23 Outcomes following aspirin discontinuation in chronic aspirin users: a study-level meta-analysis

• Published in: Heart (British Cardiac Society) Vol. 109; no. Suppl 6; pp. A25 - A26
• Main Authors: Campbell, R, Nelson, M, McNeill, J, McEvoy, J W
• Format: Journal Article
• Language: English
• Published: London BMJ Publishing Group LTD 01.10.2023
• Subjects: Aspirin; Meta-analysis
• ISSN: 1355-6037; 1468-201X
• DOI: 10.1136/heartjnl-2023-ICS.23

IntroductionBased on recent evidence (1–3), guidelines no longer recommend aspirin for primary prevention of cardiovascular disease (CVD). Implementing these guidelines may result in a number of patients discontinuing this medication. These trials predominantly tested the null hypothesis that, among persons naïve to aspirin (i.e. non-users), randomization to start aspirin vs placebo has no impact on the occurrence of a first CVD event. A related but unanswered question, and the basis of this meta-analysis, is the effect of aspirin discontinuation among chronic users.MethodsA subgroup study-level meta-analysis of baseline aspirin users (n=7222) in the ASPREE (ASPirin in Reducing Events in the Elderly) (1) and ASCEND (Effects of Aspirin for Primary Prevention in Persons with Diabetes Mellitus) trial (3) was performed. Outcomes included fatal and non-fatal CVD events and major bleeding. We used a random-effects model with restricted maximum likelihood for the meta-analysis. We used I2 statistics to assess for heterogeneity across trials. All analyses were conducted with Stata 16 (StataCorp., College Station, TX, USA). P-values <0.05 (two-sided) were considered statistically significant.ResultsThe results of this study suggest an increased risk of CVD events among aspirin users randomised to cease aspirin and take placebo vs aspirin users randomised to continue aspirin (HR 1.21, 95% CI 1.05 to 1.39, p=0.01). This risk is evident shortly after discontinuation and persists over time, suggesting an unmasking of anti-thrombotic protection rather than any rebound hypercoagulable state (4). In terms of major bleeding events, no significant difference was evident among aspirin users randomised to stop aspirin vs continue aspirin (HR 0.86, 95% CI 0.68 to 1.10, p=0.24). (figure 1). Self-selected tolerance for aspirin is the most likely explanation for the lack of significant difference in major bleeding events in the aspirin continuation group.Abstract 23 Figure 1Subgroup study-level meta-analysis of aspirin users (n= 7222) randomized to stop aspirin vs to continue aspirin[Figure omitted. See PDF]ConclusionIn conclusion, we believe the results of this analysis should prompt further investigation into the effects of aspirin discontinuation in chronic users through an individual participant level meta-analysis of ASPREE, ASCEND and other historical aspirin prevention trials.