there is no evidence that aspirin effectively reduces thrombotic events or is safe in patients with heart failure (in whom it may even result in more hospitalizations for heart failure compared to warfarin). A recent meta-analyses suggests that aspirin use for primary prevention significantly reduces the incidence of some cardiovascular events but results in no significant reduction in overall mortality. Also, aspirin significantly reduces the incidence of ischemic stroke but at the price of a significant increase in the incidence of hemorrhagic stroke, with an overall adverse effect on stroke mortality. Finally, aspirin significantly increases gastrointestinal and other extracranial bleeding. Overall, the evidence base for a role for aspirin in primary prevention in atherothrombotic disease is not compelling, and there is certainly a coexisting increase in bleeding complications.

In this issue of The American Journal of Cardiology , Cannon et al analyzed the United States cohort of 25,686 patients from the Reduction of Atherothrombosis for Continued Health (REACH) registry, an international, prospective, observational study investigating the use of aspirin and antithrombotic agents in outpatients with atherothrombotic disease. Their aims were to describe the use of aspirin and other antithrombotic drugs, to identify independent predictors influencing its use, and to highlight subgroups remaining at risk because of the underuse of aspirin in a “real-world” setting. Subjects had coronary artery disease (59.8%), cerebrovascular disease (21.3%), peripheral arterial disease (9.3%), or no symptoms with >3 cardiovascular risk factors (25.8%), and 71% received aspirin, either alone or in combination with other antithrombotic drugs, most commonly in low doses (75 to 100 mg). They found that aspirin use was significantly higher in subjects with symptoms (75%) compared to those with risk factors alone (59%). Antithrombotic use was also positively associated with being Caucasian, having atrial fibrillation or vascular disease, taking other risk-modifying medications, and being under the care of a cardiologist. Interestingly, 25% of subjects with established cardiovascular disease were not treated with aspirin for secondary prevention, and the strongest predictor of not being on aspirin was the use of an oral anticoagulant; also interestingly, 1% received dual-antiplatelet therapy and warfarin. However, 15% of subjects overall received no antithrombotic medication, including 9.1% with known disease and 34.5% with risk factors only. Predictors of lack of antithrombotic use included being female or a current smoker and having diabetes.

Apart from its cross-sectional registry design, the major limitation of this study is the absence of information regarding past bleeding complications or perceived future bleeding risk. This is likely to represent the most important factor influencing physicians not to prescribe antithrombotic medication in those with the potential to benefit. Aspirin is a common culprit in adverse drug reactions, and there may be sound reasoning underlying its perceived underuse. Nonetheless, given the limited benefit of aspirin for primary prevention over more compelling cardiovascular drugs used in primary prevention, such as statins and angiotensin-converting enzyme inhibitors, perhaps the incremental benefit of aspirin is limited, with the potential for harm from bleeding. However, Cannon et al highlight important subgroups in which aspirin appears to be underused despite current recommendations, suggesting that large populations potentially remain at inappropriately increased risk. Efforts should undoubtedly be made to reinforce aspirin use in these subgroups if the omission of antithrombotic drugs is not justified. However, we must be vigilant to the phenomenon of “indication creep,” whereby aspirin use spreads beyond the boundaries of evidence if the data are inappropriately extrapolated. An intuitive approach to prescribe aspirin for primary prevention at the first presentation of a patient with cardiovascular risk factors may be disabling or fatal. Aspirin, thus, should only be prescribed to disease-free populations if there is clear evidence of efficacy and safety. Larger, appropriately powered studies with “hard” clinical end points are vital to evaluate whether aspirin has an overall beneficial role in primary prevention, especially in cohorts already taking statins.

More attention must be paid to the adverse effects of aspirin and other antithrombotic drugs, specifically the risk for bleeding complications. Clinicians should stratify for bleeding risk and individually tailor therapy. However, this aspect is clearly complex, as risk factors for cardiovascular events are also risk factors for bleeding, including advancing age and diabetes.

To make things even more complicated, the phenomenon of nonresponsiveness to aspirin is increasingly recognized as a clinically relevant entity. Indeed, the laboratory-defined inability of aspirin to sufficiently inhibit platelet activity can adversely affect the risk for cardiovascular complications. For example, in patients with previous ischemic attacks or strokes, the incidence of aspirin resistance was much higher compared to asymptomatic patients with known cerebrovascular disease. Accordingly, patients “resistant” to aspirin are at a greater risk for clinically important cardiovascular morbidity over the long term than patients who are sensitive to aspirin. In another study, aspirin nonresponders had a 10-fold increase in the risk for recurrent vascular events compared to aspirin-sensitive patients after 2-year follow-up. Unfortunately, at present there is no reliable and widely available clinical tool to identify those patients whose platelets will not be sufficiently inhibited by aspirin.

International guidelines regarding aspirin in primary prevention should be reconsidered in the light of emerging evidence, including assessment of bleeding risk. Perhaps aspirin may be reserved for those at highest cardiovascular risk. Alternative antiplatelet agents, such as clopidogrel, prasugrel, and ticagrelor, as well as those in earlier stages of development are emerging and may develop a role in our future management of atherothrombotic disease.