There was no significant difference in cardiovascular events or major bleeding between patients with established coronary heart disease (CHD) assigned to a daily aspirin dose of 81 mg and those receiving a dose of 325 mg in the 15,000-patient ADAPTABLE trial.
Although substantial dose switching occurred in the trial, particularly from the higher to the lower dose, lead investigator W. Schuyler Jones, MD, believes the results support the use of the 81-mg dose in most patients.
“While we would have liked to see higher adherence to the assigned doses, we think the results of the trial are reliable,” Jones told theheart.org | Medscape Cardiology.
The real-world, open-label, pragmatic trial also involved an innovative low-cost design allowing researchers to identify and communicate with eligible patients directly, opening up a new cost-effective method to conduct clinical research going forward.
Jones, a cardiologist and associate professor of medicine at Duke University Medical Center, Durham, North Carolina, presented the ADAPTABLE results on May 15 at the virtual American College of Cardiology (ACC) 2021 Scientific Session. They were simultaneously published online in the New England Journal of Medicine (NEJM).
Jones noted there were mixed signals in the results. “For example, the main intent-to-treat analysis showed a trend to a lower rate of all-cause death in the 81-mg group, but the subgroup of patients who stayed on the 325-mg dose throughout the study had a lower event rate. But overall, there was no difference.”
Jones said the investigators had the following take-home messages to patients: “If a patient is already taking 81 mg, staying on this dose is probably right given the similar study results for the primary endpoint and that we didn’t find conclusive evidence that 325 mg is better; but for patients who have tolerated 325 mg long term, then they may want to stay on this dose as it may be associated with moderate benefit.”
Jones pointed out that, overall, patients who switched doses tended to do worse, but he suggested this may have been more to do with underlying reasons for switching rather than the different dose itself. “For example, switching often happens after bleeding or bruising, which can also often pre-empt an ischemic event, and other illnesses, such as cancer or AF [atrial fibrillation] can also lead patients to change doses.”
“With the caveat that this trial did not include new patients (the vast majority of patients had been taking aspirin previously) the results support the approach of starting new patients on 81 mg, which is what we have been seeing in evolving clinical practice in recent years,” he added.
Jones explained that the trial set out to answer the simple but important question about the best dose of aspirin in patients with heart disease.
“Aspirin has been established as an appropriate long-term medication for patients with ischemic heart disease since the 1980s, but we really don’t have any good information on the correct dose.”
He noted that the U.S. guidelines suggest any dose in the range of 81 mg to 325 mg daily can be used, whereas the European guidelines recommend 81 mg daily, although this is mainly based on observational data and expert opinion; there is little hard, randomized-trial evidence.
The ADAPTABLE trial randomly assigned 15,076 patients with established heart disease to receive 81 mg or 325 mg of aspirin. Before randomization, 96% of those with available information were already taking aspirin, 85% of whom were taking 81 mg.
After a mean follow-up of 26 months, the primary efficacy endpoint — a composite of all-cause death, myocardial infarction (MI), or stroke — had occurred in 7.28% of the 81-mg group and 7.51% of the 325-mg group (hazard ratio [HR], 1.02; 95% CI, 0.91 – 1.14).
The main safety endpoint, hospitalization for major bleeding with an associated blood transfusion, occurred in 0.63% of the 81-mg group and 0.60% of the 325-mg group (HR, 1.18; 95% CI, 0.79 – 1.77).
“The bleeding safety endpoint looked similar, which may be counterintuitive to what may have been expected,” Jones commented. “However, the safety endpoint was very stringent. We still haven’t analyzed all the less serious ADR [adverse drug event]/bleeding data, but overall, it does appear to be balanced.”
He added: “Most cardiologists probably may not have expected to see much difference in efficacy between these two doses but would maybe have anticipated a lower bleeding rate with the low dose. I was a little surprised to see such a low bleeding rate in the 325-mg group.”
Patients assigned to 325 mg had a higher incidence of dose switching (41.6%) than those assigned to 81 mg (7.1%) and were more likely to discontinue treatment (11.1% vs 7.0%). This resulted in fewer median days of exposure to the assigned dose in the 325-mg group (434 vs 650 days).
“This was an open-label study, and such studies always suffer from a degree of infidelity to the assigned treatment group,” Jones said. “In ADAPTABLE, this was unbalanced in that a much greater number of patients switched from 325 mg to 81 mg than the other way round.”
