A recent issue of Revista Española de Cardiología published comments by the Working Group of the Spanish Society of Cardiology1 on the new atrial fibrillation (AF) guidelines of the European Society of Cardiology.2

These guidelines make only a few changes regarding anticoagulant therapy; however, biomarkers could aid decision-making in inconclusive cases where the patient has only 1 risk factor, as mentioned by the Working Group of the Spanish Society of Cardiology.

One change made in the guidelines is that the clinical pattern of arrythmia should not affect the thromboprophylaxis indication, which is listed as a class III recommendation, rather than as the class IIa recommendation included in the 2010 guidelines.3 This recommendation is supported by a single reference published in 2000 and is based on the Stroke Prevention in Atrial Fibrillation (SPAF) studies,4 in which the authors state verbatim in the limitations section that the risk of “stroke associated with intermittent AF is likely to be related to the frequency and duration of paroxysms... [which were] not accurately ascertained in this study.”

In recent years, several articles have discussed the influence of AF pattern on thromboembolic risk, with most concluding that nonpermanent (mainly paroxysmal) AF carries a lower risk. The same conclusion was also supported by analyses of patient subgroups included in key studies on direct-acting anticoagulants.

In the ARISTOTLE trial5 analysis, the yearly rate of stroke or systemic embolism in patients with paroxysmal AF, compared with persistent and permanent AF, was 0.98 vs 1.52 (hazard ratio [HR], 0.65; 95% confidence interval [95%CI], 0.48-0.87; P=.003).

In the ROCKET trial,6 patients with paroxysmal AF had a significantly lower rate of stroke: 1.73 vs 2.18 (HR, 0.78; 95%CI, 0.61-0.99; P=.045).

In the ENGAGE trial,7 the primary endpoint was reached less often in paroxysmal AF (1.49%/y) than persistent (1.83%/y) or permanent (1.95%/y) AF, with these differences being statistically significant.

However, although the event rate in these trials was lower in paroxysmal AF, it was still high enough to warrant anticoagulant therapy, even though we should keep in mind that these studies included high-risk patients.

Last, a relevant aspect in this topic is the design of the NAVIGATE8 and RESPECT9 trials. Both of these trials included patients with a history of stroke and compared rivaroxaban or dabigatran therapy with anticoagulant treatment. In both cases, prior monitoring was performed for at least 20 hours to rule out AF episodes longer than 6 minutes; in other words, the authors considered that patients with a history of stroke and AF lasting <6 minutes could be treated with anticoagulants.

While we agree that thromboembolic prophylaxis should be determined by the thromboembolic risk profile rather than the clinical pattern of arrythmia, we believe that the available information means that the AF type can be considered in decision-making about inconclusive cases where the patient is not at high risk, with a level of evidence not inferior to the evidence level attributed to biomarkers.