Ashkan Shoamanesh, MD, is an associate professor of medicine and the Marta and Owen Boris Chair in Stroke Research and Care at McMaster University.

Ashkan Shoamanesh, MD

European Society of Cardiology (ESC) Congress 2026

, found no net benefit from edoxaban among unselected patients with atrial fibrillation (AF) and a prior intracranial hemorrhage (ICH), with excess bleeding offsetting reductions in ischemic events.¹

Ashkan Shoamanesh, MD, of McMaster University, principal investigator of the trial, discussed with 

 how clinicians might individualize care in this population going forward.

Not All Intracranial Hemorrhage Is the Same, With Ashkan Shoamanesh, MD

Not All Intracranial Hemorrhage is the Same

Asked whether prior ICH is too broad a category for clinical decision-making, and whether hemorrhage location or underlying cerebral small vessel disease should factor in instead, Shoamanesh agreed.

"Not all ICH are the same. The natural history and underlying disease that drives subdural hemorrhages is very different than the natural history and underlying disease that drives lobar ICH or convexity bleeds," he said, noting non-lobar ICH is predominantly driven by hypertensive arteriopathy and arteriosclerosis, distinct from the cerebral amyloid angiopathy typically underlying lobar and convexity hemorrhage.

Shoamanesh said a trial powered for a single ICH subtype in isolation would be nearly impossible to conduct. He noted the global initiative behind ENRICH-AF, activating 240 sites and enrolling for years to accumulate its combined study population, illustrates just how difficult recruitment is in this population, with an average enrollment rate of roughly 2 patients per center per year across the broader, lumped strategy the trial ultimately used.²

Balancing Thromboembolic and Bleeding Risk

The trial enrolled patients with high-risk AF, defined by a CHA2DS2-VASc score of ≥ 2, though prior ICH itself did not contribute points toward the score; only an actual history of ischemic stroke or transient ischemic attack did.² The average CHA2DS2-VASc score was 4, and the annualized ischemic stroke rate was approximately 5.5%, reflecting an already high-risk population for ischemic stroke.² Even at higher CHA2DS2-VASc levels, Shoamanesh said stratified analyses showed no clear anticoagulation benefit.

Shoamanesh said stroke risk is unlikely to be the key differentiator for treatment decisions. "I really think that the key discrimination here is on the safety side, not on the ischemic stroke side, and... it's the lower-risk bleeding groups that may have benefit... or a lower dose of an anticoagulant that can have greater safety, but still have some efficacy," he said, adding that factor XI inhibitors, currently being studied for AF more broadly, represent a compelling potential option given their theoretically safer bleeding profile.

On left atrial appendage occlusion (LAAO) as an alternative strategy, Shoamanesh said he now refers patients with lobar or convexity ICH for consideration, prescribing edoxaban 15 mg for those who are not candidates or decline the procedure but have sufficient ischemic stroke risk. He cautioned against assuming LAAO is the default answer, however. "Just assuming that LAAO is the answer is the wrong approach. I think we need an RCT showing that LAAO is the preferred approach in this population," Shoamanesh said, pointing to the CLOSURE-AF trial, testing LAAO in the frailest population studied with the device to date, which was recently stopped for futility. "In the interim, we do our best clinically through guesswork."

On what the ongoing ASPIRE trial, a double-blind comparison of apixaban with aspirin, might add beyond ENRICH-AF, Shoamanesh said he expects dosing, rather than drug choice, to be the key differentiator. He cited prior excess bleeding signals with apixaban in the smaller APACHE-AF trial and the larger PRESTIGE-AF trial as evidence the different DOACs are unlikely to behave differently from edoxaban in this population.³ He described ENRICH-AF's broader lesson as a caution against assuming findings from general AF populations generalize to frail, high-risk subgroups such as ICH survivors, a caution he said extends to LAAO as well as anticoagulation.

Editors’ note: Shoamanesh reports relevant disclosures with AstraZeneca, Bayer AG, Daiichi Sankyo, and others.

European Society of Cardiology. Safer stroke prevention strategies are needed for patients with atrial fibrillation after intracranial haemorrhage. Published August 29, 2026. Accessed August 30, 2026.

Shoamanesh A, et al. ENRICH-AF: edoxaban in intracranial hemorrhage survivors with atrial fibrillation. Presented at: ESC Congress 2026; August 29, 2026; Munich, Germany.

Veltkamp R, Korompoki E, Harvey KH, et al. Direct oral anticoagulants versus no anticoagulation for the prevention of stroke in survivors of intracerebral haemorrhage with atrial fibrillation (PRESTIGE-AF): a multicentre, open-label, randomised, phase 3 trial. Lancet. 2025;405(10482):927-936. doi:10.1016/S0140-6736(25)00333-2

Videos

Ashkan Shoamanesh, MD, discusses individualizing stroke prevention after ENRICH-AF, from LAAO to factor XI inhibitors to the ASPIRE trial.

