Aspirin Vs Placebo In Cardiovascular Event Reduction In Diabetes ESC 2018: ASCEND trial

A randomized 15,480 patient strong UK based trial comparing Aspirin (ASA) versus placebo in diabetics has shown that the absolute reduction in cardiovascular events from aspirin is offset by a similar absolute increase in major bleeding.

Aspirin is universally accepted as a therapy in the secondary prevention of cardiovascular disease. In healthy individuals with low cardiovascular risk, the benefits of ASA are largely outweighed by the increased risk of bleeding. For diabetics who have a higher cardiovascular risk as compared to nondiabetics, there has been major uncertainty about the value of ASA over the years. Additionally, ASA has different rates of use in different parts of the world. There is a stark difference in the guidelines with the US guidelines being more encouraging of ASA for primary prevention, in contrast to the European guidelines which are more cautious. Lately, there is also emerging evidence that ASA might be able to protect patients against cancer. In order to address the vascular question, and simultaneously test the cancer question, Armitage et al. conducted a randomized, factorial design, pragmatic trial to evaluate aspirin compared with placebo among diabetics with no known cardiovascular disease (CVD).

Out of the total of 15, 480 patients, diabetics ≥40 years of age with no known CVD were randomized to aspirin 100 mg daily (n = 7,740) versus placebo (n = 7,740). As the drug was not patentable, the trial was funded generously by the Major UK charity, the British heart foundation over 3 consecutive grants. The trial was run entirely remotely, with the help of investigators in the UK. Patients with diabetes were identified from regional diabetes registers, generally set up for diabetic eye screening and from general practice. A large number of screening questionnaires were sent around the UK and if people returned those questionnaires and indicated their interest in participation, they were invited to a 2-year pre-randomization run-in period. They were then sent placebo tablets and capsules to be taken daily. After filling yet another questionnaire, they were considered eligible for the study and subsequently randomized. While the study period was initially kept at 5 years, the duration of mean follow-up had to be extended to seven and a half years due to lower event rates than anticipated. During the study, the participants were sent a six monthly questionnaire asking about SAEs, adherence to the study treatment and other treatments of interest. For the aspirin arm particularly, the use of other antiplatelet and anticoagulants was thoroughly explored. Complete follow up of patients was ensured using a combination of participant questionnaires, contacting the GP and looking at centralized records and hospital episode statistics, if all else failed.

The primary efficacy endpoint was a composite of NSTEMI, ischemic stroke/TIA, and cardiovascular death. Ischemic stroke and cardiovascular death excluded any intracerebral or intracranial bleeding. The primary safety endpoint was a composite of a major bleed, any intracranial bleed, life-threatening eye bleeding, serious GI bleeding or any serious extracranial bleed. All bleeds were followed up and adjudicated centrally by physicians who were blind to the treatment allocation as were all the other primary endpoints.

The primary efficacy outcome occurred in 8.5% of the aspirin group compared with 9.6% of the placebo group (p = 0.01). The primary safety outcome was found to occur in 4.1% of the aspirin group compared with 3.2% of the placebo group (p = 0.003). Moreover, the reduction in adverse events or increase in bleeding events was similar across different categories of baseline risk. The rate of GI hemorrhage was 1.8% with aspirin vs. 1.3% with placebo (p < 0.05) and the rate of GI cancer was 2.0% with aspirin vs. 2.0% with placebo (p = NS). Out of 1700 patients with reported incident cancer, there was no overall effect, with an event ratio of 1.

The investigators concluded that the primary efficacy endpoint of major adverse cardiovascular events was significantly reduced by ASA allocation as compared to placebo by 12% ( p=0.01). There was a significantly high risk of major bleeding by 29% in the aspirin group, primarily due to GI hemorrhage. It was difficult to discern in any particular group where the benefits outweighed the risk. This reinforced that fact that the use of aspirin among diabetics with no known CVD needed to be individualized.

Speaking of the implications of these findings, Dr. Jane Armitage says, “We have to bear in mind that we do not want to cause harm. Having said that, the other issue is how do you balance having a major ischemic stroke against having a major bleed? They are not comparable and it becomes a matter of judgment. All we can say is, this is what the data shows.” The population of diabetics is a well managed, contemporary population of patients with 75% of patients on statins, a vast majority on blood pressure lowering medications and with a very low smoking rate among patients. All of these interventions are safe and have a major impact on cardiovascular risk. Among them, the study found no additional benefit from adding ASA.

While she believes that the results of this study will definitely impact guidelines and physicians’ judgments worldwide, Dr. Armitage stresses on the fact that “The primary focus should be on getting statins and blood pressure lowering medications well established to minimize cardiovascular risk. If the risk is still high, there might be an argument for considering aspirin. But the focus really needs to be on ensuring patients’ cardiovascular risk is reached using safe, evidence-based treatments.”

Source: https://www.nejm.org/doi/full/10.1056/NEJMoa1804988