The Food and Drug Administration approved Bayer's Sunharbo, or asundexian, on Nov. 9 for adults who have had a recent mild-to-moderate non-cardioembolic ischemic stroke or a high-risk transient ischemic attack. The drug is the first Factor XIa inhibitor approved in the United States. It is given as a 50 milligram tablet once a day, alongside antiplatelet therapy, to reduce the risk of another stroke.
Stroke is a leading cause of death and disability, and recurrent strokes are common even with antiplatelet therapy. The standard secondary-prevention regimen after a non-cardioembolic stroke is aspirin, clopidogrel, or a short course of dual antiplatelet therapy. Asundexian is an add-on to that regimen, not a replacement. The trial's population was selected for recent stroke or high-risk TIA, where the risk of recurrence is highest in the first weeks and months.
The approval is narrow by design. Sunharbo is not approved for stroke prevention in atrial fibrillation, the much larger market for blood thinners. That restriction reflects the drug's history. Asundexian failed in atrial fibrillation in 2024, when a large trial stopped early because it was inferior to apixaban. The new approval is a comeback in a different population, and it rests on a trial that showed a meaningful reduction in recurrent stroke without a statistically significant increase in major bleeding.
The trial
The approval is based on OCEANIC-STROKE, a Phase 3, randomized, double-blind, placebo-controlled trial published in the New England Journal of Medicine in April 2026. The trial enrolled 12,327 adults within 72 hours of a non-cardioembolic ischemic stroke with a National Institutes of Health Stroke Scale score of 15 or less, or a high-risk transient ischemic attack. Patients were treated at 702 sites in 37 countries. All received antiplatelet therapy, and half also received asundexian 50 milligrams once daily.
Recurrent ischemic stroke occurred in 6.2% of patients on asundexian and 8.4% on placebo, a 26% relative reduction with a hazard ratio of 0.74 and a 95% confidence interval of 0.65 to 0.84. The result was statistically significant. Disabling or fatal stroke fell 31%, from 3.0% to 2.1%. The absolute reduction was about 2.2 percentage points over a median follow-up of roughly 567 days, which works out to about 50 patients treated for a year to prevent one recurrent stroke.
The safety result was the one the field was watching. ISTH major bleeding occurred in 1.9% of patients on asundexian and 1.7% on placebo, a hazard ratio of 1.10 with a confidence interval of 0.85 to 1.44. The difference was not statistically significant. Minor bleeding, hemorrhagic stroke, symptomatic intracranial hemorrhage, and fatal bleeding were all similar to placebo. Gastrointestinal bleeding was more common with asundexian, especially lower gastrointestinal bleeding, and the FDA's label notes that major bleeding increased in patients older than 75 and during the first 90 days of treatment.
There is a caveat. A 90-day secondary outcome missed statistical significance, with a hazard ratio of 0.84 and a confidence interval of 0.69 to 1.02. That halted the trial's hierarchical testing plan. The investigators attributed the result to statistical power rather than a lack of effect, and the primary result at the longer follow-up was robust. But the early signal is not as strong as the overall result.
The benefit was consistent across subgroups. The trial reported similar effects by age, sex, region, stroke subtype, baseline NIHSS score, and whether patients received single or dual antiplatelet therapy. A separate analysis presented at the American Heart Association meeting showed consistent benefit across vascular risk strata, suggesting that clinicians may not need a risk score to decide who should receive the drug.
Why the class matters
Factor XIa is part of the intrinsic coagulation pathway. The hypothesis behind the class is that blocking it can reduce thrombosis with less bleeding than blocking Factor Xa, the target of apixaban and rivaroxaban. The idea has attracted billions of dollars in investment because bleeding is the main limitation of existing anticoagulants. A drug that prevents clots without causing as many bleeds would be a genuine advance.
Asundexian is an oral, once-daily inhibitor with a half-life of roughly 14 to 17 hours. The 50 milligram dose was chosen for the stroke trial after earlier studies. The FDA granted the application Fast Track and Priority Review designations, and the agency's announcement quoted Hylton V. Joffe, director of the office that reviewed the application, calling it "a meaningful advance for patients and health care providers." The approval follows China's decision in October to clear asundexian, making the United States the second major market.
The class's promise is not yet fully proven. The stroke trial showed that asundexian can reduce ischemic events without a significant increase in major bleeding, but the confidence interval around the bleeding result still allows for a modest increase. The gastrointestinal bleeding signal is real. The drug is not a free lunch, and the label's warnings reflect that.
The gastrointestinal signal is the one clinicians will watch. Lower gastrointestinal bleeding was more common with asundexian, and the label notes the increase in patients older than 75 and in the first 90 days. For patients with a history of GI bleeding, the risk-benefit calculation may favor staying on antiplatelet therapy alone. For patients with recurrent strokes despite antiplatelet therapy, the calculation may point the other way.