Beyond Edoxaban: Rethinking Stroke Prevention, With Ashkan Shoamanesh, MD

, is the largest randomized trial yet conducted in patients with atrial fibrillation (AF) and a prior intracranial hemorrhage (ICH), a population historically excluded from landmark AF anticoagulation trials.¹ Ashkan Shoamanesh, MD, of McMaster University, principal investigator of the trial, discussed the results, and a dose-specific signal within them, with 

Why Edoxaban Failed in ENRICH-AF, With Ashkan Shoamanesh, MD

Reconciling ENRICH-AF's Mixed Signal on Stroke and Bleeding

Over an average follow-up of 28 months, edoxaban did not significantly reduce the primary composite endpoint of stroke or systemic embolism compared with no anticoagulation (11.8% vs 12.8%; HR, 0.88; 95% CI, 0.61-1.26; 

 = .48), while the primary safety outcome, ISTH major bleeding, occurred in 11.6% of patients receiving edoxaban compared with 5.2% receiving no anticoagulation (HR, 2.23; 95% CI, 1.39-3.59; 

Asked how clinicians should reconcile the reductions in ischemic stroke and myocardial infarction (MI) with the nearly 3-fold excess in hemorrhagic stroke, Shoamanesh described a net washout effect: "The myocardial infarction signal that we saw was actually novel and surprising, that there's a 80% reduction in MI as well. However, MI was fairly infrequent,” he said. "What that 40% reduction in ischemic stroke versus that threefold excess in hemorrhagic stroke ends up doing, in an absolute sense, is creating a washout... In all-comers, there was no net benefit that we identified. There was also a twofold excess risk [of ISTH major bleeding], and then concerningly, for fatal bleeding, there was a 4.5 fold excess risk."

Lobar Hemorrhage and a Critical Dose-Response Finding

Regarding which hemorrhage locations drove the excess bleeding risk, Shoamanesh pointed to a specific subgroup identified during the trial: patients with lobar intracerebral hemorrhage or convexity subarachnoid bleeding.

"These are locations of bleeding that typically occurred due to underlying cerebral amyloid angiopathy in this age group," he explained. "Our DSMB, after we had enrolled about two thirds of the population, recommended that this subgroup be excluded because a safety review identified that their rates of recurrent ICH were unacceptably high... 11% of those with edoxaban had a recurrent intracranial hemorrhage within this subgroup versus 1% in those who received non-anticoagulant medical therapy, and the majority of the recurrent ICH on edoxaban were fatal."

By the end of the trial, the annualized rate of recurrent ICH in patients with lobar or convexity hemorrhage was 7.5% with edoxaban versus 3.5% with non-anticoagulant therapy, compared with a much smaller absolute difference in patients with non-lobar ICH (2.5% vs 1.2%).² A separate and, per Shoamanesh, clinically important finding emerged when results were examined by edoxaban dose.

"All of the excess major bleeding was seen in patients who indicated the 60 milligram dose of edoxaban, whereas when we looked at the 30 milligram dose of edoxaban, there was no excess major bleeding. That was a statistically significant interaction for the primary safety endpoint," Shoamanesh said. "And there was a suggestion for the primary efficacy endpoint, where those that were indicated the 30 milligram dose, there was about a 30% risk reduction for the composite of stroke or systemic embolism. What this suggests is that the standard dosing of anticoagulation, particularly with edoxaban, may be too high, and that a lower dose... may actually have been a more favorable dose in this population."

Shoamanesh noted a lower dose, such as the 30 mg dose, or potentially even the 15 mg dose evaluated in the ELDERCARE-AF trial of very elderly Japanese patients ineligible for standard anticoagulation, could prove more favorable in this population.³ The findings raise a question future prospective trials, rather than ENRICH-AF itself, will need to answer directly: whether standard-dose anticoagulation in AF is simply too aggressive for patients with a prior ICH.

Okumura K, Akao M, Yoshida T, et al; ELDERCARE-AF Committees and Investigators. Low-dose edoxaban in very elderly patients with atrial fibrillation. N Engl J Med. 2020;383:1735-1745. doi:10.1056/NEJMoa2012883

Ashkan Shoamanesh, MD, explains ENRICH-AF's mixed stroke and bleeding results, and a dosing signal that may point to a safer approach.

