Acetylsalicylic Acid (ASA) for Prevention of Vascular Events

This set of recommendations is intended to provide guidance for the use of acetylsalicylic acid (ASA) for primary prevention of a first vascular event. The use of ASA for the secondary prevention of vascular disease has been in practice since the 1950’s, and research on the use of ASA in primary prevention since the late 1970’s and early 1980’s. This information was widely available to the public and ASA is available without prescription, therefore many people may have initiated daily ASA without discussions with healthcare professionals, such as in cases where there is a family history of some form of vascular disease.

In recent years the risk-benefit profiles of using ASA for primary prevention in an individual without vascular disease has come under scrutiny. Three recent randomized trials have consistently found that the risk of bleeding is higher and potentially outweighs the benefits achieved for vascular risk reduction in otherwise healthy populations. The decision to start, stop or continue ASA in individuals should be made through an informed discussion between the individual and their healthcare team, weighing individual risk, benefit and preferences.  Some decision-making research is emerging that indicates some people may choose a risk of bleeding over a risk of a heart attack or stroke in choosing whether or not to take ASA.  No specific sex and gender differences in outcomes have been reported in the current evidence.

1. Proportion of people taking acetylsalicylic acid for primary vascular prevention.
2. Proportion of people taking acetylsalicylic acid for primary vascular prevention who experience a bleeding complication (gastrointestinal, intracerebral).
3. Proportion of people taking acetylsalicylic acid for primary vascular prevention who experience a vascular event (stroke, heart condition).
4. Population indicator: public perceptions on use of ASA for primary prevention (based on polling Q TBD) – include valid measure of patient reported risk factors
5. Polling question: Have you heard about new research indicating that ASA is not being recommended for primary prevention of vascular diseases now? – would you stop, did you stop?

Measurement Notes:

• Lack of centralized database to obtain data for these indicators; however, they remain important and healthcare professionals are urged to collect this data at least at a local level to help drive knowledge gaps and improvement opportunities.
• Public polling methods should be considered to obtain data for indicators 1, 4, 5.

Health Care Provider Information

• Canadian Stroke Best Practice Recommendations
• Canadian Stroke Consortium
  • Education site: contains archived learning modules and knowledge breaks
• Canadian Cardiovascular Society
  • Guidelines
  • The Use of Antiplatelet Therapy in the Outpatient Setting: Canadian Cardiovascular Society Guidelines
• Thrombosis Canada
  • Clinical guide
• Canadian Society of Hospital Pharmacists
• Canadian Pharmacists Association
  • Education site
• Shared Decision-Making
  • Should I Take Aspirin?
  • Eliciting patient values and preferences to inform shared decision making in preventive screening.

Information for People who have Experienced a Stroke, their Families and Caregivers

• Infographic ASA for Prevention of Vascular Events
• Community of Survivors
• Care Supporter’s Community
• Are you at Risk for Heart Disease or Stroke?
• Post Stroke Checklist
• Services and Resources Directory
• Your Stroke Journey
• Living Well with Heart Disease
• Managing Your Blood Pressure
• How to Manage your Cholesterol

Evidence Table and Reference List

Acetylsalicylic Acid (ASA) for Secondary Prevention

The benefit of long-term acetylsalicylic acid (ASA) or aspirin use for secondary prevention is well established. Daily, low-dose ASA reduces the risk of vascular events including myocardial infarction (MI), stroke, and vascular death in patients who have experienced a previous vascular event or who are at high risk of vascular disease. A meta-analysis conducted by The Antithrombotic Trialists’ Collaboration (2002) included the results of 287 RCTs (n=135,000) examining any antiplatelet therapy for the prevention of vascular events in high-risk patients. In 9 of these trials, long-term aspirin monotherapy was examined in patients who had experienced a previous stroke or TIA. In these trials, fewer patients receiving ASA therapy experienced a vascular event (8.2% vs. 9.1%) representing an 11% odds reduction. In 65 trials examining ASA monotherapy, the mean percentage odds reduction of any vascular event, across doses ranging from <75 mg to 1,500 mg, was 23%. Treatment with ASA reduced the number of serious vascular events by 36 per 1,000 per year over two years in patients with a previous myocardial infarction and by 36 per 1,000 per year in patients with a previous history of stroke or transient ischemic attack, compared with placebo. In patients with peripheral arterial disease, treatment with ASA reduced the odds of serious vascular events by 23%, compared with placebo. Similar risk reductions were seen for patients with stable or unstable angina.