The atrial fibrillation failure
The approval cannot be understood without OCEANIC-AF, the trial that nearly ended the program. That study enrolled 14,810 patients with atrial fibrillation and a high risk of stroke. It compared asundexian 50 milligrams once daily with apixaban 5 milligrams twice daily. The trial was stopped early because asundexian was inferior. Stroke or systemic embolism occurred in 1.3% of patients on asundexian and 0.4% on apixaban, a hazard ratio of 3.79. Major bleeding was lower with asundexian, 0.2% versus 0.7%, but the efficacy gap was too large to accept.
The failure raised questions about the class. One possibility is that Factor XIa inhibition is not enough to prevent clots in the left atrium, where blood flow is slow and thrombin generation may need more suppression. Another is that the 50 milligram dose was too low for atrial fibrillation. Another is that switching patients from a Factor Xa inhibitor to a Factor XIa inhibitor creates a gap in protection. The ongoing milvexian program, from Bristol Myers Squibb and Johnson & Johnson, is testing a different Factor XIa inhibitor in atrial fibrillation and stroke. Its results will shape how the field reads asundexian's failure.
For now, the label is clear: Sunharbo is not approved for atrial fibrillation. The approval is for secondary prevention after a non-cardioembolic stroke or high-risk TIA, a population where antiplatelet therapy is standard and recurrent events remain common. That is a large group, but it is not the anticoagulant market.
What the label says
The label recommends 50 milligrams once daily, with antiplatelet therapy. It warns about bleeding, including the higher risk in patients older than 75 and in the first 90 days. It notes that gastrointestinal bleeding is more common. It advises against use with combined P-glycoprotein and strong CYP3A4 inducers such as carbamazepine, and in patients with moderate or severe hepatic impairment. Treatment should be paused at least 24 hours before elective surgery or invasive procedures and at least 24 hours after thrombolysis for acute ischemic stroke.
Those details matter for clinical use. The drug is not a simple replacement for aspirin or clopidogrel. It is an add-on for a specific population, and the bleeding risk has to be weighed against the stroke risk. The NNT of about 50 is favorable for a secondary-prevention drug, but the number needed to harm depends on the patient's bleeding risk, age, and gastrointestinal history.
The reaction
Bayer has positioned the approval as a comeback. The company spent years and billions of dollars on the Factor XIa program, and the atrial fibrillation failure was a major setback. The stroke approval gives the program a commercial path and a clinical rationale. The drug will compete with generic antiplatelet therapy and with other approaches to secondary prevention, and its price and reimbursement will determine how quickly it is adopted.
The commercial path is narrower than the class's original ambition. Bayer had hoped to challenge apixaban and rivaroxaban in atrial fibrillation, the largest anticoagulant market. The stroke label gives the drug a smaller but well-defined population, and it may be used in patients who cannot tolerate existing anticoagulants or who have recurrent strokes despite antiplatelet therapy.
Clinicians will want more data. The trial's 90-day secondary outcome was not significant, and the bleeding signal in older patients and the first 90 days deserves attention. Post-marketing studies will show how the drug performs in routine practice, where patients are older and have more comorbidities than in a controlled trial. The label's warnings are a starting point, not the final word.
The competitive landscape is also moving. Milvexian is in late-stage trials for stroke and atrial fibrillation. Other Factor XIa inhibitors are in development. If milvexian succeeds in atrial fibrillation, the class's story changes. If it fails, asundexian's narrow approval may be the class's ceiling.
What to watch
The first months after approval will show how clinicians use the drug. Watch the uptake in stroke centers, the patient selection, and the bleeding events in real-world use. Watch the pricing and the payer coverage. Watch the European and Japanese reviews, which are still pending. Watch the post-marketing studies and any label changes.
The broader question is whether Factor XIa inhibition can deliver on its promise. The stroke trial is a proof of concept: a 26% reduction in recurrent stroke without a significant increase in major bleeding. The atrial fibrillation failure is a reminder that the mechanism is not universal. The class may end up as a niche therapy for patients who cannot tolerate existing anticoagulants, or it may become a new standard for secondary prevention. The answer will come from more trials and more real-world data.
The first Factor XIa inhibitor is approved. The label is narrow, the bleeding question is not fully settled, and the class still has something to prove. That is not a disappointment. It is how a new mechanism earns its place.
Primary sources
- The FDA's Nov. 9, 2026 approval announcement for Sunharbo, for the indication, dosing, trial summary, safety warnings, and approval pathway.
- The New England Journal of Medicine for the OCEANIC-STROKE trial results and the OCEANIC-AF comparison.
- STAT for the approval and the clinical context.
- TCTMD for the OCEANIC-STROKE subgroup analysis and the trial details.
- Bayer's press materials for the program history and the regulatory timeline.
- Clinical Trials Arena for the trial design and the atrial fibrillation failure.