Atrial Fibrillation beyond “All or None”: The Missing Piece in Understanding Atrial Fibrillation?

It has been more than a century since Thomas Lewis first explained and documented the irregularly irregular pulse we now know as atrial fibrillation (AF),¹ and we still try to grasp the various pathophysiological basis and mechanistic relationships of this complex disease. The AHA Scientific Statement on AF burden by Chen and colleagues² is an example of the ongoing efforts to expand our ever-evolving understating of this common rhythm disorder which currently affects 2.7 to the 6.1 million Americans, and the number expects to double by 2030.³

Over the past decades, AF has been mostly investigated and managed as an “all or none” condition; present or absent regardless of type or pattern. However, with the increasing use of long-term ambulatory ECG recording that accompanied recent advances in technology, more data became available on AF burden (quantity, amount or duration of AF relative to total ECG monitoring time). This triggered questions regarding the relationship between AF burden with outcomes, factors modifying AF burden, and the optimal methods to define and quantify AF burden. The AHA Scientific Statement by Chen and colleagues² systematically addresses these questions using published literature, and also point out gaps in knowledge and future research needed to bridge these gaps. The Statement is not intended to provide specific formal clinical recommendations on AF burden, but it is a comprehensive review of the current state-of-the-art science of AF burden.

There are three main observations one can make from this AHA Statement on AF burden.²

First- There is uncertainty about the presence of a linear dose-response relationship between AF burden and outcomes. Although several studies showed that persistent AF is associated with increased risk of stroke than paroxysmal AF, this was not always the case. Also, while brief periods of device-detected AF have been associated with increased risk of stroke, it was unclear whether the risk increases continuously in a linear dose-response fashion or whether a threshold exists.

Second- There is lack of enough data from well-designed large clinical trials to make informed inference about factors that could potentially modify the burden of AF in a meaningful way that impacts its associations with outcomes. Unlike AF as a binary condition, the literature on the impact of cardiovascular risk factors and lifestyle modification on AF burden is scarce. Issues with design, definition of AF burden, and statistical power are noticeable in the limited clinical trials and large observational studies examining risk factor modification of AF burden.

Third- There is no standardized, validated methods for detection and defining AF burden that fit the current technology of markedly long-term ECG monitoring. Almost all studies mentioned in the Statement used different arbitrarily cutoffs for AF burden, and all of them were not empirically derived. This complicates the comparison of the results among studies and subsequently makes it difficult to synthesize an informed clinical recommendation in the future.

Clearly, the AHA Statement on AF burden² is a call for rethinking our traditional understanding of AF. So, maybe we should ask first whether AF burden is really a forward-thinking concept with the potential to enhance our knowledge of AF and its relationship with its most common and most devastating complication, stroke. Investigating AF as a quantitative entity (i.e., AF burden) is definitely a paradigm shift from the century-old thinking of AF as a binary condition. From this perspective, AF burden is a forward-thinking concept.

On the other hand, AF burden in its core is defined based on detection of the abnormal rhythm, and hence all AF burden relationships with outcomes still have to be explained in the context of AF associations with the presence of the abnormal rhythm similar to AF as a single condition. From this perspective, our understanding of the relationship between AF and outcomes such as stroke seems to rest on a century-old hypothesis that the abnormal rhythm is the primary trigger for thromboembolism, a hypothesis that did not hold when tested in clinical trials.^4-8 The largest of these trials is the AFFIRM trial which showed that rhythm-control offered no better survival than rate-control and that more patients in the rhythm-control group were hospitalized than those in the rate-control group⁴. Also, in the few randomized clinical trials of prolonged cardiac monitoring after cryptogenic stroke, there was no significant benefit of oral anticoagulation for recurrent stroke prevention in patients with newly detected subclinical AF. ^9-11 This suboptimal response to anticoagulants when prescribed appropriately based on detection of abnormal rhythm in AF patients after cryptogenic stroke further supports disconnection between the abnormal rhythm and thromboembolism. The inconsistent temporal association between AF and ischemic stroke^11-13 is another indication for the disconnection between abnormal rhythm and stroke.

If abnormal rhythm does not explain the AF-stroke relationship, why should we expect AF burden, for which abnormal rhythm is its hallmark definition, to be the missing piece in our understanding of AF and its relationships with outcomes? As forward-thinking AF burden could be compared to AF as a single entity, both share the inability to explain several of the reported inconsistencies in AF-stroke relationship. This is probably because both depend on that old unproven hypothesis that abnormal rhythm is needed for thromboembolism. That is why there has been a call for a new model that does not depend on abnormal rhythm.^14,15 This model, atrial cardiopathy, does not consider AF as a standalone disease, but as a stage or a component of a continuum of atrial disease (AF is a marker of atrial disease) in which the risk of stroke exists but vary in severity at any stage of that continuum (atrial disease is the source of thromboembolism even in the absence of AF), and earlier stages of atrial disease could advance to AF at any time point either before or after stroke. This explains several of the observed inconsistencies in the AF-stroke relationship.^14,15

Markers of atrial disease in the atrial cardiopathy model include markers indicating structural, anatomical, electrophysiological or neuro-hormonal insult to the atria.^14,15 The electrophysiological markers include AF as well as atrial arrhythmias. The obvious increased risk of stroke in AF compared to other factors that could impact the atria accord with the fact that it is an indication of severe atrial disease. Nevertheless, since the extent of the underlying atrial pathology is the critical factor dictating the risk of stroke per the atrial cardiopathy model, it is expected that not all AF associations with stroke will be the same as observed among lone AF, post open-heart-surgery AF and typical AF. Differences in the underlying atrial pathology may also explain the inconsistent dose-response relationship between AF burden with outcomes in different populations and individuals. The basis for the atrial cardiopathy model has been tested in several cohorts, and the use of atrial cardiopathy to guide anticoagulants therapy after cryptogenic stroke is currently undergoing in the AtRial Cardiopathy and Antithrombotic Drugs In Prevention After Cryptogenic Stroke (ARCADIA) trial.¹⁶

So where to go from here? The elegant AHA Statement on AF burden by Chen and colleagues² fulfills its purpose of increasing knowledge and awareness about AF burden. Although it does not provide specific recommendations on AF burden or directly answers questions like whether AF burden is the missing piece in our understanding of AF, it sets the stage for intelligent discussions on these matters and subsequently making appropriate plans for future observational studies and trials that hopefully will provide more definitive answers and recommendations.