Switching to a different direct oral anticoagulant (DOAC) after a breakthrough ischemic stroke in patients with atrial fibrillation (AF) was not associated with improved short-term outcomes compared with continuing the same agent.

Results of a large multicenter cohort study showed the 90-day net clinical benefit outcome — defined as recurrent ischemic stroke or moderate-to-severe bleeding — occurred in 5.1% of patients who continued their prestroke DOAC and 4.9% of those who switched to anticoagulant therapy.

“Our study challenges a widespread but weakly justified practice — switching anticoagulants after a breakthrough stroke should no longer be considered a default response. Continuation is not only simpler but also avoids unnecessary therapeutic disruption,” study investigator Simona Sacco, MD, professor of neurology, Department of Biotechnological and Applied Clinical Sciences, University of L’Aquila in L’Aquila, Italy, told Medscape Medical News. 

The findings were published online on April 28 in JAMA Network Open.

Continuation vs Switching

DOACs are the preferred therapy for stroke prevention in patients with AF, yet breakthrough ischemic stroke occurs in approximately 1% of treated patients annually, with recurrence risks estimated at 5%-9% per year.

Despite limited randomized trial evidence, switching anticoagulants is common, often based on the assumption that an alternative agent may provide greater protection, but these events frequently reflect factors beyond treatment failure.

To address this gap, investigators conducted a multicenter, registry-based cohort study emulating a target trial design to compare outcomes associated with switching vs continuing DOAC therapy after a breakthrough event.

A breakthrough stroke was defined as one that occurred despite confirmed DOAC use within the prior 7 days, with the last dose taken within 48 hours prior to symptom onset.

The analysis included 1006 patients (mean age, 80.4 years; women, 50%) from 35 stroke centers across nine countries in Europe and North Africa between February 2020 and February 2025.

Among them, 463 patients continued their prestroke DOAC, while 543 switched to another anticoagulant, either a different DOAC or a vitamin K antagonist.

The primary outcome was net clinical benefit at 90 days. Secondary outcomes included recurrent ischemic events, symptomatic intracerebral hemorrhage, extracranial bleeding, all-cause mortality, and vascular death.

To account for baseline differences between groups, the researchers used inverse probability of treatment weighting (IPTW). A prespecified noninferiority margin of 3.0 percentage points was used for the primary outcome.

No Clinical Benefit at 90 Days

At 90 days, the primary outcome occurred at similar rates in both groups. After IPTW adjustment, the net clinical benefit was 4.9% in the switching group and 5.1% in the continuation group, corresponding to a risk difference of -0.3 percentage points.

Rates of recurrent ischemic stroke were also comparable (2.9% vs 3.1%; risk difference, -0.2 percentage points). Similarly, symptomatic intracerebral hemorrhage occurred in 0.9% of patients who switched vs 1.2% who continued therapy (risk difference, -0.4 percentage points).

Moderate-to-severe extracranial bleeding rates were similar (2.2% vs 2.7%; risk difference, -0.6 percentage points).

However, noninferiority was not demonstrated for all-cause mortality (8.7% vs 9.3%; risk difference, -0.6 percentage points) or vascular death (4.9% vs 4.4%; risk difference, 0.5 percentage points).

Importantly, no clinically meaningful advantage occurred across different switching strategies, including switching within the same class, across mechanisms, or to vitamin K antagonists.

Because this was an observational study, residual confounding cannot be excluded, and unrecorded treatment changes may have biased results toward no difference, the investigators noted. Moreover, the short 90-day follow-up and specialized centers may also limit generalizability to routine practice.

Focus on Modifiable Risk Factors

No clinically meaningful benefit was seen with switching, including between drug classes or to vitamin K antagonists, suggesting that “breakthrough strokes are often not primarily driven by failure of the anticoagulant itself, but rather by underlying stroke mechanisms or residual thromboembolic risk that is not mitigated by simply changing agents,” study investigator Lucio D’Anna, MD, PhD, consultant neurologist at Imperial College London in London, England, told Medscape Medical News.

Instead of routinely switching therapy, D’Anna suggested clinicians focus on identifying modifiable factors such as adherence, dosing, drug interactions, or competing stroke mechanisms.

Although no clear harm signal was observed, the investigators cautioned that switching could theoretically introduce transient risks, including suboptimal anticoagulation or dosing errors, particularly in the early poststroke period.

However, D’Anna emphasized that “switching may still be appropriate in selected cases, including intolerance, contraindications, or specific patient-related factors.”

He also acknowledged the contribution of Matteo Foschi, MD, a consultant neurologist, who D’Anna said has played an instrumental role in the research.

Ongoing Clinical Uncertainty

The findings are unlikely to immediately change current clinical practice, Larry B. Goldstein, MD, Ruth L. Works Professor and chair of the Department of Neurology at the University of Kentucky in Lexington, Kentucky, shared with Medscape Medical News.

“This is a cohort observational study, not a randomized controlled trial,” he said, noting that, although it was a well-designed study, residual confounding remains a concern despite robust statistical adjustment. The near-even split between patients who continued vs switched therapy reflects the ongoing clinical uncertainty and the tendency to act in response to prevent a recurrent event.

Goldstein agreed with the investigators that rather than routinely switching anticoagulants, clinicians should focus on optimizing modifiable factors.

“It is important that the patient not miss doses of a DOAC, which have a relatively short duration of action,” said Goldstein, co-director of the Kentucky Neuroscience Institute in Lexington. “Evaluation for causes other than cariogenic embolism that may coexist with AF is also important.”

A large prospective follow-up study, ASPERA-P, is ongoing and is expected to yield more robust results in the future.

The study was supported by the ASPERA project. Disclosure information for study authors is available in the original study publication. Goldstein reported having no relevant financial disclosures.