ENRICH-AF: A Dosing Clue Emerges After Edoxaban's Miss, With Ashkan Shoamanesh, MD

Secondary Analyses of OCEANIC-STROKE Confirm Efficacy of Asundexian

Asundexian reduces stroke incidence and requires fewer interventions in patients after a non-cardioembolic 

 or high-risk transient ischemic attack (TIA), according to 4 late-breaking analyses presented at the European Stroke Organisation Conference (ESOC) 2026.

The original OCEANIC-STROKE study confirmed the efficacy of asundexian, an investigational oral Factor XIa (FXIa) inhibitor, for the prevention of ischemic stroke in patients following non-cardioembolic ischemic stroke or high-risk TIA. Despite having an overall minor role in the formation of hemostatic plugs to seal leaks at vessel injury sites, FXIa is believed to also contribute to the formation of pathological thrombus growth and vessel blockage as well.

“We’ve never had a trial previously where we’ve had such a large number of major and minor bleeds, and to see no excess risk with an antithrombotic versus placebo,” Ashkan Shoamanesh, MD, associate professor of medicine and Marta and Owen Boris Chair in Stroke Research and Care at McMaster University and co-principal investigator of the OCEANIC-STROKE trial, told 

 in an exclusive interview. “This would be the first time in history that we’ve seen this, and it really does make me hopeful about not only bringing this drug to market to implement the results we’ve seen thus far, but also what it could mean for cardiovascular disease on a broader scale.”

OCEANIC-STROKE was a phase 3, double-blind, event-driven, randomized, placebo-controlled trial conducted at 702 centers in 37 countries. Patients were eligible for inclusion if they had a non-cardioembolic ischemic stroke or high-risk TIA within 72 hours after the onset of symptoms. Those with ischemic stroke had a maximum score of 15 on the National Institutes of Health Stroke Scale (NIHSS), and high-risk TIA was defined as an ABCD2 score of 6 or 7. Patients who had undergone intravenous thrombolysis or mechanical thrombectomy were eligible if ≥24 hours had passed since treatment.

Eligible patients were randomly assigned in a 1:1 ratio to either asundexian 50 mg daily or matching placebo on top of planned dual or single antiplatelet therapy. The primary outcome was first occurrence of ischemic stroke, while secondary endpoints included any stroke, a composite of death from myocardial infarction, cardiovascular causes, or stroke, death from any cause, and TIA, among others.

A total of 12,327 patients were randomized, with 6162 receiving asundexian and 6165 receiving placebo. By the median follow-up of 567 days, a total of 1624 patients had discontinued asundexian and 1589 had discontinued placebo. Of the remainder, ischemic stroke occurred in 384 patients in the asundexian arm compared to 518 in the placebo arm (HR, 0.74; 95% CI, 0.65 to 0.84; 

In the present analyses of OCEANIC-STROKE, investigators examined several factors including the proportion of patients with a NIHSS score ≥8, the number of disabling strokes in each arm, the level of stroke severity in each arm, and the need for acute intervention in each arm. The results presented at ESOC 2026 largely supported the existing data of asundexian, reinforcing its efficacy for this population.

Key findings from these analyses included the following data:

Acute intervention: Of the ischemic strokes that did occur, 40 patients required acute intervention via intravenous thrombolysis and/or endovascular treatment in the asundexian group compared to 82 in the placebo group

Disabling/fatal stroke risk: Disabling stroke occurred in 2.1% of the asundexian group and 3%% in the placebo group (HR, 0.69; 96^ CI, 0.55-0.87)

Stroke severity: Moderate-to-high severity strokes (NIHSS ≥8) occurred in 88 patients in the asundexian group versus 157 in the placebo group

Ischemic stroke with fatal outcomes: 16 patients experienced fatal outcomes in the asundexian group versus 28 in the placebo group (HR, 0.57; 95% CI, 0.31-1.06)

With these data established and presented, Shoamanesh believes that the next step in the investigational and approval process for asundexian is confirmation of safety data.

“There’s no reason to believe that there’ll be any loss of safety over time,” Shoamanesh said. “If anything, we typically see with antithrombotics that the bleeding is front-loaded, and the further you go away from the time of initiation, the less the bleeding – that’s what the curves would suggest thus far, and that’s what we’ve seen in OCEANIC-STROKE.”

Editors’ Note: Shoamanesh reports disclosures with the Heart and Stroke Foundation of Canada, the Marta and Owen Boris Foundation, Bayer AG, Daiichi Sankyo, AstraZeneca, Octapharma, Inc., and others.

Bayer. Bayer announces new late-breaking data from OCEANIC-STROKE trial of asundexian. May 6, 2026. Accessed May 22, 2026. 

https://www.bayer.com/en/us/news-stories/data-from-oceanic-stroke-trial-of-asundexian

Sharma M, Dong Q, Hirano T, et al. Asundexian for secondary stroke prevention. 