ASA also helps to reduce the risk of recurrent vascular events following an acute stroke. Rothwell et al. (2016) included the results of 12 trials comparing ASA vs. placebo, of which 11 trials included comparisons of ASA monotherapy versus placebo, stratified by time periods (< 6 weeks, 6–12 weeks, and >12 weeks). ASA doses ranged from 50 to 1,200 mg per day. ASA monotherapy significantly reduced the risks of any ischemic stroke, disabling or fatal ischemic stroke, any stroke and any fatal stroke up to 12 weeks post event, with large risk reductions ranging from 40% to 70%. The greatest reduction in early stroke recurrence associated with ASA monotherapy was among patients presenting with mild or moderately disabling stroke. There was no reduction in risk of recurrent ischemic stroke with ASA use after 12 weeks (OR= 0·97, 95% CI 0·84–1·12, p=0·67). In an updated Cochrane review, Sandercock et al. (2014) included the results of the CAST (1997) and IST (1997), the two largest trials testing ASA, which contributed 98% of the data.  ASA therapy, initiated within 48 hours of stroke onset, was associated with a significant reduction in the odds of being dead or dependent at final follow-up (OR= 0.95, 95% CI 0.91 to 0.99) and in the odds of death at a final follow-up (OR=0.92, 95% CI 0.87 to 0.98).  ASA therapy was associated with a significant reduction in the odds of recurrent stroke during treatment (OR=0.77, 95% CI 0.69-0.87) and marginally increased odds of intracranial hemorrhage (OR=1.22, 95% CI 1.00- 1.50). For every 1,000 people treated with ASA, 13 fewer people would avoid death or dependency, 9 fewer would avoid death and 7 fewer would avoid a recurrent stroke. The results of an older meta-regression analysis including the results of 11 RCTs published up to 1996 (Johnson et al. 1999), suggested that the effectiveness of ASA is uniform across a wide range of doses (50-1,500 mg per day), although doses in the range of 75-100 mg are more typical of what is used in clinical practice currently. ASA was associated with a 15% reduction in recurrent stroke.

ASA for Primary Prevention

While low-dose ASA therapy for primary prevention of cardiovascular disease was once commonly recommended, it is now being reconsidered in light of recent evidence. Currently 2019 ACC/AHA guidelines on the primary prevention of cardiovascular disease suggest that low-dose ASA (75-100 mg/day) might be considered among selected adults, aged 40-70 years at higher risk of cardiovascular disease and should be avoided in persons >70 years (Arnett et al. 2019).  This language was modified from the 2014 recommendation, which stated that “the use of aspirin for cardiovascular (including but not specific to stroke) prophylaxis is reasonable for people whose risk is sufficiently high (10-year risk >10%) for the benefits to outweigh the risks associated with treatment.” (Meschia et al. 2014). The U.S. Preventive Services Task Force made age specific recommendations, suggesting that persons aged 50 to 59 years initiate low-dose ASA if their 10-year cardiovascular risk is >10% and the risk of bleeding is not increased. For persons aged 60-69 years, the recommendation was for the decision to be a personal one, given a similar 10-year risk, while the authors suggested the evidence is insufficient to make recommendations for person younger than 50 years or older than 69 years (Bibbins-Domingo et al. 2016). The 2016 European Guidelines on cardiovascular disease prevention (Piepoli et al. 2016) explicitly state that antiplatelet therapy is not recommended in individuals without cardiovascular disease due to the increased risk of major bleeding, as did older guidelines issued by the Canadian Cardiovascular Society (Bell et al. 2011).This is the first time that recommendations for ASA monotherapy in the context of primary prevention have been issued by the Canadian Stroke Best Practice Recommendations writing group.

