ISC 2024 | ANNEXA-I: superior hemostatic efficacy with andexanet alfa vs SoC in FXa inhibitor-related ICH

So ANNEXA-I was a Phase IV, multicenter, prospective, randomized, open label, blinded endpoint trial in patients with acute ICH treated with factor Xa inhibitors. Patients were eligible if they were over 18 years of age, presenting with an acute intracerebral hemorrhage within six hours of symptom onset, who had taken the last dose of factor Xa inhibitor within 15 hours of randomization, and the ICH volume was confined to 0.5 to 60 cc’s. These patients were randomized 1:1 to receive andexanet alfa or usual care, as per the discretion of local treating physician and investigator. The primary efficacy endpoint for this trial was effective hemostasis at 12 hours, and this was defined by meeting all three of the following criteria: having 35% or less growth in hematoma volume between baseline and 12 hours, having no greater than a six-point increase on the NIHSS at 12 hours, and then also no rescue therapy between 3 and 12 hours after randomization. Secondary endpoints also included the change in percent from baseline to nadir in anti-factor Xa activity during the last two hours after randomization, as well as selected safety endpoints such as thromboembolic events at 30 days and 30-day mortality.

However, in June of 2023, the study was stopped early for overwhelming efficacy with andexanet alfa versus usual care in achieving hemostatic efficacy. And then by the time the decision was made, an additional 80 participants had been enrolled for a total study population of 530 participants. Baseline variables were evenly distributed between both treatment arms, and the mean age was about 80, which is not unsurprising in this elderly high risk population who has bleeding in the brain. Roughly half of participants were female. 1 in 10 had prior history of MI and 1 in 5 had prior history of stroke. The indication for anticoagulation was predominantly atrial fibrillation at 87%, and this was a high-risk population with the CHADS-VASc score median of four. Baseline hematoma volume was roughly 10cc’s, which is just a bit less than what you would expect in this population of factor Xa inhibitor-associated ICH, where cohorts have suggested roughly about a 13cc median and this is likely due to the fact that we excluded patients with higher baseline ICH volumes above 60cc’s, as well as those that had presented with low GCS. The time from symptom onset to baseline scan was 2.2 hours, scan to randomization time was one hour, and door- to-needle time was two hours, and the median time from symptom onset to treating was about four hours.

Overall study results indicated that andexanet alfa led to a large absolute reduction in not achieving hemostatic efficacy. So for every 100 patients treated, 13.4 patients were more likely to achieve hemostatic efficacy with andexanet versus usual care. And it’s important to note that the usual care arm actually included 87% of PCC. So in the people assigned to usual care, 87% received PCC. And in the patients where we have information on the type and dose of PCC they received, over 90%, received 4-factor-PCC and the median dose was 3000 units, which is usually the max dose that’s highlighted in guidelines. So we’re confident that this is an adequate comparison of andexanet alfa versus the best standard of care PCC currently in use. Those proportions that I just mentioned in terms of absolute benefit with andexanet for achieving hemostatic efficacy was in the initial population of 450 participants, where that led to the interim analysis termination of the study. However, in the extended population of 530 participants, it was 11 patients per every 100 treated who benefited from achieving hemostatic efficacy.

And this was actually predominantly led by excellent hemostatic efficacy, which is even more conservative at no greater than 20% growth in the hematoma. And when we also looked at another outcome in an exploratory fashion because of its high correlation with poor outcome at 90 days, and that’s an absolute hematoma increase of 12.5cc’s or greater, and this has been previously shown to have an 80% positive predictive value for an mRS of 4-6 at 90 days, andexanet reduced the likelihood of a hematoma growing beyond a 12.5cc threshold in 7.4 of every 100 patients treated.

In a subgroup analysis of efficacy, what was kind of interesting to see was that the subgroup of patients who were destined to receive PCC versus the subgroup of patients in the usual care arm who was destined to receive no PCC, behaved essentially no differently in that the effect size of andexanet versus those that received PCC was statistically the same as those destined to receive no PCC. And actually, numerically, it was even a bit greater in those destined to receive PCC. And when we combine that with the fact that the baseline anti-factor Xa activity was reduced by 95% with andexanet alfa versus only 24% in usual care, and also some previous data suggesting that, one, PCC does not reduce anti-factor Xa activity any more effectively than placebo, and that ,two, although there is an improvement in return to endogenous thrombin potential or normal levels of endogenous thrombin potential with PCC versus placebo, it takes about eight hours for it to do it, in contrast to andexanet alfa, where you have immediate normalization of endogenous thrombin potential. All this together lends me to at least question whether PCC is providing any benefit whatsoever in mitigating hematoma growth in patients presenting with acute ICH.

However, the benefits of andexanet alfa came at some cost and that was excess thrombotic events at 30 days. There was for every 100 patients, five more roughly thrombotic events with andexanet alfa versus usual care. The rates of thrombotic events at one month were 10% in those who received andexanet alfa, versus about 6% in those who received usual care. And this was particularly actually driven by ischemic stroke that occurred in 6.5% of those assigned to andexanet alfa versus about 1.5% of those assigned to usual care. There was no effect in all-cause mortality within andexanet versus usual care at 30 days and we also found no difference in poor outcome on the mRS at 30 days. But it’s worth knowing that this trial was never powered to detect a difference in mortalit or mRS when it was initially sampled from 900 participants, and particularly after being stopped early with the total sample now of 530 participants, we really don’t have the power to give any good estimates of treatment effect for the mortality outcome. And when it comes to mRS, not only are we limited by power, but also we know that to see a treatment effect of acute interventions manifest themselves in patients with acute ICH, typically we need six months or maybe even 12 months of follow up for treatments to clearly differentiate themselves. And this study was just not designed to detect that with the 30-day follow up which was quite brief.

Now, when looking at subgroup analyzes of patients who had treatment response to thrombotic events and looking for any heterogeneity in treatment effect of andexanet versus usual care for this outcome, what was interesting is that almost all of the excess thrombotic events with andexanet alfa versus usual care was seen in patients who entered the study with a prior history of an arterial ischemic event, either an ischemic stroke and MI. So one way to perhaps use this data in clinical practice, when trying to weigh the benefits of andexanet alfa for reducing hematoma expansion versus its potential excess risk of thrombotic events is excluding patients who have a history of arterial ischemic events where the rate of thrombotic events is the highest.

So in general, I’ll say that an ANNEXA-I has, one, proven the efficacy of andexanet alfa versus PCC in terms of reversal of anti-factor Xa activity. It has also proven its efficacy for achieving hemostatic efficacy relative to PCC. And that the subgroup analyzes and overall totality of data raise significant questions about whether PCC is providing any benefit at all in patients with acute ICH occurring on anti-factor Xa activity.