Ashkan Shoamanesh, MD, discusses a series of abstracts presented at ESOC 2026 reinforcing asundexian’s protective effect in patients at risk of ischemic stroke or TIA.

Asundexian Reduces Stroke Incidence Without Increasing Bleeding, With Ashkan Shoamanesh, MD

Asundexian, an investigational Factor XIa inhibitor, lowers the risk of stroke in patients who recently experienced non-cardioembolic 

 or a transient ischemic attack (TIA) without increasing bleeding, according to data from OCEANIC-STROKE.

Despite secondary prevention strategies, 5.1% of patients with ischemic stroke or TIA have a recurrent stroke within 1 year – after 5 years, 22% of patients who had survived the event are disabled or dead. Dual antiplatelet therapy incompletely reduces this risk, and long-term use is associated with higher risks of major hemorrhage. FXIa inhibition approaches the condition from a new angle, potentially circumventing these issues.

“Not only are we not seeing excess in major bleeding, but we also didn’t see excess in minor bleeding,” Ashkan Shoamanesh, MD, associate professor of medicine and Marta and Owen Boris Chair in Stroke Research and Care at McMaster University, told 

 in an exclusive interview. “Although minor bleeding is a less immediate clinical consequence regarding health, it certainly still is very significant, because it results in substantial psychological distress for patients and leads to non-compliance, which leads to less efficacy. It’s frankly historic to have a trial of this size and to have confidence in the results, and to see no excess risk with an antithrombotic versus placebo.”

OCEANIC-STROKE was a phase 3, double-blind, event-driven, randomized, placebo-controlled trial conducted at 702 centers across 37 countries. Patients were eligible if they were ≥18 years old with a non-cardioembolic ischemic stroke or high-risk TIA within 72 hours after the onset of symptoms and had a plan for treatment with dual or single antiplatelet therapy. Those with a history of atrial fibrillation or other conditions with an indication for anticoagulation, active nontrivial bleeding, or a qualifying stroke related to a procedure or other specific cause were excluded.

After enrollment, patients were randomly assigned in a 1:1 ratio to either asundexian 50 mg daily or matching placebo, on top of planned dual or single antiplatelet therapy. The trial’s primary outcome was the first occurrence of ischemic stroke, which was determined to have occurred if there was a new focal neurologic deficit or clinically meaningful worsening of an existing deficit persisting for ≥24 hours. The primary safety outcome was major bleeding, including fatal bleeding, symptomatic bleeding in a critical area or organ, and clinically overt bleeding resulting in a hemoglobin reduction of ≥2 g/dL.

A total of 12,327 patients were randomized, with 6162 assigned to the asundexian group and 6165 to placebo. A total of 45 patients withdrew consent, and 39 were lost to follow-up. The mean age of the final group of patients was 68 years. 11,677 patients presented with ischemic stroke, while the remainder had TIA. Median follow-up was 567 days (interquartile range [IQR], 377-729).

Ultimately, Shoamanesh and colleagues found that ischemic stroke incidence was substantially lower in the asundexian group than the placebo group (6.2% vs 8.4%; cause-specific HR, 0.74; 95% CI, 0.65-0.84; 

 <.001). The incidence of a composite of death from cardiovascular causes, myocardial infarction, or stroke was also lower in the asundexian arm. Major bleeding incidence was similar between both arms (1.9% and 1.7%, respectively; cause-specific HR, 1.1; 95% CI, 0.85-1.44). Adverse events occurred in 69.3% of the asundexian group and 70.1% in the placebo group – serious adverse events had an incidence of 19.2% and 19.5%, respectively.

The team concluded that treatment with asundexian substantially reduced the risk of ischemic stroke and major cardiovascular events compared with placebo, without increasing the risk of major bleeding in patients. Shoamanesh and colleagues believe that these data mark the beginning of a major shift in ischemic stroke treatment, and in cardiovascular disease overall, as a result of FXIa inhibition’s success.


“I think this is just the beginning of a large paradigm shift in our treatment of patients with cardiovascular disease,” Shoamanesh said. “Certainly, to date, there have been challenges with factor XI and other disease states. But I suspect the reason for that is that we just haven’t found the right recipe yet – either with higher doses of existing agents or with combination treatments, factor XI is going to be used more widely across the spectrum of cardiovascular disease and venous thromboembolic disease.”

McMaster University. New anti-clotting medication lowers risk of stroke without added bleeding. 

. April 15, 2026. Accessed April 20, 2026. 

https://www.eurekalert.org/news-releases/1123838

Shoamanesh discusses the positive data from OCEANIC-STROKE, a phase 3 trial comparing the investigative FXIa inhibitor to placebo in patients with prior ischemic stroke.

Ashkan Shoamanesh, MD

Articles