The AF and ICH Anticoagulation Conundrum

This transcript has been edited for clarity. Dear colleagues, I’m Christoph Diener from the Faculty of Medicine at the University of Duisburg-Essen in Germany. This month, I would like to concentrate on a particular clinical problem and that is patients with atrial fibrillation who have had an intracranial or intracerebral hemorrhage. If we don’t initiate anticoagulation, these patients have a high risk of having an ischemic stroke or a recurrent ischemic stroke.Stroke Reduction, Bleeding Risk Data From Prior Studies And if we initiate anticoagulation, they also have a high risk for intracerebral hemorrhage. So what do we know about this problem from randomized trials? The first trial was the SoSTART trial in the UK. They recruited 203 patients with atrial fibrillation and intracerebral hemorrhage, and compared oral anticoagulation vs no oral anticoagulation. They found that recurrent symptomatic intracerebral hemorrhage was higher with anticoagulation (8 vs 4), and ischemic strokes were significantly more frequent without oral anticoagulation (19 vs 3 with anticoagulation). The second study was the APACHE-AF study in Netherlands. They also included patients with atrial fibrillation and intracerebral hemorrhage. Fifty patients were treated with apixaban, and 51 patients received no anticoagulation. For the primary endpoint of nonfatal stroke and vascular death, there was no significant difference (26% vs 24%), but bleeding complications were more frequent with apixaban.There was a first meta-analysis called COCROACH involving 412 patients with atrial fibrillation and intracranial hemorrhage. When they compared anticoagulation with no anticoagulation, there were 9 major ischemic events vs 38 for no anticoagulation, and there were 12 intracranial hemorrhages vs 5. So they found a reduction in ischemic stroke and an increase in intracranial hemorrhage with anticoagulation.The next study was PRESTIGE-AF. This compared direct oral anticoagulation vs no anticoagulation in 319 patients. Again, there was a significant reduction in ischemic stroke with direct oral anticoagulants and an increased risk for intracerebral hemorrhages. The number needed to treat to prevent an ischemic stroke was 13, and the number needed to harm to cause an intracerebral hemorrhage was 24.An updated meta-analysis that included now four randomized trials in 653 patients with atrial fibrillation and intracerebral hemorrhage found a relative risk reduction for ischemic stroke of 0.23, and an increased risk for intracranial hemorrhage with a relative risk for 3.60. Fortunately, there was no difference in mortality.Enter ENRICH-AF at ESC Now, at the European Society of Cardiology (ESC) Congress in Munich a few days ago, the ENRICH-AF study was presented. This is by far the largest study. The study included 948 patients with atrial fibrillation and intracranial hemorrhage. These patients were randomized to edoxaban or no edoxaban. And no anticoagulation meant either nothing or antiplatelet therapy.At an interim analysis, patients with lobar hemorrhages were no longer included because they clearly had an increased risk for recurrent bleeding. After 26 months, there was no significant difference for stroke or systemic embolism, with rates of 0.8% for edoxaban vs 12.8% without edoxaban. There was a significant reduction in ischemic stroke and myocardial infarction with edoxaban, and a significant increase in major bleeding complications (11.6% vs 5.2% without edoxaban).Now, if we take all these studies together, they do not answer the critical clinical question: What should we do with patients with atrial fibrillation who have had an intracerebral or intracranial hemorrhage? If we anticoagulate them, they have an increased risk for intracerebral bleeding, and if we do not anticoagulate them, they have an increased risk for ischemic stroke.In a risk-benefit analysis, I think the consequences of an intracerebral hemorrhage are more severe than the consequences of an ischemic stroke, and this would be a strong argument to forgo anticoagulation these patients. But we have to wait for the next meta-analysis. Perhaps, if we combine all of these trials, we can identify a subgroup of patients who benefit from anticoagulation and have only a minor risk for recurrent intracerebral hemorrhage. This next COCROACH meta-analysis may not be available until the end of the year. So, dear colleagues, this is a very difficult issue, and at the moment, it’s unresolved in terms what to do best. I am Christoph Diener from the Faculty of Medicine at the University of Duisburg-Essen. Thank you very much for listening and watching.