Until now the optimal dosing of aspirin in terms of efficacy and safety for secondary prevention of atherosclerotic cardiovascular disease had not been previously well studied in randomized clinical trials. US Guideline recommendations permitted doses ranging from 81 to 325 mg daily.
Rather than utilizing a traditional clinical trial design, ADAPTABLE (Aspirin Dosing: A Patient-Centric Trial Assessing Benefits and Long-Term Effectiveness) was designed to be a pragmatic, open label, randomized trial utilizing PCORnet, the US-based National Patient Centered Clinical Research Network, utilizing electronic health record systems for patient identification via a computable phenotype, data collection, and clinical event ascertainment. There was a call center option and patient-reported outcomes were also collected. The trial has 40 participating US centers. Importantly, the trial design, protocols, and all patient-facing materials had patient partners in the study providing comprehensive input and serving as advocates for the trial.
Enrollment in the trial ran from April 2016 to June 2019 during which time 450,000 eligible patients were contacted for potential enrollment and 50,245 visited the trial patient portal. 15,076 patients provided consent, were enrolled and underwent randomization to either aspirin 81 mg daily or 325 mg daily.
The primary clinical effectiveness endpoint was time to first event of death from any cause, hospitalization for myocardial infarction, or hospitalization for stroke. The primary safety outcome was hospitalization for major bleeding. Over a median follow-up of 26.2 months, there were no significant difference for either the primary effectiveness or safety endpoints in the 81 mg vs 325 mg aspirin daily arms.
This is among the first large scale pragmatic comparative effectiveness studies conducted in the US utilizing electronic health record systems for trial conduct and to evaluate a cardiovascular therapy at two different doses. The design allowed this large scale randomized trial to be conducted at a substantially lower cost compared to conventional randomized clinical trials.
However, the open label design, in which patients and their clinicians were aware of which dose of aspirin was being taken, likely contributed to a high rate of dose switching during the course of the trial. 41.6% of patients assigned to 325 mg of aspirin switched to lower dose whereas only 7.1% assigned to 81 mg switched to higher dose. This differential crossover resulted in significant differences in median days of exposure which in turn could have influenced the findings.
There is considerable interest in having clinical trials where enrollment reflects the type of patient encountered in daily clinical practice and utilizing efficient/lower cost/scalable design features. ADAPTABLE has provided an important example that multicenter trials can leverage electronic health record systems, direct to patient engagement via web portals, patient reported outcomes, and claims data. However, there are important lessons learned including ways to further improve trial design and conduct, along with the potential need for additional steps to avoid high crossover rates.