Low-Dose Aspirin for Individualized Cancer Prevention in Older Adults: A Secondary Analysis of the ASPREE Randomized Clinical Trial

Importance The role of low-dose aspirin (LDA) in cancer prevention among older adults is unclear. Identifying individuals likely to experience benefit or harm is crucial for guiding personalized prevention strategies. Objective To identify subgroups of the older population who may benefit most from LDA for cancer prevention by developing and validating an effect score model, including individual characteristics such as clonal hematopoiesis of indeterminate potential (CHIP) and other factors, to predict the individualized treatment effects (ITEs) of LDA in cancer incidence. Design, Setting, and Participants This comparative effectiveness study is a nonprespecified secondary analysis of the Aspirin in Reducing Events in the Elderly (ASPREE) randomized clinical trial conducted between 2010 and 2014. Australian community-dwelling ASPREE participants (aged ≥70 years) provided biospecimen samples for CHIP testing. CHIP was measured in 18 genes using a targeted sequencing assay. Analyses were conducted between May 2023 and July 2025. Exposures LDA (100 mg per day) or placebo. Main Outcomes and Measures The primary outcome was overall cancer incidence at 5 years. Twelve candidate models for predicting ITEs were assessed; the final model was selected to maximize cancer risk reduction and internally validated using bootstrapping. Participants were grouped into treatment-favorable (ITE >0) and treatment-unfavorable (ITE <0) subgroups based on the final predictive model. Results A total of 9350 participants were included in the analysis (median age, 73.7 [IQR, 71.6-77.1] years; 5021 females [53.7%]) to LDA (4667 [49.9%]) or placebo (4683 [50.1%]). The median follow-up was 4.5 (IQR, 3.3-5.5) years. Factors associated with LDA benefit included older age, nonsmoking status, CHIP with a variant allele frequency of 10% or greater, family history of cancer, and lower body mass index. CHIP was the strongest predictor of benefit. Personalized treatment based on the model improved 5-year absolute cancer risk reduction by a median of 2.3% (IQR, 0.7%-3.7%) compared with a treat-all approach (LDA to all participants). LDA was associated with lower cancer risk (hazard ratio [HR], 0.85 [95% CI, 0.72-1.00]) in the treatment-favorable subgroup and higher cancer risk (HR, 1.14 [95% CI, 0.95-1.38]) in the treatment-unfavorable subgroup (P = .02 for heterogeneity). Conclusions and Relevance The findings of this study suggest significant heterogeneity in the effectiveness of LDA for cancer prevention among older adults, with CHIP identified as a key predictive factor. Further study is needed to fully understand the implications of these findings.