Optimizing oral anticoagulation management for patients with AF

Hello. It's a pleasure to talk to you about optimizing oral anticoagulant management in patients with atrial fibrillation.

Before I do, just my disclosures which include a number of companies relevant to this talk.

We've got some excellent guidelines for the management of patients with atrial fibrillation from the ESC released in 2020. And these promote the use of NOACs instead of VKAs in patients with atrial fibrillation excluding those with mechanical heart valves or moderate to severe mitral stenosis where there is better evidence for the VKAs like Warfarin.

And they promote a risk factor based approach to deciding whether to use oral anticoagulation, here using the CHA2DS2–VASc score. And those with a score of nought in men or one in women, don't necessarily need to be offered any antithrombotic therapy. Whereas those with a score of two in men or three in women should be with a class one recommendation, recommended oral anticoagulation.

Now how good is this score for predicting stroke? It's a well-established score, but we are learning more. So there are studies now showing that biomarkers may provide more information about stroke risk and give a more refined estimate of stroke risk and potentially guide anticoagulant therapy in the future.

And the ABC AF stroke score shown here shows if you add NT-proBNP or cardiac troponin T to age and clinical history that can provide a more accurate prediction of stroke risk. So this is something we should see in the future. But for those with high NT-proBNP, that is potentially an indication for antithrombotic therapy in someone who may not otherwise be indicated for it.

In those with intermediate to CHA2DS2–VASc scores, one in men and two in women. They have a class IIa recommendation for anticoagulant therapy. But we also have recommendations related to bleeding risk. And so it's recommended to assess bleeding risk in patients. Firstly, so that you can consider modifying bleeding risk, but secondly, also to have more prolonged or more frequent follow-up of patients to review their tolerance of treatment. And the HAS-BLED score is recommended in the guidelines.

Now again, more recent evidence here, showing the ABC AF bleeding score shows how adding biomarkers may refine risk prediction. Looking at hemoglobin, cardiac troponin T, but also GDF-15, which seems to discriminate bleeding risk and may help to refine risk prediction in these patients.

So this is something potentially for the future. As to how we're becoming more sophisticated in predicting both stroke and bleeding risk. Those recommendations for when you should switch from VKAs to NOACs, such as if patients have suboptimal time in therapeutic range on a VKA. That's a reason to switch to a NOAC with predictable response.

But it's important that you select the right dosis as I'll discuss in a minute. But it's also important to emphasize that antiplatelet therapy alone is not an alternative to oral anticoagulation. There's no particular gain in safety and it's less effective. And also the pattern of atrial fibrillation should not influence the decision for anticoagulant therapy.

And here are the dose selection criteria for the different NOACs. We have a range of dose reduction criteria. And if you note for apixaban, you have to have at least two or three criteria to reduce the dose from five mg twice daily to two point five milligrams twice daily. If the patients just have one of these criteria, that's not sufficient to reduce the dose. And likewise, it's important to follow these data reduction criteria accurately and not inappropriately use the lower dose.

We have a big challenge with patients, who have an indication for dual antiplatelet therapy, to treat or prevent or both coronary thrombosis. And those patients with atrial fibrillation who have a risk of cardiac thromboembolism. And traditionally, we use anticoagulants for AF and we use dual antiplatelet therapy for preventing and treating coronary thrombosis.

But if we combine dual antiplatelet therapy with anticoagulants, then we're having quite a profound effect on hemostasis and driving a high bleeding risk. And partly this relates to the fact that anticoagulants indirectly or directly inhibit not only coagulation but also thrombin induced platelet activation. And so if you use an anticoagulant with two antiplatelet drugs, aspirin and a P2Y12 inhibitor, you're actually targeting three pathways on platelets. And that may be too much to cause an excess bleeding risk and potentially not necessary in terms of preventing coronary thrombosis.

But what about the evidence? We've had trials looking at all the NOACs in patients with AF and undergoing PCI or ACS patients or both. And these trials have been powered for assessing bleeding outcomes, but they've been underpowered for ischemic endpoints, which leaves some doubt and uncertainty in the evidence base.

The AUGUSTUS study was a two-by-two that allowed us to show firstly, reduced bleeding with a NOAC Apixaban uses its license dose for atrial fibrillation versus a VKA such as Warfarin. And it also showed the profound reduction in bleeding, when using placebo instead of aspirin in a double-blind fashion. So the study clarified the bleeding risk associated with firstly, the anticoagulant class and secondly, the use or not of aspirin.

But what about ischaemic outcomes? There was a trend towards less stroke with Apixaban versus VKA. Death in ischemic events was similar in terms of rates. But when we look at aspirin versus placebo, again, the rates of death in ischemic events are similar but there was numerically, although not statistically, less stent thrombosis with aspirin versus placebo. The majority of these patients received clopidogrel.