Acetylsalicylic acid increases the risk of major bleeding. In a meta-analysis that examined intracranial bleeding outcomes exclusively, Huang et al. (2019) included the results of 13 RCTs (n=134,446) including persons without preexisting symptomatic cardiovascular diseases and compared low-dose ASA (≤100 mg/day, for ≥6 months) vs. placebo, or no treatment and. The use of ASA was associated with a significantly increased risk of any intracranial bleeding (RR=1.37, 95% CI, 1.13-1.66; n=8 trials; 2 additional intracranial hemorrhages in 1,000 people). In a sensitivity analysis, excluding the results from ASPREE (2018), which included elderly people only (≥70 years), the risk became nonsignificant. ASA was not associated with a significantly increased risk of intracerebral hemorrhage or subarachnoid hemorrhage. In subgroup analysis, Asians and persons with a BMI <25 taking ASA were at significantly higher risk for intracerebral hemorrhage. Another systematic review including the results of 15 trails (Abdelaziz et al. 2019) also reported an increased risk of major bleeding (1.47% vs. 1.02%; RR= 1.50; 95% CI: 1.33 to 1.69), intracranial bleeding including hemorrhagic stroke (0.42% vs. 0.32%; RR= 1.32; 95% CI: 1.12 to 1.55), and major GI bleeding (0.80% vs. 0.54%; RR= 1.52; 95% CI: 1.34 to 1.73).

In terms of efficacy to prevent ischemic strokes or TIA, three systematic reviews and meta-analyses have been published recently (Abdelaziz et al. 2019, Mahmoud et al. 2019, Zheng & Roddick 2019). The number of included trials ranged from 11 to 15. All reviews included the results of the ARRIVE, ASCEND and ASPREE trials, with much overlap among the remaining included trials. In two of the reviews the risk of ischemic stroke was reduced significantly with aspirin therapy. Abdelaziz et al. (2019) reported the relative risk of TIA and ischemic stroke were 0.79 (95% CI: 0.71 to 0.89) and 0.87 (95% CI: 0.79 to 0.95, respectively), with associated NNTs of 370, and 500. Zheng & Roddick (2019T) also found the use of aspirin was associated with a significant reduction in ischemic stroke (HR=0.81 [95% CrI, 0.76-0.87]; absolute risk reduction 0.16% [95% CI 0.06 to 0.30]; NNT=540). In contrast, in the third review (Mahmoud et al. 2019), the risk of ischemic stroke was not reduced significantly with aspirin (1.7% vs. 1.8%; RR=0.94, 95% CI 0.86-1.04). The inclusion criteria in these reviews were restricted to participants who had no previous cardiovascular disease (Zheng & Roddick 2019) or preexisting cardiovascular diseases (Abdelaziz et al. 2019) and those without a prior history of atherosclerosis (Mahmoud et al. 2019).

