We performed some secondary analyses that we were just presented in the International Stroke Congress a couple of weeks ago. And that was looking at predictors of hematoma expansion and what kind of patient variables can be used to select patients who are at greatest risk of expansion. where we anticipate the benefits of an X and an alpha to be amplified. So in these analyses, we selected only patients who were participating in an X and I who had baseline and follow-up imaging and who had an intrapranchymal or intraventricular hemorrhage. So we excluded patients with subdurals and subarachnoids as the primary location of bleeding.
And that left us with about 460 participants. Baseline demographics were evenly distributed amongst those assigned to index-net alpha unusual care in this particular sample. And really, the values of these baseline demographics was quite generalizable to the overall trial population.
And we looked at the primary outcome that we defined as a growth of hematoma from baseline to 12 hours of no more than 12.5 cc's or greater as an absolute measure kind of cutoff or a proportional cutoff of 35 percent or greater and as i mentioned earlier this 12.5 cc cutoff is well established to have a high positive predictive value for poor outcomes at 90 days mrs4 to 6 and 90 days in particular a 90 positive predictive value for this and And the 35% or 33% cutoff has about a 70% positive predictive value for an MRS 4 to 6 at 90 days. And we looked at this in the predictors in two logistic regression models. In model one, in addition to treatment with indexinit, we looked at time from symptom onset to treatment, baseline anti-faculty activity. as a measure of degree of anticoagulation and then three baseline ICH volume.
And in the second model, we looked at a metric of average pre-scan hematoma growth. This has also been termed in the literature as ultra-early hematoma growth and was initially proposed by David Rodriguez Luna in Barcelona. And... This was calculated by dividing the baseline hematoma volume by the time from symptom onset to baseline scan. In these logistic regression models, we found that there was an inverse association between time from symptom onset to treatment and hematoma growth, in that the earlier patients presented, the more likely they were to expand.
Two... that there was a positive association with baseline ICH volume. So the larger the baseline ICH volume, the more likely they were to expand. And then three, that the strongest predictor was actually baseline hematoma growth rate or pre-scan hematoma growth rate. And that the higher the average growth rate prior to the first baseline scan, the more likely the patients were to expand as well.
Now, although we found an association that had a strong effect size in univariate analysis with anti-factor tanning activity, this was really washed out. At least its significance was washed out, although the effect size was maintained in the regression model. And part of that may have been due to collinearity with some of the other variables. And when we mapped the risk of expansion across the range of these metrics, essentially we found a consistent...
kind of almost linear relationship with increasing baseline ICH volume and risk of expansion, reaching close to 60% risk of expansion in patients who were at the highest quartile for baseline ICH volume, which was about 22 cc's. Similarly, in those that had the highest quartile of baseline pre-scan hematoma growth rate, which was 11.4 cc or greater. There were about a 60% risk of expansion. And that across time as well, those who presented within their first quartile of time from symptom to treatment, they were at that the first quartile was less than 3.3 hours. They were also at roughly 50% risk of expansion.
Now, this doesn't mean that those that were at the kind of lowest quartile of hematoma volume or expansion rate or highest quartile of time had negligible expansion rates. There were about one in five for time and one in four for baseline hematoma volume and growth rate. So still about a 20, 25% expansion rate, even in those that were deemed to be low risk in our analyses.
And what really is the question we're asking is, well, what was the effect of indexinit versus usual care across these metrics? And it was actually constant. So the proportional reduction was quite constant across the range of these metrics. However, because patients were at higher risk, if they had larger ICH volumes in particular, or if they had higher growth rates, those that reached the highest quartile for these two measures had a much greater absolute benefit from adexanet, where it was 25 patients who benefited for every 100 treated. And that actually translates to a number near three to four.
So quite a robust number near three to prevent hematoma expansion. And importantly, none of these metrics had a relationship with excess thrombotic events. And thus, overall, our analyses would suggest that the One, the number needed to treat across all quartiles was less than 10 for any of these metrics, so robust across the range. Even more amplified in those that had the highest quartile baseline ICH volume and baseline hematoma growth rate, and that was a four number needed to treat. And a number needed to harm stayed around 26 across the range of any of these metrics.
And thus, by selecting even higher risk populations, Through the use of these baseline variables, we can even amplify the benefit of indexing it. Now, one important piece is that because we did not see a difference in clinical outcomes and we don't have the power for it, to interpret this data, the clinicians also want to know what was the clinical consequence of ischemic stroke so that they could balance it versus the clinical consequence of hematoma expansion. And that's going to be the focus of a subsequent secondary analysis that we're going to be presenting at the ESOC, which I think will really help bring all this data together.