FDA Approves Sunharbo, a Factor XIa Inhibitor: Ischemic Stroke Down 26%, Major Bleeding Up 0.2 Points
The FDA approved Bayer asundexian on November 9. In 12,327 patients it lowered ischemic stroke from 8.4% to 6.2%, major bleeding was 1.9% against 1.7%, and lower gastrointestinal bleeding rose from 1.0% to 1.7%.

On November 9 the FDA approved Sunharbo (asundexian), Bayer's oral factor XIa inhibitor, to reduce the risk of stroke in adults after a recent mild-to-moderate non-cardioembolic ischemic stroke or a high-risk transient ischemic attack. It is number 48 on the FDA's list of novel drug approvals for 2026.
The question that decides whether a drug like this matters is simple: how many strokes does it prevent, and what does it cost in bleeding? The trial that supports the approval, OCEANIC-STROKE, answers both with numbers we can check.
What did the trial test?
OCEANIC-STROKE randomized 12,327 patients within 72 hours of a non-cardioembolic ischemic stroke or high-risk TIA. Half (6,162) received asundexian 50 mg once daily and half (6,165) received placebo, both on top of planned single or dual antiplatelet therapy. The primary efficacy outcome was ischemic stroke; the primary safety outcome was major bleeding.
This is a different group from the one in which the drug first failed. Bayer's earlier trial in atrial fibrillation, OCEANIC-AF, was stopped early. The approved indication is stroke or TIA of non-cardioembolic origin, not atrial fibrillation.
How large is the benefit, and how large is the bleeding cost?
Asundexian lowered ischemic stroke by 2.2 percentage points and left major bleeding within 0.2 points of placebo, but lower gastrointestinal bleeding rose from 1.0% to 1.7%.
Percentage of patients with each outcome in OCEANIC-STROKE. Efficacy: all 12,327 randomized. Bleeding: 12,254 treated (6,124 asundexian, 6,130 placebo). CRNM is clinically relevant non-major bleeding.
Source: Asundexian for Secondary Stroke Prevention, N Engl J Med 2026; OCEANIC-STROKE bleeding analysis, Circulation 2026. The stroke outcome and its hazard ratio (0.74, 95% CI 0.65 to 0.84) are from the NEJM abstract. Bleeding categories use ISTH definitions and a while-on-treatment analysis, from the Circulation paper. Lower GI bleeding: 102 patients on asundexian, 59 on placebo.
View data
| outcome | asundexian | placebo |
|---|---|---|
| Ischemic stroke | 6.2 | 8.4 |
| Major bleeding | 1.9 | 1.7 |
| Major or CRNM bleeding | 5.5 | 5 |
| Any bleeding | 12.5 | 12.3 |
| Any GI bleeding | 3 | 2.4 |
| Lower GI bleeding | 1.7 | 1 |
The efficacy result is clear. Ischemic stroke occurred in 6.2% of patients on asundexian and 8.4% on placebo, a cause-specific hazard ratio of 0.74 (95% CI 0.65 to 0.84, P below 0.001). That is a 26% relative reduction and 2.2 percentage points in absolute terms. Taking those rounded percentages at face value, about 45 patients would need treatment over the trial period to prevent one ischemic stroke. That is our arithmetic, not a figure the authors report.
On safety, major bleeding occurred in 117 patients (1.9%) on asundexian and 107 (1.7%) on placebo (hazard ratio 1.10, 95% CI 0.85 to 1.44, P=0.46). The interval includes no difference, and it also includes a 44% higher risk, so the trial does not rule out a modest increase. Symptomatic intracranial hemorrhage was 41 patients (0.7%) against 36 (0.6%).
Where did the extra bleeding show up?
Not in the headline categories. Any bleeding was almost identical (12.5% against 12.3%). The signal is in the gut: any gastrointestinal bleeding was 186 patients (3.0%) against 147 (2.4%), hazard ratio 1.28 (95% CI 1.03 to 1.59). Lower gastrointestinal bleeding was 102 (1.7%) against 59 (1.0%), hazard ratio 1.75 (95% CI 1.27 to 2.41), mostly rectal bleeding.
In per-1,000 terms, that is about 7 more lower-GI bleeds per 1,000 patients and about 2 more major bleeds per 1,000, against about 22 fewer ischemic strokes per 1,000. The trade is favorable on these numbers, which is presumably why the FDA approved it. It is not free, and the lower-GI result is a detail clinicians will want to see in the label.
Why does a factor XIa approval matter?
Factor XI sits in the intrinsic clotting pathway. The idea behind inhibiting it is that clots can be prevented with less effect on the clotting that stops ordinary bleeding. Reviews of the field describe a mixed record so far, with neutral phase 2 studies and the early termination of OCEANIC-AF. OCEANIC-STROKE is a phase 3 result in which the bleeding profile held up beside a clear efficacy gain, and an approved drug now exists to test that idea in routine care.
For people tracking the bioeconomy, the approval also lands in an already busy year. The FDA list stood at 46 novel approvals on October 9, as we covered in FDA Has Approved 46 Novel Drugs by October 9. Two more have followed on the agency's list: bepirovirsen on October 20 and Sunharbo on November 9.
What this does and does not show
The numbers show that, in this trial population, adding asundexian to antiplatelet therapy lowered ischemic stroke without a statistically significant rise in major bleeding. They do not show how the drug performs in older or sicker patients than those enrolled, in patients on other anticoagulants, or over years of use. The trial was funded by Bayer. We read the NEJM and Circulation abstracts rather than the full papers, so subgroup results, the length of follow-up and the exact definition of the time window are not covered here. As of November 16, ClinicalTrials.gov listed the study as completed (primary completion October 23, 2025) with no results posted in the registry. The 2.2-point difference and the 45 figure use rounded percentages and ignore how long patients were followed.
Sources
- FDA, Novel Drug Approvals for 2026: Sunharbo (asundexian), approved 11/9/2026, number 48; indication wording; Hibsago (bepirovirsen), 10/20/2026.
- Asundexian for Secondary Stroke Prevention, N Engl J Med 2026, doi:10.1056/NEJMoa2513880 (abstract read via Europe PMC): 12,327 randomized; 6.2% vs 8.4% ischemic stroke; hazard ratio 0.74; major bleeding 1.9% vs 1.7%.
- Bleeding With Asundexian Versus Placebo in Patients With Noncardioembolic Ischemic Stroke or High-Risk TIA, Circulation 2026, doi:10.1161/CIRCULATIONAHA.126.081836: all bleeding counts and hazard ratios.
- ClinicalTrials.gov, NCT05686070: enrollment 12,327, start January 26, 2023, primary completion October 23, 2025, results not posted as of November 16, 2026.
- FXI and FXII in experimental stroke and FXI in clinical stroke trials, Neurol Res Pract 2026: background on the early termination of OCEANIC-AF.
- Our earlier analysis: FDA Has Approved 46 Novel Drugs by October 9.
Recommended Reading

KRAS G12D Went From 2 Trial Starts to 24 in One Year. Now 17 of This Year's 20 Are Phase 2 or 3

Optogenetics Won the Nobel With 1,745 Papers a Year, but Only Nine Trials, All in the Eye

Nearly 70% or 20% to 43%? How to Read Regeneron's Muscle Data for GLP-1 Users
Discussion (0)
Sign in to join the conversation and share your thoughts
No comments yet. Be the first to start the discussion!