In Terms of Stroke Risk, Not All Atrial Fibrillation Is Created Equal

John Mandrola


November 17, 2014

A study presented today at the American Heart Association (AHA) 2014 Scientific Sessions may have upended one of the longest-running dogmas in cardiology. For years, decades even, cardiologists have considered the stroke risk of short-duration atrial-fibrillation episodes equal to that of persistent or permanent AF. Current guidelines for AF treatment do not distinguish types of AF.

That decree has never made sense to me. Have you tried explaining to a curious well-informed AF patient that his intermittent short-duration AF episodes are as risky as long-lasting AF? It's not easy. You often end up saying it just is, or trust me.

Think about it. If stasis in the atria is the mechanism of thrombus formation and stroke, then, on statistical grounds alone, long-duration episodes should be riskier. Another intuition is that persistent AF associates with progressive atrial structural disease, and it's well established that atrial structural disease correlates with stroke risk.

Let me show you one study from AHA, and then two supporting abstracts from the latest European Society of Cardiology Congress. Then you decide if we should distinguish on AF type.

The AHA Study

A group of Canadian researchers, including Dr Stuart Connolly (McMaster University, Hamilton, ON), performed a systematic review of existing literature[1]. They found 18 papers with 134 847 patients included in studies that separated out paroxysmal AF (PAF) or permanent AF. The first analysis looked at stroke rates in patients with PAF vs perm AF in patients off oral anticoagulation (OAC.) The odds ratio favored PAF patients at 0.75 (95% CI 0.61–0.93). The second analysis considered the same comparison in patients on OAC. The odds ratio again favored PAF patients at 0.77 (95% CI 0.68–0.88). The third analysis looked at patients on mixed OAC use. Again the odds ratio favored PAF patients, although the authors point out there was substantial heterogeneity in studies with mixed OAC use.

The researchers concluded that existing literature suggests a consistently lower stroke risk in patients with PAF than with permanent AF. They admit that whether AF pattern is an independent predictor of stroke or rather a reflection of a patient's stroke risk profile cannot be concluded from this review due to the heterogeneous study designs and AF populations.

I'm no statistician, but these findings look like a strong signal. And it aligns with what clinicians suspect. Now to the two studies presented at ESC 2014.

Two Studies from ESC 2014

ESC Abstract 4871 : Investigators from the US and UK used data from the ROCKET-AF trial to assess the risk of stroke and death in patients with paroxysmal vs persistent AF receiving oral anticoagulation[2]. Recall that ROCKET-AF compared warfarin and rivaroxaban (Xarelto, Bayer Pharma/Janssen Pharmaceuticals) in 14 062 patients, 11 548 (82%) with persistent AF and 2514 (18%) with paroxysmal AF.

Patients with persistent AF had significantly higher adjusted rates of stroke and all-cause mortality. For stroke, the numbers were 2.18 vs 1.73 events/100 patient-years (P=0.048). For all-cause mortality, the numbers were stronger (4.78 vs 3.52; P=0.006).

These relationships were not altered by assignment to either rivaroxaban or warfarin.

The authors concluded patients with persistent AF and moderate stroke risk have a higher risk of events and worse survival compared with paroxysmal AF.

ESC Abstract 4182 : Researchers from Canada, Germany and the Netherlands analyzed rates of stroke in the 6563 aspirin-treated AF patients included in the ACTIVE-A/AVERROES database according to AF presentation[3]. They found "yearly stroke rates were 4.2%, 3.0%, and 2.1% for patients with permanent, persistent, and paroxysmal AF, respectively, with an adjusted hazard ratio of 1.83 (P<0.001) for permanent vs paroxysmal and 1.44 (P=0.02) for persistent vs paroxysmal." In multivariate analysis, AF type was an independent predictor.

They concluded that in a nonanticoagulated contemporary group of AF patients, the type of AF was a strong predictor of stroke risk, and therefore the clinical presentation of AF may be helpful in assessing the benefits and risk of anticoagulants, especially in low-risk patients.


Can you see the narrative forming here? It's always been intuitive that persistent forms of AF would confer more risk than short-duration, infrequent episodes. Now there looks to be compelling data that are consistent through multiple studies, use contemporary patients, and hold up in both nonanticoagulated and anticoagulated groups.

Consider the low-risk patient with permanent AF. Is he really that low risk? Likewise, consider two patients with a CHADS-VASC of 1 or 2. One has 30 minutes of AF weekly and the other has permanent AF. Are the risks the same? Shouldn’t the type of AF factor in the risk/benefit decision of anticoagulation? These studies suggest the answer is yes.

I would be interested in your comments on this strong and potentially practice-changing data.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.