ESOC 2026 | Incident stroke and time course of benefit of asundexian: analyses of the OCEANIC-STROKE trial

There are two major analyses that we’ve done for the OCEANIC trial. The first one was looking at incident strokes, strokes that occurred during the course of the trial by treatment allocation. And the second one was to look at the duration of benefit. So there have been a number of questions as to whether or not this treatment should continue indefinitely once it’s approved, or if there’s a time limit.

Just starting with the incident stroke, which was presented this morning. In OCEANIC stroke, recall that we randomized people within 72 hours of symptom onset of either a non-cardioembolic ischemic stroke with an NIHSS of up to 15 or high-risk TIA with an ABCD2 score of 6 or 7. All of these patients were randomized within 72 hours of symptom onset and everyone was on antiplatelet therapy. During the course of the trial, we had 902 incident ischemic infarcts. Asundexian reduced the occurrence of ischemic stroke with a hazard ratio of 0.74. Our interest in this analysis was to characterize those strokes that occurred by severity, by treatment, and by outcome. Now the story here is actually quite nice. What we found was strokes that occurred with asundexian were milder in terms of the NIH stroke scale than strokes that occurred with placebo. If you look at the category of strokes that have an NIH SS of greater than eight, asundexian was lower in terms of this category by about 7%. So put it another way, the proportion of strokes which fell into this category, 7% lower with asundexian treatment than with placebo. And this was on top of long-term antiplatelet therapy. There was a shift towards milder stroke severities with asundexian, so proportionally more in the category of NIHSS of three or less, and also more in the moderate category between four and seven.

The other thing we looked at was how acute strokes were treated. Now, we gave guidance to investigators such that if there was a stroke, which was a large vessel occlusion, they were to proceed directly to mechanical thrombectomy without a need for unblinding. If there was a consideration for intravenous thrombolysis and the patient was still taking study medication, then there was the option of unblinding if the investigator and the treating physician felt it would guide treatment decisions. What we found was overall acute treatment was used in about 10% of patients on asundexian versus almost 16% on placebo. So fewer patients on asundexian had acute treatment by either modality, IVT, EVT, or a combination of the two. Now some of that was due to the fact that there was less IVT as people were unblinded, found to be on asundexian, where we do not have a lot of data in terms of potential interaction. But also some of this was due to the fact that the strokes were milder. We looked at discharge NIHSS, and that is less lower in patients treated with asundexian than with placebo. So, you know, the mean was about three with placebo and two with asundexian at seven days of hospital discharge, whichever came first. We did not see a trend to a worsening of this measure with asundexian treatment, which really suggests to us that treatment isn’t being complicated by hemorrhage, resulting in worse NIHSS early on. Now we do need more data because the total number of participants treated with asundexian was 40 and only 13 was IV thrombolysis alone. So we will be looking further into this.

The second analysis was to look at the duration of benefit. And we’ve been asked questions about early benefit as well as how long it lasts. So one of the things that you notice is when you blow up the early part of the cumulative incidence curve, the Aalen–Johansen curve that was published in the New England Journal on April 16th, what you see is that those curves start to diverge at about day 10. There is some random variation and then the divergence becomes well established and continues for as long as we have good data in the trial. We’ve looked at this as well with piecewise hazard ratios, hazard ratios which cover only a specific time period. Now those are almost like small trials running simultaneously. So the point estimate varies, but really we see consistent benefit with asundexian all the way through the trial. Finally, we did a restricted cubic spline, which smooths out the curve. And again, apart from the tails, which are generally unreliable, a very consistent benefit for asundexian. So right now, in terms of the time course, it looks like asundexian is active very quickly, and we do not see a drop-off in its efficacy. We did look at three different pre-specified net clinical benefit outcomes, all of which were in favor of asundexian. These included measures which had all-cause mortality, ischemic stroke, as well as hemorrhagic stroke. All of them favored asundexian. So thus far, the data is looking very consistent and very good. This should be an easy medication once approved to translate into practice.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.