AF and HF Together: A Vicious Electromechanical Cycle

LaPrincess C Brewer, MD; Bernard J Gersh, MD


May 13, 2015

Editorial Collaboration

Medscape &

This feature requires the newest version of Flash. You can download it here.

Atrial Fibrillation and Heart Failure: What's the Connection?

LaPrincess Brewer, MD: I am Dr LaPrincess Brewer, an advanced cardiology fellow here at Mayo Clinic. During today's Mayo Clinic Talk, we will be discussing atrial fibrillation and heart failure, both of which have emerged as cardiovascular epidemics over the past decade. I am joined by Dr Bernard Gersh, professor of medicine, who specializes in clinical electrophysiology.

Atrial fibrillation and heart failure share common risk factors and frequently coexist. What is the pathophysiology behind atrial fibrillation and heart failure?

Bernard J Gersh, MD: It is a question of cause, consequence, or both. There is no doubt that atrial fibrillation can precipitate heart failure. You lose atrial transport and have rapid ventricular rates. Atrial fibrillation can increase mitral regurgitation, and there is the entity of tachycardia-induced left ventricular dysfunction. Studies in animals have shown that not just a rapid rate but an irregular rate can stimulate the neurohormonal system and cause heart failure. The reverse also applies, and heart failure results in atrial stretch, atrial fibrosis, diastolic dysfunction (which can also increase atrial dysfunction and atrial volume), and atrial volume overload, although a large left atrium leads to electrical inhomogeneity. I can understand why heart failure also causes atrial fibrillation. The other point is that heart failure activates the neurohormonal system and that, in turn, can cause atrial fibrillation. It's both a cause and a consequence, but I think of it as a vicious electromechanical cycle. They both feed upon each other.

Dr Brewer: What is the frequency of atrial fibrillation in patients with heart failure?

Dr Gersh: It depends on the New York Heart Association class. If you look at the trials of people with class 3 and 4 heart failure, the frequency is approximately 40%; and with class 2 to 3 heart failure, it is approximately 15%.[1] In one of our studies, and in a Framingham study over about 5 years, 20% of people with atrial fibrillation had heart failure, and the reverse was true as well.[2]

Dr Brewer: The prognostic significance of atrial fibrillation in patients with heart failure remains controversial. Can you talk about the impact on prognosis for these patients?

Dr Gersh: It is less controversial now. It's not good from either a symptomatic or a prognostic standpoint. In some of the trials and in such trials as the Framingham study,[2] the development of atrial fibrillation independently was associated with a significant increase in subsequent mortality. They may have developed atrial fibrillation because they were deteriorating. Rather than blaming it all on the atrial fibrillation; it's bad news. In someone with compensated left ventricular dysfunction, the development of atrial fibrillation is an adverse prognostic finding. There also is a very interesting group of people with "tachycardia-induced cardiomyopathy" who have no underlying structural heart disease that we know of but who develop atrial fibrillation and heart failure, and they seem to respond (at least from the symptomatic and ejection-fraction standpoint) to correction of the atrial fibrillation. I don't know whether their heart function returns to normal, but that is a group in whom atrial fibrillation is the cause.

Rhythm vs Rate Control

Dr Brewer: Atrial fibrillation and heart failure share common mechanisms. What are the benefits of medical vs interventional treatment?

Dr Gersh: The AF-CHF trial done in Canada[3] that compared rhythm control with rate control did not come up with the answers that I expected. There was no difference in outcomes between rhythm control vs rate control.

It raises several questions. One of the issues in that trial was that a surprising number of people in the rate-control arm actually maintained sinus rhythm. However, this contradicts some of the previous registry studies (which were not necessarily in heart-failure patients) that showed that patients in sinus rhythm did better.[4,5] We don't know whether that is because they were in sinus rhythm. Was the fact that they were in sinus rhythm a favorable prognostic sign? It doesn't mean that trying to maintain sinus rhythm is beneficial.

