The aspirin balance in primary cardiovascular prevention

The use of aspirin (ASA) in the primary prevention of cardiovascular diseases has been discussed in recent years for the uncertainty deriving from the results of the studies, so much so that the most recent guidelines do not recommend their use¹.
The variegated set of major cardiovascular events recognises thromboembolic phenomena as common elements, in which platelets have a prominent role, which determine their clinical manifestation. On this basis the use of ASA has been proposed whose action, at 100 mg per day, is to block the synthesis of thromboxane in the platelets through a selective inhibition of type 1 cyclooxygenase.

The risk-benefit assessment in studies

Acting on the haemostatic process by itself justifies the excess of bleeding events attributable to the use of antiplatelet drugs in clinical trials and in clinical practice. So not only the expected benefit, but also the balance between this and the risk of bleeding are the parameters on which to evaluate the opportunity to recommend in the population the chronic consumption of ASA. These considerations contained in the meta-analysis work of the Antiplatelet Trialists' Collaboration² and the absence of evidence from more recent studies conducted in diabetic patients, who have a greater cardiovascular risk than the general population, are the assumption that justified recently published randomised clinical studies.
In all of these, ASA was administered at a dose of 100 mg per day. The ASCEND³ study evaluated 15,480 subjects with diabetes mellitus, followed for an average period of 7.4 years, the appearance of a cardiovascular event identified as non-fatal myocardial infarction, non-fatal non-haemorrhagic stroke, transient ischaemic attack or death from any vascular cause (combined outcome). The number of events recorded was lower than expected. There was a benefit with ASA compared to placebo (0.88 ratio, IC95% from 0.79 to 0.97); however, this was associated with a parallel increase of roughly double in bleeding events. The net balance in favour of ASA was therefore very limited since it was necessary to treat 91 subjects to prevent a cardiovascular event, while for every 112 subjects treated a bleeding event may be expected.
Two studies were conducted in representative subjects of the general population in which the highest risk factors was age over 70, hypertenson (ASPREE study) or frequent adult morbilities such as arterial hypertension and dyslipidaemia, but not diabetes mellitus (ASPIRE study). The results of the ASPREE study on 19,114 elderly subjects followed for an average of 4.7 years were published in three different articles published in the same issue of the New England Journal of Medicine.^4-6 The mortality data were analysed separately, including those for tumour, the data of survival without incapacity (defined as dementia and physical disability) and, finally, major vascular events defined by a composite outcome similar to that of the other studies. There was no difference between treatment with ASA and placebo for any of the proposed aims, except for the highest overall mortality in the group of subjects assigned to the ASA group (risk ratio 1.14, IC95% from 1.01 to 1.29) for an excess of mortality from cancer, data that was unique, compared to previous studies. The number of bleedings was significantly higher in subjects of the ASA group (3.8% compared to 2.8%, risk ratio 1.38, 95% confidence limits from 1.18 to 1.62). Finally, the ARRIVE⁷ study evaluated the effect of acetylsalicylic acid in 12,452 subjects with moderate cardiovascular risk (mean age 63.9 years) followed for 60 months. Also in this study there was no significant effect on cardiovascular events defined by the same composite outcome as previous studies: 269 (4.29%) in patients in the ASA group versus 281 (4.48%) in the placebo group (risk ratio 0.96, IC95% from 0.81 to 1.13, p=0.6038). Mainly mild intestinal bleeding occurred in 61 (0.97%) subjects assigned to the ASA group compared with 29 (0.46%) in the placebo group (risk ratio 2.11, IC95% from 1.36 to 3.28, p=0.0007), with no difference in the total number of adverse events between the two groups. Once again, as in the ASCEND study, the cardiovascular events detected were lower than expected (20-30% at 10 years).

Considerations for clinical practice

What considerations can be transferred to clinical practice? First of all, the new studies were expected to confirm or deny what was already anticipated by the meta-analyses carried out on the basis of studies concerning subjects with low cardiovascular risk, in which different doses of ASA or other antiplatelet drugs were used. They also confirm the lack of benefit even for a condition considered previously to represent a very high cardiovascular risk such as diabetes mellitus.
The guidelines published two years ago by a joint initiative of European scientific societies active in the field of cardiovascular1 diseases already contained the recommendation not to treat with ASA subjects who did not present cardiovascular disease or did not require antiplatelet treatment for acute cardiovascular events; they also advised against the prescription of ASA to patients with diabetes mellitus. The result of the studies described provides an updated view of the changes in the last three decades in the cardiovascular risk profile of the population and the widespread availability of drugs, with an overall more effective treatment of frequently associated clinical conditions such as arterial hypertension, diabetes mellitus and hypercholesterolaemia which represent the causative factors of vascular damage. This is also associated with a probable greater awareness of the risk population and of the importance of changes to lifestyle, such as the reduction in the number of smokers. Therefore, connected factors that are improving the epidemiology of cardiovascular diseases allow the intervention on the platelet function to no longer represent a beneficial therapeutic target, while drugs with a more favourable risk profile can take its place, such as statins, which despite acting on different molecular targets, reduce thromboembolic events.
Although the residual cardiovascular risk is higher than that of the general population, the benefit for the general population of diabetics does not seem to justify the long-term prescription of ASA. The studies also confirm that the risk of gastrointestinal bleeding with ASA, if administered at a dose of 100 mg per day, is present but relatively low, even if this data is not directly transferable to the risk attributable to subjects for whom the use of ASA is strongly recommended, i.e. those at high cardiovascular risk for whom the risk of bleeding may be higher². We can therefore accept Ridker’s witty remark in the editorial accompanying the ASPREE study: “The best strategy for the use of ASA in the primary prevention of cardiovascular diseases may be prescribing a statin in its place”, remembering that the firm indication of the prescription of ASA for the secondary prevention of cardiovascular diseases remains.