“But our results do reflect what happens in normal life,” he added. “People behaved in the study like they do in the real world. They sometimes changed their dose and sometimes stopped taking aspirin altogether. So, I think the results are an accurate representation of the real world.”
A sensitivity analysis based on which dose the patient actually reported taking showed a higher risk for death, MI, or stroke in patients who took 81 mg than those who took 325 mg (HR, 1.25; 95% CI, 1.10 – 1.43). But as with any postrandomization analysis, this approach has many inherent biases, Jones cautioned.
Innovative Study Design
The ADAPTABLE study used an innovative low-cost design, which involved direct communication with the patients themselves.
Using the National Patient-Centered Clinical Research Network (PCORnet), a group of 40 U.S. centers committed to compiling data in a common format, invitations to enroll in the study were sent to eligible patients identified from medical records. Consent and randomization took place on the patient web portal.
Participants then purchased aspirin at the assigned dose themselves, and all follow-up was done virtually or on the phone, with outcomes ascertained remotely (from patient reports, electronic medical records, and insurance claims) without adjudication.
“This is a pretty neat way to do clinical research, enabling us to conduct a 15,000-patient trial on a very tight budget,” Jones commented.
He estimated that the trial cost around $18 to $19 million. “No industry funder would have sponsored such a study of aspirin, and a typical trial with this many patients conducted in the traditional way would have cost at least 5 or 10 times more,” he said.
“This is the first time this type of study has been done in the U.S. on such a large scale, and it opens up this method for future research.”
He explained that this design, communicating directly with patients, somewhat limits the questions that can be addressed. “As aspirin is purchased over the counter by patients themselves, this is a question than leant itself to be answered in this way.”
Another innovative design feature was the inclusion of “patient partners,” with one patient nominated by each center to be part of the organization of the trial. “This helped keep the research relevant to what patients care about.
They also helped with the recruitment strategy and communication with participants. I think this is something we need to continue and prioritize in clinical research going forward,” Jones noted.
Discussants of the study at the ACC presentation congratulated the investigators on conducting such an innovative trial.
Donald Lloyd-Jones, MD, chair of preventive medicine at Northwestern University Feinberg School of Medicine, Chicago, Illinois, said, “This is really a pioneering large pragmatic trial, and we’re going to need to see more of these over the next few years. The most important legacy from this trial for me is that you did it, and that you showed us many of the promises and some of the pitfalls of these large pragmatic designs.”
Akshay Desai, MD, associate professor of medicine, Harvard Medical School, Boston, Massachusetts, added: “This was an innovative approach to answering an important question for daily clinical practice.”
On the results of the study, Lloyd-Jones said, “Maybe the outcomes were not too surprising, and I certainly endorse your cautious status quo statement about patients staying on the dose that they are on.”
But he suggested that the bleeding safety outcomes were perhaps a little unexpected, being a little lower in the lower-dose group, and he asked whether there was a sensitivity analysis looking at bleeding on a per protocol basis. Jones answered that this was planned.
Desai also raised questions about the “unusual bleeding endpoint,” noting that the rates of bleeding were far lower than would be expected compared with other clinical trials.
Jones replied that the bleeding endpoint with blood product transfusion was chosen to allow the researchers to accurately identify these events in claims codes. He said the endpoint probably mirrored the GUSTO severe bleeding classification.
In an editorial accompanying the publication of ADAPTABLE in the NEJM, Colin Baigent, FMedSci, says the study provides proof of principle that large pragmatic randomized trials can be conducted in the United States.
But Baigent, who is professor of epidemiology and director of the Medical Research Council Population Health Research Unit at the University of Oxford, United Kingdom, says that the high degree of switching between dosages that occurred during the trial gives rise to some uncertainty about the results.
“Because switching was not likely to have been at random, bias arising from this degree of crossover could have obscured a true difference in efficacy or safety (or both), and moreover it is also not possible to conclude that the lack of any significant difference between the two dose groups implies equivalence of the effects of the doses,” he writes.
He suggests that a pilot study may have identified a preference for the 81-mg dose and allowed methods to facilitate equipoise, such as a run-in period with both doses, and only patients adhering being considered for randomization.
But Baigent concludes that the ADAPTABLE trial is a “major achievement” in that it paves the way for low-cost randomized trials in the United States, which should allow many more clinical questions to be answered.
The trial was supported by an award from the Patient-Centred Outcomes
American College of Cardiology (ACC) 2021 Scientific Sessions. Presented May 15, 2021.