Three trials have been published recently that assessed the potential benefit of 100 mg of ASA versus placebo in persons without pre-existing cardiovascular disease, which was defined slightly differently in each trial. All trials included large sample sizes (>12,000 to >19,000 participants). ARRIVE (2018) included men ≥ 55 years with 2 to 4 cardiovascular risk factors and women ≥ 60 years with 3 or more risk factors and excluded those with diabetes. ASPREE (2018) included men and women aged ≥65 or ≥70 years (11% with diabetes), depending on race and ASCEND included men and women ≥40 years with type 1 or 2 diabetes. Mean age of participants was 64 years in the ARRIVE (2018) and ASCEND (2018) trials, while median age in the ASPREE (2018) trial was 74 years. Current ASA +/- anticoagulants use was an exclusion criterion in all trials; however, 36% of participants in the ASCEND trial had used ASA prior to screening, while 11% used ASA previously in the ASPREE trial. Median duration of follow-up ranged from 4.7 to 7.1 years. Two trials were negative (ARRIVE, ASPREE), whereby the risks of cardiovascular events were not significantly lower in the ASA-treated group. In the ARRIVE and ASPREE trials, the hazard ratios associated with ASA use for the primary outcome were 0.96 (95% CI 0·81–1·13) and 0.95 (95% CI 0.83–1.08), respectively. In the ASCEND trial, the risk of the primary outcome (first serious vascular event [MI, stroke, TIA or cardiovascular death]) was significantly lower in the ASA group (8.5% vs, 9.6%, RR=0.88, 95% CI, 0.79 to 0.97; p=0.01). In all trials, the risk of major bleeding events was increased significantly with ASA therapy (ASPREE, HR=1.38; 95% CI 1.18-1.62; ARRIVE [any gastrointestinal bleeding], HR=2.11, 95% CI 1·36–3·28 and ASCEND, RR=1.29, 95% CI 1.09-1.52).

Marquis-Gravel et al. (2019) highlighted several areas of uncertainty that remain regarding ASA therapy, following the completion of the three latest RCTs. Factors such as body weight and sex were identified as potential effect modifiers. In a meta-analysis of pooled individual patient data from 10 RCTs (Rothwell et al. 2019), the risk of cardiovascular events associated with the use of 75–100 mg ASA decreased with increasing weight, while low-dose ASA had the greatest preventative effect among those participants weighing 50–69 kg. In the same study, ASA doses of 350 and 500 mg were associated with decreased risk of cardiovascular events in persons weighing ≥70 kg. The role of sex as a potential effect modifier is less clear. Although sex was not identified as one in in the ARRIVE, ASCEND or ASPREE trials, results from an older meta-analysis, which included the results of 6 RCTs (Berger et al. 2006) suggested that ASA reduced the risk of myocardial infarction only in men, and the risk of all stroke and ischemic strokes only in women.

ASA has also been evaluated with respect to its efficacy for reducing the risk of cardiovascular events including TIA, stroke, myocardial infarction, unstable angina or death among persons with asymptomatic atherosclerosis.  While it has been suggested that low-dose ASA leads to a change in the composition of plaque within blood vessels, transforming it from a soft foamy material to a harder material that is less likely to rupture, and reduces inflammation, the use of 325 mg of daily ASA for two years in persons with carotid stenosis (≥50%) was not associated with reductions in vascular events compared with placebo (Cote et al. 1995). Nor was the risk of the composite of initial fatal or nonfatal coronary event or stroke or revascularization reduced among persons with an ankle brachial index ≤ 0.95, randomized to receive 100 mg ASA daily for an average of 8.2 years in the Aspirin for Asymptomatic Atherosclerosis trial (Fowkes et al. 2010). The results of trials examining the use of ASA therapy for the primary prevention of cardiovascular events for persons with peripheral artery disease have also been negative (Belch et al. 2008, Catalano et al. 2007).

Since cerebrovascular disease is known to play a pivotal role in the development and progression of mild cognitive impairment, it has been suggested that ASA might prevent cognitive impairment or slow worsening of cognitive function. Unfortunately, the evidence base does not support such a role. After an average of 6 years, ASA treatment did not reduce the odds of the development or cognitive impairment, nor was it associated with better global cognitive test scores among 36,196 participants who were cognitively intact at baseline (Veronese et al. 2017). In the REGARDS study, (Kelley et al. 2015) the odds of cognitive impairment were not significantly higher among non-aspirin users who were cognitively normal at baseline, after a mean duration of follow-up of 5.9 years (OR = 0.99, 95% CI = 0.89–1.09). Negative results were also reported in subgroups analyses from the Aspirin for Asymptomatic Atherosclerosis Trial (Price et al. 2008) and a cohort study within the Women’s Health Study (Kang et al. 2007).

