What have we learned about using aspirin for primary prevention from the ASCEND and ARRIVE trials?

Abstract

Subjects with diabetes mellitus are at increased cardiovascular disease (CVD) risk. Particularly, diabetic patients with additional risk factors or target organ damage are considered as risk equivalent to patients with known coronary artery disease and are eligible for aggressive treatment prevention strategies. However, the role of aspirin in the primary prevention of CVD in diabetic patients remains an issue of controversy. In this setting, the results of the recently published ASCEND trial have been long awaited. ASCEND was a large clinical trial that randomized a total of 15 480 patients with diabetes mellitus with no known vascular disease to aspirin or placebo, with prospective follow-up for 7.4 years to test the effects of aspirin on vascular events (i.e. myocardial infarction, stroke or transient ischaemic attack, or death from any vascular cause, excluding any confirmed intracranial haemorrhage) and major bleeding events (i.e. intracranial haemorrhage, sight-threatening bleeding event in the eye, gastrointestinal bleeding, or other serious bleeding). Although aspirin significantly lowered vascular events [relative risk (RR) = 0.88; 95% confidence interval (CI) 0.79–0.97], it also increased major bleeding events (RR = 1.29; 95% CI 1.09–1.52) mainly due to excess gastrointestinal bleeding and other extracranial bleeding. In pre-specified subgroup analyses there was no evidence of variation in the effect of aspirin according to baseline characteristics, the use of n-3 fatty acids or estimated vascular risk. Also, previous reports of greater benefit of aspirin use among persons with a body weight <70 kg were not confirmed. The effects of aspirin on all-cause mortality and cancer risk were neutral. Importantly, the incidence of a major bleeding event increased with baseline vascular risk, and thus the predicted number of serious vascular events that would be avoided by aspirin was closely balanced by the predicted number of bleeding event caused. Thus, overall any absolute benefits in the reduction of vascular events by preventive aspirin administration in diabetic persons were counterbalanced by an increase in bleeding events. Interestingly, similar negative findings were simultaneously reported by the ARRIVE trial, another randomized control trial on aspirin use in primary prevention; in a total of 12 546 patients at moderate cardiovascular risk with no diabetes or incident coronary artery disease, aspirin did not have any significant effects on cardiovascular events. A number of observations can be made about these findings relevant to CVD prevention. First, a remarkable finding in ASCEND was that, in contrary to opposite expectations, the majority of deaths in diabetic patients were due to non-vascular causes. This suggests that current risk management strategies with statins, antihypertensive medication, and possibly newer antidiabetic agents too, have radically changed the landscape and clinical course of diabetic patients, who are nowadays better protected against CVD development. This was confirmed by the actual rate of events in these studies, which was lower than originally anticipated based on clinical risk score estimates. This is an important observation, which relates to the systematic overestimation of risk by the use of clinical risk score calculators. Notably, aspirin did reduce vascular events in diabetic patients, and therefore, the rationale for its use remains solid in these populations. However, the efficacy of preventive treatment interventions is highly dependent on the actual event risk of the population of interest. Therefore, tools for more accurate risk assessment could be used to identify the subjects at the highest risk to guide the deployment of preventive measures such as aspirin. Recently, novel imaging modalities for the estimation of coronary inflammation and the residual inflammatory risk, such as perivascular fat attenuation index by standard coronary computed tomography angiography, could have a role in the risk stratification for coronary events, as recently suggested by the CRISP CT study. Certainly, more evidence is needed on this issue and further translational research in the field. Another issue that warrants further investigation in basic research studies relates to aspirin non-responsiveness in patients with diabetes. A recent study suggested that the enteric coating of aspirin significantly reduces its bioavailability and its platelet inhibition efficacy in diabetes. Ways to maximize the platelet inhibitory effects of aspirin (without raising bleeding risks) and strategies to overcome aspirin resistance in patients with diabetes mellitus would be beneficial for millions of people worldwide. Interestingly, the postulated links between platelet activation and cancer were not confirmed by the ASCEND trial. Use of aspirin did not reduce the risk of gastrointestinal or any other form of cancer. US Preventive Services Task Force currently recommends aspirin for the primary prevention of colorectal cancer, but the cross-talk between platelets and cancer and the role of chemoprevention with aspirin may need to be further examined. In conclusion, the negative results of the recent randomized trials in the primary prevention of CVD with the use of aspirin in people with