So, the guidelines took a pragmatic response to this evidence base to say: Look at the risk of stent thrombosis in an AF patient, who's undergone PCI. Look at their bleeding risk. If the risk of stent thrombosis is low or the bleeding risk is greater than the stent thrombosis risk what we mean by that is defined at the bottom of the table. Then stop aspirin early, within a week, potentially even straight after PCI. But if the risk of stent thrombosis is more considerable then you can use triple therapy with aspirin and clopidogrel and preferably a NOAC. Where the total duration decided according to the assessment of these risks, but generally no more than a month. So class IIa recommendation.

But we've got a problem here because when we recommend clopidogrel to an individual patient we don't know necessarily what the response to clopidogrel is going to be. Even though we genotype the CYP2C19, that only takes us a small part of the way towards predicting the response to clopidogrel.

And in these two studies shown here, we compared on the left ticagrelor two different doses pre and post maintenance dose with placebo and on the right, we compared reusing the same assay with clopidogrel. You can see in between the green lines quite an overlap between platelet reactivity on clopidogrel and platelet reactivity on placebo. Whereas with ticagrelor you get a much more consistent and reliable level of P2Y12 inhibition.

So, unless you actually do the assay to measure clopidogrel response, you don't quite know where you are. And even if you do that in an AF patient, we don't know how that should influence the therapeutic decision making.

But the guidelines, again, support dual therapy with an oral anticoagulant and either ticagrelor or prasugrel with a IIb class recommendation as an alternative to using triple therapy with aspirin and clopidogrel and an anticoagulant in those with a moderate or high risk of stent thrombosis. So, this is quite a nice strategy, because it allows you to ditch aspirin early on and not worry about the variable response to clopidogrel using a potent P2Y12 inhibitor. And as your second antithrombotic drug, an oral anticoagulant. But there's limited data on this at the moment. So, the balance of safety and efficacy is yet to be well defined. But it's important to stress that you shouldn't use triple therapy with ticagrelor or prasugrel because of the very high bleeding risk.

What about the long term? Once we're beyond this risk of stent thrombosis in patients with AF. Should we continue any antiplatelet drugs or should we stop them all and just use anticoagulant monotherapy. The AFIRE study addressed this, looking at rivaroxaban and aspirin versus rivaroxaban alone in patients with coronary artery disease. And this showed on the left, actually, fewer ischemic events with anticoagulant monotherapy compared to combination therapy and less bleeding with monotherapy. So, as a default strategy this trial and others suggest that anticoagulant monotherapy is the optimum strategy for the majority of patients.

But there's always exceptions to the rule. So, certainly, for all these CCS patients they score at least one point with the CHA2DS2–VASc score, because of modification of the vascular disease definition, to include coronary artery disease evidence. But for most of the patients we are not going to continue antiplatelet therapy. For those who have a high risk of recurrent ischemic events and do not have a high risk of bleeding the chronic coronary syndrome guidelines gave a IIb class recommendation for aspirin possibly clopidogrel, although there is less evidence for bleeding risk with that for long-term therapy. This has to be reserved for those very special patients with high ischemic risk and without high bleeding risk, so there's not many patients. The majority will receive monotherapy.

Finally, the ESC guidelines do cover the issue of patients who have a device and have high atrial rate episodes on that or subclinical atrial fibrillation. Clear guidance in the guidelines, shown here, as to how you look, firstly at stroke risk by assessing the CHA2DS2–VASc score, potentially in the future incorporating biomarkers and then also look at the burden of atrial high-rate episodes on the device monitoring. For those who are at increased risk, according to the CHA2DS2–VASc score and have a higher burden of these atrial high-rate episodes then consider them for anticoagulation.

So, to conclude, we have emerging risk scores for stroke and bleeding that incorporate biomarkers and improve risk prediction over the currently recommended scores and that's something to watch for in the future as we become more sophisticated. Perhaps using artificial intelligence to individualize therapies.

NOACs are preferred to vitamin K antagonists in eligible patients and should be used at the licensed dose without inappropriate dose reduction. NOACs indirectly or directly inhibit thrombin-induced platelet activation, potentially allowing early cessation of aspirin after PCI in order to reduce the bleeding risk without necessarily increasing the thrombotic risk. But although clopidogrel is recommended in dual therapy with NOAC after PCI there was uncertainty about the stent thrombosis risk with this combination. So perhaps a three-month course of a potent P2Y12 inhibitor such as ticagrelor without aspirin may provide a better balance of safety and efficacy. We need more work on this. Monotherapy with anticoagulant therapy is the default option in chronic coronary syndrome patients who have atrial fibrillation but occasional patients with high ischemic risk and who don't have a high bleeding risk as defined in the guidelines, may warrant dual therapy. But combined with a proton-pump inhibitor because of the higher risk of gastrointestinal as well as other bleeding. And clear guidance, now for those with high atrial rate episodes on device monitoring in terms of considering the burden and also considering the stroke risk, the thrombotic risk, in order to guide antithrombotic therapy.

I like to thank you very much indeed for your attention to this talk. I hope you've found it of interest.