The Role of Shared Decision Making

Long-term use of ASA for primary prevention of vascular disease is not recommended for the primary prevention of a first vascular event in the current Canadian Stroke Best Practices Recommendations. Nevertheless, there is a recognition that the decision to initiate ASA therapy should be highly individualized and should be made following an assessment of the benefit/risk ratio and a clinician-patient discussion regarding potential benefits/harms, and alternatives. This process of shared decision making (SDM) is based both on clinical evidence and the patient’s informed preferences and values (Charles et al. 1997; Munro et al. 2016), recognizing that many patients want to actively participate in decisions about their own medical care.

To facilitate SMD, interventions can target the clinician, the patient, or both. Examples of activities that target clinicians include educational meetings, educational material, and educational outreach visits, while examples of activities that target patients include decision aids, pamphlets/leaflets, videos and education sessions. Unfortunately, there are no studies that have evaluated SMD for ASA use in the primary prevention of cardiovascular disease although a few case studies have been presented and a decision support algorithm and mobile application for use by physicians, described (Mora et al. 2016). Montori et al (2003) examined the personal characteristics and preferences that affected the decision to take ASA to reduce cardiovascular risk among a group of 206 patients with diabetes attending an outpatient clinic. Of the participants surveyed, 67% (n=146) were using ASA. Those using ASA were at higher risk of cardiovascular disease, knew more about the benefits of ASA, but less about the risks, and placed a higher value on preventing cardiovascular events than on avoiding the side effects.

In the broader context, the evidence for SDM interventions is weak. A Cochrane review included the results of 87 studies (Légaré et al. 2018), of which 44 targeted patients, 15 evaluated interventions targeting healthcare professionals and 28 studies targeted both patients and healthcare professionals. The authors concluded that it was uncertain whether activities to enable SDM are effective compared with usual care because the certainty of the evidence was low or very low. In this review, a wide variety of medical conditions were represented including cancer, dementia, fibromyalgia, and mental health issues. However, another Cochrane review, (Stacey et al. 2017) that included the results of 86 randomized controlled trials (RCTs), specifically examined the effectiveness of patient decision aids (PDA). Patients who were exposed to decision aids were more knowledgeable, better informed and clearer about their values leading to having a more active role in decision making and had more accurate risk perceptions.

There is a substantial literature examining the use of SDM in the context of cardiovascular risk factor reduction, the results of which have been equivocal. A systematic review of 6 studies (5 RCTs) that included adults in primary care being treated for hypertension, compared the effects of shared decision-making interventions versus any comparator, targeting either the patient or physician reported that interventions did not increase measures of SMD, patient participation or blood pressure (Johnson et al. 2018). Following a 6-hour, multicomponent program of SDM training for 36 general practitioners (GP), patients in their practices with treated but uncontrolled hypertension reported no significant changes in perceived participation (SDM-Q-9), systolic blood pressure, diastolic blood pressure, knowledge, medication adherence or cardiovascular risk score compared with GPs who treated their patients as usual, at 6, 12 and 18 months (Tinsel et al. 2013). Persons with diabetes established on metformin but with persistent hyperglycemia who were recommended to consider medication intensification had significantly larger knowledge gains from baseline (35.0% vs. 9.9%, p < 0.0001) and greater improvements in Decision Self Efficacy Scale scores (3.7 vs.−3.9, p < 0.0001) and Decisional Conflict Scale scores (−22.2 vs. −7.5, p < 0.0001) following exposure to an online PDA (Bailey et al. 2016). Among a group of primary care patients, 20% of whom had a previous cardiovascular event, there were significantly higher levels of satisfaction and participation and lower decisional regret reported when treated by physicians who had attended two interactive continuing medical education sessions that focused on cardiovascular risk reduction (Krones et al. 2008)