Category Archives: EBM Summaries

Thromboprophylaxis in AF – An Evidence-Based Overview

Thromboprophylaxis in AF – An Evidence-Based Overview

Dr Mark Garside – registrar in elderly care and stroke medicine

The Royal College of Physicians (Edinburgh) recently held a consensus conference about the management of atrial fibrillation (AF), with the aim of producing a set of updated evidence-based guidelines on the management of this commonly-encountered condition. AF carries with it a significant negative implication for mortality and morbidity. In particular, there is an increased risk of stroke in people with AF, and stroke prevention is an important treatment goal. What follows is a brief overview of the evidence surrounding treatment decisions in AF. The aim of this post is not to scrutinise each individual publication, but rather to highlight the most important point from the literature that should inform clinical treatment decisions.

Whilst it has long been accepted that ‘high risk’ patients with AF (ie. patients with other vascular risk factors) benefit from formal anticoagulation with warfarin, there has been less certainty about the best treatment for patients with fewer co-existing risk factors. Until recently, it has been common clinical practice to treat patients with aspirin as a substitute for warfarin if they are either felt to be at lower risk of stroke, or if warfarin was deemed to be unsuitable for some reason.

Risk stratification tools such as CHA2D2S-VASc can be used to quantify patients’ risk of having a stroke, but it is important to note that all but the lowest risk patients (CHA2D2S VASc=0) get a net clinical benefit from anticoagulation treatment.

An important point to make is that the new guidelines from the consensus conference recommend that aspirin should no longer be used for thromboprophylaxis in AF. Not only is this an evidence-based recommendation, but it is also consistent with the pathology of thromboembolic stroke in AF, where the thrombus is mainly fibrin-based, compared to the platelet-rich thrombi that predominate in coronary artery disease.

The evidence for the use of aspirin in AF is actually very weak. Although a meta-analysis by Hart et. al. (1) suggested there may be approximately a 20% relative risk reduction with aspirin compared to placebo, there are some problems with this. Firstly, this reduction did not quite reach statistical significance. Secondly, the methodology in the studies analysed varied widely, not least in the range of different doses of aspirin that were used. Finally, it  appears that the results of a single study, SPAF-1, are skewing the results of the meta analysis to make aspirin appear more effective than it is actually likely to be. SPAF-1, which compared aspirin to placebo in 2 different subgroups (one with patients that would have been suitable to receive warfarin treatment, and one with patients who were though to be unsuitable for warfarin) produced some anomalous results, with only one subgroup (the one with patients who we now know should have been on warfarin) showing a benefit of aspirin (2).

For low-risk patients – ie. those with “lone” AF, who would have a CHA2D2S-VASc score of 0 – an RCT in Japanese patients comparing aspirin to placebo was stopped early due to an increased rate of bleeding in the aspirin arm without any suggestion of a benefit in stroke prevention (3).

Even in the elderly population, who are often not anticoagulated due to a perceived high risk of bleeding, trial data has shown that not only is warfarin significantly more effective  than aspirin in terms of stroke prevention, but also that it is no more dangerous in terms of causing significant bleeding (4).

That is not to say that the risk of bleeding should be ignored, and the HAS-BLED scoring tool has been developed to try and quantify that risk. However, several of the components of the HAS-BLED score also appear in the CHA2D2S-VASc score, reflecting the fact that risk factors for bleeding overlap with risk factors for ischaemic stroke. The HAS-BLED score was not designed to be used as a tool to justify not treating patients with warfarin, but rather to identify high-risk patients so that they could be more closely monitored, and that modifiable risk factors could be addressed.

But warfarin is not without its drawbacks. Although we have many years of experience with it, it dose not have a consistent dose-response, requires careful and regular monitoring, has many drug interactions, is often inconvenient for the patient, and only has a net clinical benefit if over 65% of the time is spent within a therapeutic INR range.

Although they are in their infancy, there are three new oral anticoagulants which may provide a suitable alternative treatment to warfarin, and solve many of the above problems. Dabigatran (a direct thrombin inhibitor), Rivaroxaban and Apixaban (both factor Xa inhibitors) have all been shown to be at least as effective as warfarin for stroke prevention in AF, and at least as safe in terms of bleeding risk (5,6,7). In particular, it’s worth noting that the higher dose of dabigatran has been shown to be significantly more *effective* than warfarin with a similar rate of bleeding, whereas the lower dose is significantly *safer* in terms of bleeding but has similar efficacy to warfarin in terms of stroke prevention. This could prove to be important for clinicians, as the choice of doses allows for extra caution to be taken in patients felt to be at high risk of bleeding.

In addition to comparing itself against warfarin, Apixaban has also been trialled against aspirin in a patient population who were deemed too high-risk for warfarin(8). The results showed it to be significantly more effective than aspirin at stroke prevention, with a similar bleeding profile. So in the near future this may also become a realistic treatment option for patients who are not suitable for warfarin treatment.

Weighed against these advantages though is the fact that there is no specific antidote, which may be problematic in acutely unwell and bleeding patients. However, although there are concerns and as-yet unanswered questions about how well these new agents will work in the real world, the fact remains that the more pressing risk to patients is the risk of ischaemic stroke, and trying to mitigate this risk through anticoagulation should be a priority wherever possible.

In summary, there is no evidence to support the use of aspirin for thromboprophylaxis in AF, and it may even risk net clinical harm to patients. Anticoagulation is the preferred treatment, with warfarin being safer and more effective that is often perceived, even in an elderly population. Newer oral anticoagulants seem to be at least as effective as warfarin, and in many cases safer, but come with a significantly higher cost attached.

For now, the take-home clinical message is to be aware of AF as an important clinical problem, and to give serious thought about what the best treatment is likely to be for your patients.

Further Reading:

For a more detailed overview of this topic, including information about the trials mentioned above, the following review is recommended:

Bassand, JP; Review of atrial fibrillation outcome trials of oral anticoagulant and
antiplatelet agents Europace (2012) 14(3): 312-324


(1) Hart et. al. “Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation.” Ann Intern Med 2007;146:857-67

(2) Stroke Prevention in Atrial Fibrillation Investigators. “Stroke Prevention in Atrial Fibrillation study. Final results.” Circulation 1991;84:527-39

(3) Sato et. al. “Low-dose aspirin for prevention of stroke in low-risk patients with atrial fibrillation: Japan Atrial Fibrillation Stroke Trial.” Stroke 2006;37:447-51

(4) Mant et. al. “Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised controlled trial” Lancet 2007;370:493-503 DOI: 10.1016/S0140-6736(07)61233-1

(5) Connolley et. al. “Dabigatran versus warfarin in patients with atrial fibrillation.” N Engl J Med 2009;361:1139-51

(6) Patel et. al. “Rivaroxaban versus warfarin in nonvalvular atrial fibrillation.” N Engl J Med 2011;365:883-91

(7) Granger et. al. “Apixaban versus warfarin in patients with atrial fibrillation.” N Engl J Med 2011;365:981-92

(8) Connolley et. al. “Apixaban in patients with atrial fibrillation.” N Engl J Med 2011;364:806-17

Posted in EBM Summaries, Uncategorized | 1 Comment