The AF-CHF trial showed no benefit. I would certainly try to at least restore sinus rhythm in most patients, using cardioversion first and then antiarrhythmic drugs. Management has to be individualized. The antiarrhythmic agents available to us for patients with heart failure are very few. We have amiodarone, sotalol, and dofetilide. If they have coronary disease, you can't use any of the other antiarrhythmic agents. I think sinus rhythm is better. The AF-CHF trial did not show that. The drugs that we have available are not that effective, and that brings us to catheter ablation. Obviously, this is still the subject of an ongoing trial, but there are some data to suggest that if catheter ablation is successful, there is a significant improvement in ejection fraction.[6] There are also data to show that the maintenance of sinus rhythm in people with left ventricular dysfunction is much lower.[7]

There are patients in whom we clearly want to maintain sinus rhythm. They have been going along with compensated heart failure and doing well. They develop atrial fibrillation and deteriorate. There is another group of people in whom it is very hard to maintain rate control. You can't give them calcium-channel blockers because they have left ventricular dysfunction.

Rate control is not always easy. There is a third group with heart failure with preserved ejection fraction in whom atrial fibrillation is very poorly tolerated, so one has to individualize. I would certainly, at least in those patients, try to maintain sinus rhythm.

Counting Down the Heartbeats

Dr Gersh: You asked about the pathophysiology. What is the pathophysiology of tachycardia-induced cardiomyopathy? I have a colleague who believes that man is born with a finite number of heartbeats, and you shouldn't waste them on exercise. This applies to tachycardia-induced cardiomyopathy. We are using up too many heartbeats. It's actually wrong, because if you exercise, your resting heart rate will go down. That will give you more time to use up your heartbeats.

The most plausible mechanistic explanations for tachycardia-induced cardiomyopathy (and perhaps I can extrapolate that not only to tachycardia-induced cardiomyopathy but also to compensated left ventricular dysfunction in patients who deteriorate when they go to atrial fibrillation) is probably abnormal cardiac calcium regulation. There is no question that activation of the renin-angiotensin and the neurohormonal sympathetic nervous systems plays a role in the pathophysiology of left ventricular dysfunction. And then there are changes in the myocytes and extracellular matrix.

That is the pathophysiology, and it argues for attempts to maintain sinus rhythm or rate control. The other aspect of the management of atrial fibrillation and heart failure is the need to stress standard heart-failure medications. Are these patients on optimum doses of beta-blockers? If they have left ventricular dysfunction, they shouldn't be on calcium-channel blockers. Are they on angiotensin-converting-enzyme inhibitors or angiotensin-receptor blockers? Does the patient need an aldosterone antagonist? What about salt in the diet? Are they on anticoagulants? They should be if they are in atrial fibrillation. The other thing to think about is sleep apnea. Are we missing sleep apnea?

Some interesting issues have come up with medical therapy recently, and one of them is a meta-analysis[8] that suggests that beta-blockers are less effective in patients with atrial fibrillation and heart failure. This is not what I would have expected. Some trials favor digoxin[9,10]; some trials are against it.[11,12] Basically the evidence is neutral, but I do use digoxin in some patients for additional rate control. That is one of the few indications for it. I keep the patient's digoxin level on the low side. One problem with atrial fibrillation is in people with implantable cardioverter defibrillators and inappropriate shocks. Data show that inappropriate shocks are not only very uncomfortable for the patient, but they are not good for them. They are associated with adverse mortality.[13]

One other unique aspect of the treatment of atrial fibrillation in people with heart failure is cardiac resynchronization therapy (CRT). The objective of CRT is to pace the ventricles biventricularly all the time. It is always useful when you interrogate the device to see how often the patient is receiving CRT. In patients with atrial fibrillation and more rapid rates, it may be that we are not pacing often enough. Using registry data, a case has been made for atrioventricular nodal ablation to ensure 100% biventricular pacing.[14]

Dr Brewer: That's interesting.

Dr Gersh: It's an interesting condition. It is fairly common. Tachycardia-induced cardiomyopathy is not rare. It's not that common, but it's not rare. By that I mean the patient who has been in perfectly good health goes into atrial fibrillation and then heart failure. There is some evidence that they don't return completely to normal, but they improve.

What is much more common and often overlooked is the patient who is in heart failure and reasonably well-compensated, but then decompensates with atrial fibrillation.

Dr Brewer: Thank you so much, Dr Gersh. This has been a wonderful discussion on atrial fibrillation and heart failure. Thanks to our listeners for tuning into Mayo Clinic Talks for on Medscape Cardiology.


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.
Post as: