The landmark AFFIRM trial showed no difference in outcome between rate and rhythm control for atrial fibrillation (AF or Afib). But does that mean rate control should be the default treatment method?
In this episode of the AP Cardiology podcast, Andrew Perry, MD, discusses this topic and how to approach this decision with Mitch Faddis, MD, PhD, and Phillip Cuculich, MD, both of Washington University School of Medicine in St. Louis.
A transcript of the podcast follows:
Perry: Hi, everyone. Andrew here. Hope you’re enjoying the podcast series so far. I have a great episode today talking atrial fibrillation. Before we get started, I have just a few small requests that I might pose past you. If you are, indeed, enjoying the podcast and find it useful, I hope that you would consider sharing the information or links with your friends and colleagues. Additionally, if it’s not too much trouble, you can go over to iTunes and give me a review and a rating. Those do go a long way.
Second, if there are some topics that you would like to hear or changes, possibly, you think need to be made, there was a survey that I posted a couple of months back where you can find a Google form to post information and some ideas about future episodes and information that you would like to learn more about. Lastly, if you’re on Twitter, you can interact with me there. My handle is @apcardio. I do try to share links to the episodes there and share some interesting content I find related to the episodes that I discuss. With that being said, we’ll launch into today’s episode.
Back in 2002, the AFFIRM trial was published in the New England Journal of Medicine. This was a randomized, controlled trial comparing rate control versus rhythm control in patients with atrial fibrillation. From my perspective, it seemed that this trial showing no difference between those two strategies swung a pendulum in favor of rate control for atrial fibrillation for many, many years. As an internal medicine resident interacting with primarily internists and in internal medicine clinics, rate control tends to be the default method for treating atrial fibrillation.
However, it’s been about over the past year as I have read more about this topic and talked with more cardiologists about atrial fibrillation, I get the sense that the pendulum is swinging back towards a rhythm-control strategy as being a preferred method for some or many patients with atrial fibrillation. I have a lot of questions about this topic, and so I got two experts here at Washington University to discuss them with me. I visited with Dr. Mitch Faddis and Dr. Phil Cuculich, and we all had a roundtable discussion where I presented a case and talked about the points of the case and the management decisions and how many of these trials over the last 5, 10 years have influenced their thinking in the management for atrial fibrillation. They make a pretty strong case for rhythm control. When I’m talking about rhythm control, I’m talking about both antiarrhythmic drugs, but also more particularly in the last few years, [there] has been a stronger emphasis on use of catheter ablation. Without further ado, we’ll get started with today’s episode.
… Thank you both for meeting with me today. I’ve got Dr. Faddis and Dr. Cuculich. We’ll be discussing about atrial fibrillation, and some of the management decisions there. Can I have you both say your name and your title? We’ll start with you, Dr. Faddis.
Faddis: I’m Mitch Faddis. I’m an electrophysiologist here, been on staff since 1999. And I guess pertinent to this discussion I did the first ablation for AF here in 2000. That woman did very well for many, many years, so I have some biases in our discussion today.
Cuculich: Good. Andrew, thanks for the invitation. I’m Phillip Cuculich. I’m an Associate Professor of Medicine here at Washington University and Barnes-Jewish Hospital. Also, a cardiac electrophysiologist.
Perry: Perfect. Thank you both. Without much of a preamble, let’s just launch straight into our case that I have for you. We had a 63-year-old male come to see me in clinic. He has a history of obesity and diabetes. If this was a Curbsiders episode, I might say that his name was Chad Vasc or something like this.
So he comes in. He sees me as a new diagnosis of atrial fibrillation. He reports palpitations that occur a few times a month for the past year, and they could last between 5 and 60 minutes. Recently, they’ve been increasing in frequency, so another physician put him on an event monitor and captured this atrial fibrillation. The first question for you seeing him is do you pursue a rate or a rhythm-control strategy for him?
Faddis: I guess in my mind the conversation that I have with patients always starts with the idea that you have atrial fibrillation because of some other disease process that’s causing progressive fibrosis in your heart. That’s my understanding of the literature behind Afib and its pathogenesis, and that unless we identify and arrest that process, that anything we do to treat atrial fibrillation is likely to fail ultimately. We may get short-term success, but that long term you’re destined to have a recurrence unless we can identify the things that are driving this. The big four are probably present in this patient, although you mentioned just two. Obesity, diabetes, and then the other two that are the common drivers are sleep apnea, which is present in a third to half of all patients who present with Afib and hypertension.
We go on an exhaustive look for sleep apnea, certainly, because I know that treatment of sleep apnea is critical to any long-term success down on the rhythm control arm. Now our bias in looking at this case is that this fellow is really uncomfortable with AF and is going to want something done specifically to prevent him from having episodes. So rhythm control seems to be where he’s destined. I think I’ll let Phil go ahead and give his comments about early takes on this case.
Cuculich: Sure. If I saw this guy in clinic, Andrew, and we had a conversation about this, we used to frame it around the three tenets of Afib management. That is stroke risk reduction, rate control, when appropriate, rhythm control, when appropriate. Now we add that fourth tenet, that fourth pillar, if you will, of lifestyle modification that Mitch just went through very nicely, but that’s clearly an important part of what we do now.
In that conversation, I usually start with stroke risk reduction where we do talk to patients about CHA2DS2-VASc scores and make sure that they understand what their yearly risk of stroke may be if we don’t do something about it, and then what the chances of something bad happening with the medicines that we do use to try to prevent it, namely oral anticoagulants. We usually start with that stroke risk reduction because that’s the one part that we know we’ve made the most progress with and have the most data to support.
Then the decision for rate control and rhythm control, it’s largely driven by symptoms. If you are symptomatic with your atrial fibrillation, it is very much appropriate to restore quality of life with efforts at rhythm control. We’ll go dig deeper into those efforts of rhythm control, but we have about six different medications, a catheter ablation procedure, and a surgery that are at our disposal to try to keep somebody in regular rhythm and maintain most of their life in regular rhythm. That’s where I think this will ultimately go for this gentleman. If I saw him in clinic and he told me he felt his palpitations, they were bothering him, we would start to talk about the rhythm-control strategies there.
Perry: I just want to highlight something that I heard from both of you is that you kind of emphasize more on a quality of life perspective for this gentleman and his symptomatics from atrial fibrillation. I do, based on that, I want to read one section here from the 2014 ACC and AHA guidelines on management for atrial fibrillation. It states that randomized, control trials comparing outcomes of rhythm-control strategy using antiarrhythmic drugs with a rate-control strategy in patients with atrial fibrillation failed to show superiority of rhythm control on mortality for either strategy. Furthermore, when applied in patients who are candidates for both treatment strategies, rhythm or rate control, a rhythm-control strategy resulted in more hospitalizations. While there may not be a difference in mortality versus rate versus rhythm, for this gentleman, maybe we’re thinking more about his symptomatic burden and he would probably be less symptomatic with a rhythm-control strategy.
Cuculich: I want to address that because that was largely driven by a very important randomized trial called AFFIRM. In the AFFIRM trial and really in all of the clinical trials, particularly with atrial fibrillation, you have to look at the population that was included and you have to look at the outcomes that were measured. These are really important, particularly as we start to quote other trials. Because when we flash trials out there, we’re not comparing apples to apples. So let’s talk about AFFIRM for one moment: 2,200 patients, very large study, well-done study, but it was done in older patients who did not feel their Afib. This is important. These were asymptomatic patients, average age was about 75 years old. They were randomized into giving best efforts for rhythm control, and that was generally antiarrhythmic medicines, cardioversions, and efforts to stay in regular rhythm or they were given beta-blockers, calcium channel blockers, and just let be in atrial fibrillation and we would just manage the rate. Indeed, we thought that we knew. In fact, we knew. We had the audacity to know that we were going to find a survival difference, and we were humbled when we didn’t. There was no difference, in terms of, survival between those two strategies.
The upshot of that was positive and was negative in our field. The positive part was that we felt reassured that it was okay that we could keep people alive just as long if we left them in atrial fibrillation. That meant, to me, that we didn’t have to push very hard on the patients who didn’t feel their Afib. It was not going to make them live longer. But what has really been a bit of a blow to our field has been the extrapolation of that study to patients who have symptomatic Afib. On a weekly basis, I see patients who have been managed by their doctors in a rate-controlled strategy who clearly feel their Afib. They’ve been suffering from symptoms of atrial fibrillation for months or years because they thought AFFIRM applied to them. They thought that it doesn’t matter if you do rate control or rhythm control. You’re going to live the same. That is probably true, but you’re going to live with a poor quality of life. That was the downside to this particular study, and I think as you heard from Mitch and from me, symptoms really drive our decisions about trying to restore and maintain regular rhythm.
Faddis: It’s a fun study to talk about because I think it highlights, audacity is a kind way to say it, but I would say the arrogance of electrophysiologists historically to know what’s best for the patient or at least they think they know what’s best. So it was titled AFFIRM because the presumption was we’re going to prevent strokes and people will live longer. It’s clear that AF has got a pretty profound hemodynamic effect. The heart is a two-stage pump. When you lose that preload, there’s a significant impact on your maximum cardiac output, certainly if you’re an active person. You’re going to feel that with your exercise, not to mention that variations in R-R interval, every beat has a different stroke volume and that’s unpleasant and maybe pathophysiology have some impact long term. Certainly, if you’re tachycardic you can develop the cardiomyopathy.
The idea was that pathophysiology has to translate into, “If we’re aggressive with this thing, people are going to do a lot better.” It’s a recapitulation of the same ideas with PVCs. We refused to believe that getting rid of PVCs [premature ventricular contractions] was a bad thing. It sort of cycled back around again because now we’re getting rid of PVCs with catheter ablation, but that’s another talk. The other thing to remember from AFFIRM in my mind is that the trend was towards a higher mortality in the rhythm-control arm. That’s an important thing to remember. We certainly have the capacity to do harm. That trial has a lot of criticisms to be made. For example, in the rhythm-control strategy, if you felt that you had achieved rhythm control, and undoubtedly a lot had, even though we didn’t have the same sort of sophisticated way to look for Afib that was silent, you could be taken off of warfarin and put on just aspirin. 80% of the stokes happened in that group of people in that arm.
I think subsequent data suggests that the connection between Afib and the genesis of the thrombus is tenuous, at best, and that a better pathophysiology to think of is a diseased left atrium has Afib, and a diseased left atrium is thrombogenic and those two things are independent events. They don’t necessarily happen at the same time.
Perry: That might be, just to kind of expound upon that, this diseased left atrium, that may indicate also why after a cardioversion you then promote or why we prefer to put someone on anticoagulation, because we haven’t necessarily gotten rid of that underlying substrate that is thrombogenic within the left atrium.
Cuculich: Very good point.
Perry: You both raised up some very good points that I appreciated, because I think one thing that I notice as a resident spending most of my time in internal medicine clinics and with internists is that frequently, yes, the AFFIRM trial is used in this manner to say, “Somebody here with atrial fibrillation. I can manage them with rate control, and I don’t need to send them to the electrophysiologist to be put on Tikosyn [dofetilide] or sotalol or to get an ablation or whatnot. I can just manage them with my metoprolol or my diltiazem for longer periods of time.” I do want to bring up one other portion from those guidelines, as well, because we mentioned about a symptomatic difference.
The guidelines mention this MANTRA-PAF style: the Medical Antiarrhythmic Treatment or Radiofrequency Ablation in Paroxysmal Atrial Fibrillation. What they state in here, I’ll just read from this. This trial compared AF catheter ablation with antiarrhythmic drug therapy as first-line therapy in 294 patients, so a smaller study. At the 24-month follow-up, more patients of the ablation group were free from any AF or symptomatic AF and quality of life was significantly better. What I found interesting, though, is this following statement where they state, “However, total AF burden was not significantly different between the two groups,” which seems as somewhat of a disconnect in my mind. What are your thoughts about that?
Cuculich: No, let me help you with that. That was an important study it was a Danish study, I believe. Nearly 300 patients and you have to look very closely at the patients that are going into the trial, and you have to look very closely at the outcomes that you’re measuring afterwards. In the exact opposite of AFFIRM, which looked at 75-year-old asymptomatic patients, MANTRA-AF looked at roughly, I think they were 52 years old was the average going into that.
Perry: Much younger.
Cuculich: They were essentially first episode of paroxysmal Afib or dealing with the early stages of Afib in the youngest patients. It’s a different patient population altogether. If I did anything to this patient population, if I gave them flecainide or Tikosyn or did a catheter ablation or gave them gumdrops, chances are atrial fibrillation is not likely to be a big burden for them over the course of the next year. Afib doesn’t progress that quickly. I expect of a 52-year-old, 50-, 55-year-old population, first presentation of Afib, they’re likely to have a low burden. Chances are it’s going to be difficult if that’s the case to demonstrate superiority of one technique over another if the chances of Afib coming back are only 10% in the next year. Maybe in the other arm it’s only 15%. Statistically, it’ll be very difficult to determine a difference. That’s the population.
Then we look at what did they study? This is really important for any arrhythmia study. The thing that they used as their primary endpoint was burden. The groups were then looked and they looked at the average, the mean burden. When you look at a mean burden, it can be skewed dramatically by a few outliers. That is if most of the patients have a burden of 1% or 2%, but you have a handful of patients who are in a 80% or 90% or 100% burden, it skews it dramatically and it’s not a very accurate representation of each patient’s individual effect. I think if you’re going to use burden in any arrhythmia study, it really would be helpful to look at each individual patient. Did their specific patient or did their specific burden come down? That is if I started with a 10% burden and then it became 2%, that was probably a win for me. If I started with 10% and I went to 9%, I don’t know if that was much of a win for me. There’s ways statistically to look at this.
Perry: Like a waterfall plot or something like that.
Cuculich: Yes, that’s a good way to do it. They looked at other ways in this study and they were secondary endpoints that they had pre-specified, and they reported and when you looked at those, it was actually statistically important differences in terms of the number of patients who didn’t have any Afib. If you looked at that, catheter ablation, I think it was a 2-year endpoint, had an 85% chance of not having any more Afib. The medicines have about a 70% chance, and that was a statistical difference. You’re looking at a 15% absolute improvement between 85% and 70%.
They also looked at symptomatic Afib, which was probably the thing that is most important for patients and those numbers remained statistically important. I think it was 92% to 83% or something like that. There are other ways to look at it. I think as we continue to look and I’m sure we’ll end up talking about CABANA during this podcast because you can’t really talk about Afib in 2019 without talking about CABANA. The preponderance of data now supports the idea that if you’re trying to fix Afib, if you’re trying to minimize Afib in your patients, catheter ablation is a better way to do it than a lifetime of medications. Now the question is does that help everybody or does that help hard endpoints? That’s up for some debate.
Faddis: There’s another caveat to that trial. I like that it’s sort of a worst-case analysis of the failings of our treatment, but technically it isn’t what we do in 2019. And what a lot of people were doing at the time of that study, which is to get pulmonary vein isolation, electrical isolation, which, to me, means entrance and exit block out of the pulmonary vein antrum, which is the zone of the left atrium where the pulmonary veins insert. I think definition holds true for most people in the world that do this on a frequent basis. What they did was essentially a technique which was the first pulmonary vein circumferential ablation technique proposed in the early 2000s, which is an abatement of voltage guided by a circular mapping catheter, so they actually said, “We’re going to ablate in the antrum, try to avoid the veins, until all the signals we see are less than 0.2 of a millivolt.” What does that mean? I’m not sure. I think that would be proarrhythmic in some senses because it’s really whether or not the signals can exit that zone and enter into the rest of the left atrium. I would say that they were doomed to failure because of the naivety of the technique that they were using. They also were using, not in a majority, but in some, they used this 8-mm tip catheter, which has a lot of problems with it. I don’t think it represents the technology that was available in 2010 or the technique that was available in 2010, so I thought it was interesting that it was quoted in the guidelines.
Cuculich: This sadly is always going to be the case in a rapidly evolving field. The technology continues to change. Our mapping systems evolve. Our ablations systems evolve. Our knowledge of the disease evolves. By the time we get our hands on some data, a well-run, randomized clinical trial, the very first caveat that we throw at it is, “Wow, that’s not how we’re doing it anymore.”
Faddis: This technology is 5, 7 years old now at this point by the time it’s published.
Perry: Interesting. Good caveats that you both bring up there. Let’s kind of come back to them, back to our case, this 63-year-old male. Dr. Faddis had wanted to make sure that we do our lifestyle modifications. He’s obese. Let’s make sure we start counseling him on some weight loss. He’s diabetic. Let’s make sure his diabetes is under control. You were also particularly concerned, I think, about sleep apnea. He’s coming for a sleep study. I think everybody has a positive sleep study who I’ve ever sent, so he probably needs some CPAP [continuous positive airway pressure], but typical patient I see is that he doesn’t like his CPAP. Can’t find a mask that fits, so he persists in not using his CPAP and continues to have burden from his atrial fibrillation. You both mentioned rhythm control. What might be your initial strategy in pursuing rhythm control for him, medications versus catheter ablation? He’s 63 years old.
Faddis: Sure. I think Phil also started the case with an important point, which is that medically the only thing we’re obligated to do is protect him from stroke, and that guideline has recently changed. The novel oral anticoagulants have been elevated to the first-line treatment. They, before, had been an alternative to warfarin, but now the difference in outcomes from all the large, randomized clinical trials, you just can’t ignore the fact that survival seems to be better with the novel oral anticoagulants. The risk of intracerebral hemorrhage has dropped by 50%. Strangely, at least, and thinking about it, in people who take warfarin, 20% of the strokes that happen while you’re on warfarin are because of intracranial hemorrhage. We cited that. That’s an important positive endpoint that’s caused by a side effect of the drug.
If he doesn’t tolerate CPAP, that’s a huge problem. There are studies that suggest that ablation is not effective or at least compared to non-ablation, it’s not effective in people who aren’t getting treatment. Now that doesn’t generally stop us from hoping against hope that we’ll still get ahead of things, so maybe with weight loss and positional therapy or maybe an oral appliance, maybe there’s some workaround that we can kind of treat his CPAP even though he can’t tolerate the CPAP, and that that’s going to power some positive results for the rhythm-control strategy.
Perry: Let’s, maybe, modify that case, then, with that caveat in place. Let’s say that he can tolerate CPAP because it sounds like that would be a major distraction for you in possibly offering one of your therapies. If that’s not an issue for you, what might you be able to offer?
Faddis: This is where a lot of factors come into play, and I think one of those are tolerance of drugs, tolerance of side effects, perceptions about the Afib and impact on quality of life. Some of that is physical. Some of that is mental. Some people are just so distraught about having Afib they will not rest comfortably unless they know they’ve done everything they can to prevent it, so a lot of this is very personal and involves in-depth conversations of the therapies that are out there, sort of arming the patient with some conception of the spectrum of treatments. It’s fair to say that the treatment is more effective the higher the risk you take or I’ll say it another way. The most-effective treatments have the highest risks of side effects. Telling patients that and getting a sense of their risk tolerance and the impact on the Afib, sometimes they’ll listen to that conversation and say, “Yeah, I don’t want to take a lot of risk here. I’m not that uncomfortable.” That’s probably a patient I’d try a drug, antiarrhythmic drug that is.
Or another patient who’s, say, a very fit person who is a marathoner and their resting heart rate is already bradycardic, so I know I can’t use anything. Tikosyn was not labeled for paroxysmal Afib because it was so ineffective. Now it’s fair to say that paroxysmal Afib, paradoxically are less sensitive to antiarrhythmic drugs, at least in my experience, so we’re already starting out at a disadvantage. That patient might be best served by a catheter ablation, probably would be best served by catheter ablation and I would sort of lobby for that based on those caveats. I’ll let Phil take over.
Cuculich: If I’m seeing this guy in clinic, we have a pretty prescribed opportunity here for treatment, and the things that we talk about are we talk about medicines, catheter ablation, or surgery. We talk about the risks and the benefits of each of those things. For this guy, if he has no other surgical needs, if his valves are working appropriately, he doesn’t have surgical coronary disease, the risk of a maze procedure and the invasiveness are too much compared to what we already have, so that usually falls off the table pretty quickly in an otherwise healthy paroxysmal Afib consultation.
Now we’re left between medicines and catheter ablation. Medicines can be taken every day to prevent the Afib from happening. Alternatively, if your burden of Afib is low, if you only have it one every few months, you can take the pill-in-the-pocket approach. That is you’re not medicating yourself everyday to prevent it, but you’re keeping the pills in your pocket on the days that you have atrial fibrillation, you can take that medication. Studies have shown that decreases the time that you remain in atrial fibrillation. I like it because it’s a way to empower certain patients to be able to help manage their atrial fibrillation, but not have to deal with the longitudinal side effects of every day medicines. That works well if people’s burden is low and if they can manage this. In general, that’s usually once a month or less in my mind is where I make that cut.
Perry: Kind of low.
Cuculich: If somebody is having once every 2 or 3 months, we might start with the pill-in-the-pocket approach, but have the conversations about the others. I agree with everything Mitch said about there’s personal factors that go into this. When you get to the point of Afib happening more frequently and you have to decide between medications and catheter ablation, it’s important to have the conversation that it is really a lifetime of medications that you’re prescribing at that point, too.
If you had a toothache, you’d get a week of amoxicillin and it would clear up the infection. If you have atrial fibrillation, you don’t just get a week of sotalol to clear up the electrical abnormality. You are suppressing that electrical abnormality from here to the finish line. It doesn’t seem to remodel in positive ways most of the time. Now we’re learning more with lifestyle modification, in particular I think, weight loss. Of all of the lifestyle modifications, to me, weight loss is the most important and I do set a very hard and fast goal of a 10% weight loss. That obviously means something to somebody if you’re 400 pounds, if you’re 300 pounds, you can do that math. You can achieve that goal. That is, in the studies, the type of metabolism change that affects improvement in atrial fibrillation. It turns out if you drop 10% of your body weight, you are likely to be on fewer diabetic medications, you’re likely to be on fewer antihypertensive medicines, and there’s some electrophysiologic changes that adapt as well with that.
We talk about the 10% weight drop. We talk about the medicines. We talk about the catheter ablation. For this guy who’s having symptoms for the patient that you gave us, who’s having symptoms many times a month, I would probably start with offering him either daily medication or an opportunity for a catheter ablation, depending on what his feelings were about procedures or about medicines. You’ve seen it, too, Andrew. I mean you’ve seen people who don’t want to take pills. They want to swear off pills. They would much rather have a fix now for something or a perceived fix now for something and others who would never choose to have an elective cardiac ablation. They would rather cycle through medicines and keep the invasive nature of treatment away. It’s really a personal decision. You do have to get to know your patients and understand what makes them tick.
Perry: Sure, okay. I appreciate both of those comments. Part of what kind of spurred my interest in this topic is I feel like, as you mentioned earlier, this AFFIRM trial back in 2002 or around there. It’s similar between rate and rhythm control, but I feel like I’ve been hearing these murmurings of swaying of the pendulum back towards in favor of rhythm control. One thing I wanted to highlight here that I thought was very interesting is this idea that rhythm control can reduce the rates of dementia. There was one study I found. It’s Intermountain group out in Utah where they followed longitude in all these patients who were treated for atrial fibrillation, and it appeared that a rhythm-control strategy reduced the risks of dementia there. What are your thoughts about this point about rhythm control?
Cuculich: I’ve got some strong feelings about this and I would just caution you about the verbs that you use. It hasn’t really been shown that rhythm control reduces that. That’s a statement that you can make if you do a positive, if you do a randomized trial and you’re studying this closely. So let’s couch it with the fact that the Intermountain group did a beautiful study. But it is a retrospective, paired analysis study where they took 4,000-plus patients who had atrial fibrillation who underwent catheter ablation, and they paired them in, I think, it was a 1-to-4 trial, some other way to be able to retrospectively compare them to other patients that are age-matched, gender-matched, as matched as possible as you can in a retrospective study.
They showed an association with, so you can’t really say causal — an association with. Their risks of dementia appeared to be the same as somebody who had never had atrial fibrillation. If you had a catheter ablation, your risks of dementia were the same as if you didn’t have any Afib at all. When you looked at patients who had Afib that didn’t get catheter ablation, their risks of dementia were considerably higher. There is inherently a bias in this. We are inherently taking some of the healthiest patients for catheter ablation, and we are inherently not offering catheter ablation to the sickest patients. Retrospective studies can never overcome those biases. They tried as hard as they could, but you can never overcome those biases.
I think it’s thought provoking. I think it’s hypothesis generating. I think it may be true, but the only way to really understand the cause and effect of it is to put it out on the line and do it in a prospective way. That is, follow dementia markers in a randomized trial of patients that you intervened on in some way and a control arm that you didn’t. That’s a much harder trial to execute, but that would really give you the definitive cause and effect.
Faddis: It’s an important observation because it really is the elephant in the room. It’s one of the elephants in the room, so the one thing that we haven’t mentioned is the oft-repeated observation that if you have Afib your mortality risk is higher. It’s twice as high. Probably the best study looking at that is from Framingham, the longitudinal study of a community of people looking at mortality risk as a function of different disease markers and trying to capture the natural history of cardiovascular disease. What they saw is a two-fold mortality increase. That’s been seen in other trials as well.
Now the question is is there a cause and effect here? If we treat the Afib, can we make that mortality risk go away? Most electrophysiologists feel there may be a cause and effect. I’m sort of more skeptical. I think that it’s likely that the sickest people have Afib and they’re getting a higher mortality risk because of those other pathophysiology pieces. Afib is a marker of that, but it’s not really the cause and effect. That’s the other side of the possibility. The same thing is true with stroke. Is Afib the thing that we need to shoot at for stroke prevention? AFFIRM said no. I’m skeptical. And that Intermountain Health Registry study, they didn’t take a careful look at anticoagulation status. That’s the obvious thing. Maybe this is the difference.
Phil brings out the point about selection bias. There’s no doubt that played a role as well. You wouldn’t take a demented person to a catheter ablation, so I think maybe early in dementia you just sort of can tell from your interview that that would not be a great candidate. That would bias your outcome. But whether or not patients need to continue to be on anticoagulation after a successful ablation, our party line and our understanding of the data is that yes, we need to pay attention to CHA2DS2-VASc. You’ll notice that CHA2DS2-VASc says nothing about paroxysmal or persistent or efficacy of rhythm control. That’s not in the risk. I think that speaks to the fact that it’s the substrate. The substrate is thrombogenic and that’s the problem there.
Perry: I’ve wondered about that and when you were talking about the CHA2DS2-VASc score in that there’s nothing to input about the burden of atrial fibrillation that you’re in. Is your risk different if you’re always in atrial fibrillation versus if you’re going in once a month, per se? You just have Afib and then you get this score. It’s an interesting point that you bring up from there.
Circling back to our case, so certainly when he comes and sees you, you order an echocardiogram or he gets an echocardiogram at some point. He is found to have a reduced left ventricular ejection fraction. Let’s say less than 35%, right around 35%. There have been a couple of important trials that we’ve alluded to to this point in the discussion. I want to bring up now, in this case, in this setting forum. We’ll say that for our gentleman he has a non-ischemic cardiomyopathy. There were two studies, the CABANA trial and then more recently the CASTLE-AF trial, both having somewhat of different responses or maybe different answers as to maybe what the better option is or what the “right” management decision for him would be. What would be your thoughts about these two trials? Maybe first we should explain what these two trials are.
Cuculich: Yeah, let’s just think about the patient first. Somebody shows up with new atrial fibrillation and they have a reduced ejection fraction, for no other purposes. You have to think that atrial fibrillation could be driving that cardiomyopathy. That can happen through a number of mechanisms. It could just be a rate-related component. This could be a tachycardia-mediated cardiomyopathy. We’ve seen ejection fractions improve with just slowing down the heart rate even in atrial fibrillation. Taking somebody from 140 to 90 as an average oftentimes improves the cardiomyopathy. But beyond that, there actually has been now a number of studies that have shown restoration of sinus rhythm seems to also improve the cardiomyopathy, improve the ejection fraction, and, interestingly in some of the imaging studies, restoration of sinus rhythm also has an effect on fibrosis in the ventricle.
Looking backwards at that, you might say, “The more atrial fibrillation you have, the more likely you are to have fibrosis laying down in the ventricle,” which is a bad thing. There’s this push of data now that say not only if we think about the mechanisms of this, it could be rate-related, but it could actually be rhythm-related as well. If somebody from a practical standpoint, you see somebody with new onset atrial fibrillation and a reduced ejection fraction, that gives me even more power, more enthusiasm to restore and maintain regular rhythm and see what the effect is on their cardiomyopathy. We might fix a lot of problems with maintaining regular rhythm in this case.
In some ways, this is reminiscent of the CASTLE-AF study that you alluded to here. This is a relatively new study published in the New England Journal of Medicine. And it was a group of patients who have reduced ejection fractions, had defibrillators in. I think they were all for primary prevention, and it was a mix. You’re right. It was a mix of primary and secondary prevention. Half of those patients were given a rate-control strategy or a medication-control strategy for their Afib. That was, they could take antiarrhythmics or they could take beta-blockers, calcium channel blockers, but it was called a medication control. Then the other arm got catheter ablation. Catheter ablation oftentimes meant that you were also off your antiarrhythmics afterwards. It’s important to think that it’s not just the intervention. It’s the intervention plus the medical treatment that goes on over the course of time. It looked at hard endpoints, which I think is nice. That’s the question that we’re asking now in our field is, “Do we see hard endpoint changes? Do we see hospitalizations or deaths or stroke change with restoration and maintenance of regular rhythm?” In this study, it was pretty profound that the combined endpoint of death and cardiovascular hospitalizations was, indeed, reduced if you underwent a catheter ablation and got off your antiarrhythmics versus staying on medical therapy over the course of those. That became different over the course of about 2 to 2 1/2 years.
It was a very profound effect. It was unlikely to be a statistical variant. I’d like to think that if we saw a similar sort of effect with a heart failure medication it would be a big deal in the heart failure world — that we’re reducing cardiovascular hospitalizations, we’re reducing death. It’s would be a prescribed medicine in anywhere else in cardiology. Yet, I think there’s been some pushback in our field about how that trial was performed and how far we can extrapolate. And similar to what I have always said, you have to look at the patients going in and the endpoints you’re measuring on the out. The patient is going in in this situation were patients who had left atria that weren’t too dilated. I think it was a 6-cm difference I think is what they said, so anything under 6 cm, and it was a mix of paroxysmal and persistent. It was heart failures that avoided the end-stage class 4’s as well.
It was a fairly prescribed group of patients, but for those patients, it was an important endpoint to see a difference. In my mind, this applies very much to where we would sit with this patient. He doesn’t yet have a defibrillator. I would like to restore sinus rhythm to avoid the need to offer him one and maybe improve a lot of things: his risk for sudden cardiac arrest, his risk for cardiomyopathy, his risk of heart failure, but I think it’s reasonable to extrapolate that CASTLE-AF study into where we are here. Mitch, your thoughts about CASTLE-AF?
Faddis: Yeah, I’ve regarded it as a very strange trial since it was published. A dramatic outcome. I think it’s telling that the committee that wrote the guidelines used that trial, considered it a gold standard design, randomized-control trial, not sham controlled, which would have been even better. But a randomized-control trial nonetheless, and now include that as a IIb indication for ablation. Why did they do that? I think if you look carefully they actually published the subgroup analysis. It’s a tiny trial in comparison to other trials. The trial, to get into it, you had to either fail a drug, antiarrhythmic drug, or you had to say, “I don’t want to take an antiarrhythmic drug.” I think a number got in. They didn’t break that out in the original publication, so whether or not it’s in the supplementary data, I don’t know.
But in the subgroup analysis, if you’ll look, you’ll see that a third of the patients in both arms were on amiodarone at the end of the trial, and that in that group, there was no difference in outcome. Primary outcome was mortality and heart failure hospitalization. Now you’re cutting it pretty thin when you take a third of an already small group. The other thing is that all of the benefits seemed to be focused in patients who were younger than 65 and that had an EF that was over 25% and that had class II New York Heart Association class heart failure. The healthiest people are where the benefit is. Those are important caveats.
Then the other thing that sort of sticks in my mind is why is this trial so dramatically positive compared to, for example, AF-CHF, a much bigger trial done with largely amiodarone in heart failure patients, not placebo controlled, but controlled against rhythm control? That trial did not show any sort of benefit in mortality or EF. The other thing about that trial is that 70% efficacy for amiodarone to maintain sinus rhythm. Now they gauged that by intermittent ECGs, so you could say, “Well, they’re missing all that silent AF.” That’s a possibility. But I think there were a lot of successes.
Is it that burning the atrium is giving you a mortality advantage? I can’t buy that. I don’t think that’s it at all. I think there’s something else here. One thing that occurred to me is that we know that shocks, all these people in CASTLE-AF had ICDs. We know that shocks damage the heart and increase mortality. We have good data for that. Maybe it’s that the people who didn’t get catheter ablation had more ICD shocks and that led to the mortality signal that we’re seeing here. We don’t see anything about the veracity of the rhythm-control strategy or rate-control strategy. I think there are a lot of questions. There are other signals from other tiny trials that suggest that rhythm control does have a mortality advantage. Again, there’s this question, “Is this selection bias? Are we really just looking at the healthiest people?” Similarly, as Phil pointed out, there’s also a signal consistent in small trials that EF improves with rhythm-control strategies.
Still questions out there. I don’t know. I don’t think that CASTLE-AF is the final say on this trial. We didn’t talk about CABANA. We had hoped that CABANA would be a final say on this. Very large trial as opposed to CASTLE-AF. How many, 2,000?
Faddis: 2,200 randomized. A multinational study. The combined endpoint is the trouble. It included major bleeding events, cardiac arrest, stroke.
Perry: And death. I think it was those four.
Faddis: Those four. Patients were randomized to catheter ablation or medical therapy, not specified rhythm versus rate control. It was not significantly different. Now there was a huge crossover rate. I think something like 30% crossover from medical to ablation, which diminishes the legacy. We really wanted to know whether or not mortality or stroke were affected by catheter ablation. At least from my perspective, whether we’re doing harm is a significant question here. But it was so diluted. A majority of the major enrollers for the center were Russia. That doesn’t reflect my practice, probably. I don’t know that they have the same patients or access to technology or strategies for following the patients. I’m disappointed that CABANA wasn’t more definitive. You can say, “The suggestion that intention-to-treat was positive. Yes, it does look like intention-to-treat was positive in CABANA.” But again, that’s now hypothesis generating. It wasn’t definitive because we’re looking at subgroups here. Phil, your comments?
Cuculich: I’ll just do a point/counterpoint a little bit. I am higher on CASTLE-AF and I am lower on CABANA than I think the way that you have interpreted it. From CASTLE-AF, I was surprised that the guideline directors made it a IIb indication because it is a randomized trial that was carefully done. It’s also not just in isolation. There are other, as you mentioned, smaller trials. The AATAC-AF study that showed the exact same type of survival benefit and hospitalization benefit, so it’s not entirely swimming upstream. There’s now three studies, two small randomized and one prospectively randomized larger, which is what CASTLE-AF is, so I’d give it a little bit more credit that it is a randomized trial.
I think the hard part about it is that it does swim upstream from the way that we think about heart failure patients, and it goes back to what Mitch had said about is Afib causing the problems or is it a secondary bystander to an already sick heart? We’ve seen a lot of patients who have advanced heart failure who also have Afib. Whether we restore regular rhythm or not, it doesn’t seem to do anything to their more advanced heart failure stats. It doesn’t seem to change their cardiomyopathy, the number of times they’re frequently coming into the hospital. I think our bias is that we’re used to seeing a picture of a AF/HF patient, in my mind, is somebody who has really advanced heart failure. I have a hard time thinking that we’re going to fix that patient. There is a concern in our field that there’s going to be an overextrapolation of the data from CASTLE-AF to every patient with heart failure and AF, and that would not be appropriate use of that data.
Perry: It could be that these patients are too far end-stage in their heart failure. Their substrate is so far remodeled that there’s…
Cuculich: Correct. I think if you pooled a lot of cardiologists, a lot of electrophysiologists, I think that would be the general take on why CASTLE-AF hasn’t gained more traction. There’s fear that this is going to extrapolate to every patient with heart failure. Mitch points out nicely, they’re in the subgroup. You can tell who those patients and let’s think about this. This is death and cardiovascular heart failure or cardiovascular admissions. This is a big deal. Those are really impactful endpoints. We owe it to our patients to know how we can prevent those things. In some ways, CASTLE-AF and CABANA completely align on this. That is, when you look at the subgroup analysis of CABANA, the groups that do have the differences in that combined mortality of the death, stroke, cardiac arrest, and severe bleeding. When you look at that, it, indeed, is the younger patients as well, the youngest tertile in that group as well.
It’s likely that you have to do something early enough in the Afib experience to alter the later effects of death and heart failure and hospitalization. It’s likely that there’s an impact. You have to live long enough to achieve the benefit. And that if you had an 85-year-old person with advanced heart failure, it doesn’t seem intuitive that if we did a catheter ablation on that person we’re going to make them live longer or stay out of the hospital. I think, as I mentioned before, I think the population that you’re studying matters most. I think what you’re studying afterwards matters. I think that’s how I put together those two.
Perry: Great. I think we’ve had a rich discussion about this case and a number of different trials that have influenced both of your thinking and the field’s thinking in regards to rate and rhythm control, and then the use of catheter ablation. Maybe just a couple thoughts as a summary for where you stand maybe on this topic. We’ll start with Dr. Faddis.
Faddis: I might have come off as being against catheter ablation, but that’s certainly the majority of my practice is catheter ablation of atrial fibrillation. So I think we’re on our firmest footing when we say that catheter ablation is the most definitive treatment for symptomatic Afib, short of maze surgery, and that the patients who are suffering from very poor quality of life related to atrial fibrillation are best served by catheter ablation. That’s where I’m the firmest footing. Lots of uncertainty about whether or not rhythm control changes the natural history of cardiomyopathy and affects mortality, so I think we’re still in the dark, to some degree, in that area.
Cuculich: I would just sum it up by saying in 2019 we have more treatment options now than we ever have had in the history of mankind to treat atrial fibrillation. We have more medicines, more catheter ablations, different technologies to treat atrial fibrillation. It turns out that they’re getting more and more effective so that catheter ablation, in general, is our sweet spot for risk and efficacy when we talk about trying to control atrial fibrillation, and that’s largely driven by symptoms. I would say here in 2019 you should not let a symptomatic patient with atrial fibrillation suffer. We have opportunities to help expand and treat their quality of life.
There are many places that we’ve got to push further, though. We need to understand the mechanisms of atrial fibrillation in order to be able to tailor treatment to each patient. We need to understand the link between atrial fibrillation and stroke, and I think Mitch has a very good way of thinking about that with the diseased left atrium. I think that we’re going to be able to continue pushing this forward by understanding mechanisms, by better phenotyping our patients, realizing that one size does not fit all in atrial fibrillation. It really requires a tailored treatment program for each individual patient.
Perry: Great. Those are great summaries from both of you. Lastly, thank you, again, for your time.
Faddis: Thank you.
Cuculich: Thank you, Andrew. Appreciate it.
Perry: I’ll now summarize a few of the main points from our discussion. First, one of your primary goals in treating a patient with atrial fibrillation should be stroke risk reduction, and that’s using anticoagulation. One pearl that I learned from Dr. Faddis is about how the stroke risk appears to come from a diseased left atrium, and that is the thrombogenic portion, and not so much of a contribution from the irregular rhythm.
Next, when you’re considering a patient, when treating them for atrial fibrillation, probably more emphasis needs to be placed on managing their comorbidities. Particular emphasis was placed on obesity and obstructive sleep apnea. Dr. Cuculich advised that he strives for a 10% weight loss reduction in his patients, and they also focused on good control and good treatment of obstructive sleep apnea with CPAP.
Now when also considering treatments between rate and rhythm control, one point that often gets missed is that patients who were included in the AFFIRM trial were asymptomatic. There may be patients in your clinic or in your practice who are not asymptomatic from their atrial fibrillation, yet are being treated with a rate-control strategy. Next, both Dr. Cuculich and Dr. Faddis emphasized a shared decision model with their patients. For example, when discussing rhythm control for their patients, there are options between antiarrhythmic medications and catheter ablation. And really it comes down to a discussion between the patient’s desires and preferences and their hopes and their fears to help guide you in which would be the most appropriate management decision there.
Then lastly, it appears that the most effective treatment for atrial fibrillation in restoring sinus rhythm is via catheter ablation. This is a rapidly evolving field with new technologies and new techniques expanding every year. We are certainly only going to learn more about who the better candidates are, the underlying mechanisms, and the processes by which catheter ablation is most effective. Thanks for listening and we’ll see you next time.
Andrew Perry, MD, is a resident physician at Barnes-Jewish Hospital and Washington University School of Medicine in St. Louis.