In stroke patients with atrial fibrillation (AF) and cardiovascular disease (CVD), adding an antiplatelet to an oral anticoagulant (OAC) does not reduce ischemic events and more than doubles bleeding risk, results of a randomized clinical trial, which was halted for futility, showed.

“These findings suggest that an anticoagulant monotherapy strategy appears safer for secondary stroke prevention in this high-risk population,” study investigator Hiroshi Yamagami, MD, PhD, Division of Stroke Prevention and Treatment, Institute of Medicine, University of Tsukuba, Tsukuba, Japan, told Medscape Medical News.

The trial was terminated for futility after the planned interim analysis.

The findings were published online on October 6, 2025, in JAMA Neurology.

Lack of High-Quality Evidence

Nonvalvular AF (NVAF) and atherosclerotic CVD (ASCVD) share common risk factors, such as hypertension and diabetes, which are frequently comorbid. Individuals with ischemic stroke and both NVAF and ASCVD face an elevated risk for recurrent stroke.

The optimal treatment approach for these patients remains uncertain. OACs are standard for preventing cardioembolic events, while antiplatelet therapy is typically used to manage atherosclerotic disease.

“In patients who have both AF and atherosclerotic cardiovascular disease, clinicians sometimes add an antiplatelet agent to cover the atherosclerotic component, hoping to further reduce ischemic events,” said Yamagami.

“The challenge is that combining agents increases bleeding risk, and high‑quality evidence about the net benefit in this dual‑pathology population has been limited,” he added.

The ATIS-NVAF trial included 316 patients with NVAF and ASCVD — defined as ischemic stroke due to large artery atherosclerosis or small vessel occlusion, intracranial or carotid stenosis, or coronary or peripheral arterial disease — across 41 centers in Japan. Participants had a mean age of 77.2 years, and 28.5% were women.

Participants had experienced an ischemic stroke or transient ischemic attack within the past 8 days up to 1 year.

Researchers randomly assigned participants to either OAC monotherapy or combination therapy with an anticoagulant plus an antiplatelet. The median time from stroke onset to randomization was 21 days.

The choice of anticoagulant and antiplatelet agents was at the discretion of the treating physician. Treatment was initiated within 3 days of randomization and was to continue for 2 years.

Apixaban was the most prescribed anticoagulant, used in 38% of the combination group and 46% of the monotherapy group, while warfarin was used in 6% of each group. In the combination therapy group, the most common antiplatelet agent was aspirin (52%), followed by clopidogrel (31%) and cilostazol (17%).

Clinical assessments were conducted at baseline and at 6, 12, 18, and 24 months. All events were adjudicated by a blinded, independent central review committee.

The primary outcome was a composite of ischemic cardiovascular events — including stroke, myocardial infarction, vascular death, systemic embolism, or urgent revascularization — or major bleeding.

An interim analysis for efficacy and futility was planned at 50% of the required events.

The cumulative incidence of the primary outcome was 17.8% in the combination therapy group and 19.6% in the monotherapy group (hazard ratio [HR], 0.91; 95% CI, 0.53-1.55; P = .64).

The primary outcome did not differ significantly across subgroups, including age, sex, history of ischemic stroke or major bleeding, presence of carotid or intracranial artery stenosis, hypertension, diabetes, dyslipidemia, and congestive heart failure.

For secondary outcomes, ischemic cardiovascular events occurred in 11.1% of the combination therapy group and 14.2% of the monotherapy group (HR, 0.76; 95% CI, 0.39-1.48; P = .41). Most events were ischemic strokes, occurring in 9.2% of the combination group vs 13.0% of the monotherapy group (HR, 0.67; 95% CI, 0.32-1.39; P = .28).

Unexpected Bleeding Rate

The modest reduction in ischemic cardiovascular events with combination therapy was offset by a substantially higher bleeding risk. Major or clinically relevant nonmajor bleeding occurred in 19.5% of the combination group vs 8.6% of the monotherapy group (HR, 2.42; 95% CI, 1.23-4.76; P = .008).

The results for cardiovascular events were disappointing but not particularly surprising, although the higher bleeding rates were somewhat unexpected, said Yamagami. “We had anticipated a significant reduction in bleeding events.”

Yamagami noted that prior trials generally favored direct OAC monotherapy over combination therapy because it reduced bleeding without compromising efficacy. This new trial differs by enrolling stroke patients with both AF and ASCVD, a group at particularly high risk for early recurrence.

The most notable hemorrhagic events in the study were gastrointestinal bleeding, occurring in 6% of the combination group vs 3% of the monotherapy group, followed by intracranial hemorrhage (4% vs 3%), nasal bleeding (4% vs 0%), and hematuria (2% vs 0%).

The study’s open-label design may have introduced bias, potentially affecting patient adherence, outcome reporting, and clinical decision-making. Additional limitations included allowing treating physicians to choose the type and dosage of OACs and antiplatelet agents and enrolling only Japanese patients, among whom genetic and environmental factors may influence response to antiplatelet therapy.

Yamagami and colleagues plan several follow-up studies to explore whether specific subgroups — such as younger patients or individuals with well-controlled blood pressure — might benefit from adding antiplatelet therapy.

They also plan to investigate alternative strategies to reduce bleeding, including left atrial appendage closure, catheter ablation combined with antiplatelet therapy, and factor XIa inhibitors — an emerging class of targeted anticoagulants that may prevent clot formation without substantially increasing the risk for bleeding.

In addition, they plan to assess whether a very short, targeted course of antiplatelet therapy for specific vascular indications could meaningfully reduce risk without excess bleeding.

‘Uncertain and Unproven’ Benefit

Commenting for Medscape Medical News, Amy Guzik, MD, associate professor of neurology and director of Stroke and Telestroke Services at Wake Forest University School of Medicine in Winston-Salem, North Carolina, noted that the study represents an important addition to the literature and that its findings could be useful across a range of patient populations. 

However, she raised concerns about certain aspects of the trial, noting that most participants’ coexisting atherosclerotic disease was atherothrombotic or lacunar stroke, which typically stems from small vessel disease deep within the brain.

Clinically, these strokes are usually attributed to uncontrolled risk factors such as hypertension, and most neurologists would not routinely add antiplatelet therapy on top of anticoagulation, unlike in cases of large-vessel disease in the coronary, carotid, or major intracranial arteries, said Guzik.

Participants with symptomatic intracranial or carotid artery stenosis and ischemic heart disease represented only a small portion of the study population, making it difficult to generalize the results to these subgroups, said Guzik. She also noted that fewer than 30% of participants were women.

In an accompanying editorial, Richard A. Bernstein, MD, PhD, Department of Neurology, Northwestern University, Chicago, and a colleague said the trial shows the combination therapy in patients with stroke and AF significantly raises the bleeding risk in return for what at best is an “uncertain and unproven” benefit.

“(B)lithely adding antiplatelet agents to OAC in stroke patients should be generally avoided,” they wrote. “The benefits are unclear, and the risks are severe.”

The trial design included certain “quirks and compromises” that affect interpretation, including classifying small vessel strokes as atherosclerotic disease — something most neurologists would not do, they wrote.

The trial design had some “quirks and compromises” that “color” interpretation, one of which was counting small vessel stroke as atherosclerotic disease, which is something most neurologists wouldn’t do, they wrote.

The inclusion of both symptomatic and asymptomatic large vessel stenoses is also “troublesome,” as the risk for stroke from the former is much higher, they added.

“Again, many neurologists do not include asymptomatic large vessel atherosclerosis in their treatment choice algorithm. The authors’ comment that separating those with symptomatic vs asymptomatic atherosclerosis did not affect the results, but in a study this small, that is hard to interpret.”

In addition, the combination of ischemic events and major bleeding as a measure of net benefit should be interpreted cautiously, they noted. “Most patients would prefer a gastrointestinal bleed to a recurrent stroke or myocardial infarction, so weighing the two as equivalent may not be appropriate.”

The study was partially supported by the Japan Thrombosis Investigator-Initiated Research Program, funded by Bristol Myers Squibb and Pfizer. Yamagami reported receiving grants from Bristol Myers Squibb during the conduct of the study and receiving lecture fees from Abbott Medical Japan, Boston Scientific Japan, Bristol Myers Squibb, Daiichi Sankyo, Medtronic, Otsuka Pharmaceutical, and Stryker outside the submitted work. Guzik reported having no relevant conflicts of interest.