Jun 05 2026 This Week in Cardiology

The first topic today is chemical cardioversion of AFib in the emergency department, the FLECA ED trial. Indeed, I missed an important trial from the European Heart Rhythm meeting in Paris this past April. Fleca ED was presented as part of the late breaking clinical trials. In six emergency departments in Athens, ninety-three patients with new onset AFib considered for cardioversion.

were randomized to IV fleconide or IV amyoderone. Notably, the patients had to have documented coronary artery disease. but no residual ischemia and an ejection fraction greater than thirty five percent. This is certainly novel because phlecanine has had a black box warning against use in patients with structural heart disease. especially cornerated. This warning stems from the nineteen nineties era CAST trial, CAST.

where post MI patients with ventricular arrhythmias were randomized to one C antirhrythmic drugs or placebo. And almost all of you know the results of CAS, despite the pretrial ethical controversy about randomizing such patients to placebo. The rate of the primary efficacy endpoint of death or cardiac arrest was three point five percent in absolute terms higher in the antirhythmic drug arm. The number needed to kill was twenty nine in cash.

Now, while I normally call CAST one of the most important trials, not just in cardiology, but in modern medicine, it is also true that it has been widely over-extrapolated to patients unlike those in the trial. CAST patients had recent MI, depressed LV function, incomplete revascarization. In fact, more than a third of the patients had no revascarization. Most patients were enrolled in the thirty days post MI. Approximately eighty percent had abnormal ejection fractions, and not only that.

The target arrhythmia in cast was ventricular ectopy, not AFib. The Flecca ED trial challenges the broad ban on fleconide use in patients with CAD. The primary efficacy endpoint was successful cardioversion and sinus rhythm. The safety endpoint was any VT or bradycardia or low blood pressure at six hours.

About seven hundred and fifty patients were screened for eligibility, and exclusion of patients occurred for many reasons, such as unstable hemodynamics, low EF, no coronary disease history, ACS changes on the ECG positive chiponins or left bundle branch block. In the end. The trial has found 107 patients who had no angina but did have coronary disease or a PCI or cabbage within a year and had a recent negative stress imaging.

A few more patients were excluded due to residual ischemia and ninety-three were randomized to IV fleconide or amiotorone. Now IV fleconide is not available in the US. I've not used it, but the dose is two milligrams per kilogram up to one hundred fifty milligrams max over ten minutes and amiodorone is the normal five milligrams per kilogram over one hour. Now these patients were seventy-five years old, mostly male patients who had normal EF on average.

Their Chad's vascular was three to four, and notably about half had had prior MI, and nearly one in four patients had known obstructive CAD without prior revascarization, but they did have negative stress imaging. The AF duration was short at a mean of four hours, more than half were on anticoagulation. Now note that the inclusion criteria was for AF less than 48 hours if not on anticoagulation, but less than seven days if on anticoagulation. And these are typical uh restrictions that we use.

The results were clear, clear. 36 of 42 patients or 86% in the phleconide group converted to sinus rhythm as six hours. versus 18 of 45 or 40 percent in the amiodorone group. That's an odds ratio of nine and a highly significant p value. Safety, three patients had brachardia and these were in both arms. There was no VT. An electrical cardioversion was done in twelve percent versus thirty six percent respectively.

So my comments, I like this effort. Drugs like amioderone and diltiazam for that matter get entrenched into a therapeutic fashion. Uh it's based on tradition or preference or just simple inertia. So it's good to do such studies. Cardioversion is an easy thing to measure. You go to sinus rhythm or you don't. Harder to measure in a small study like this is safety.

The authors say that the study was powered for efficacy, but it was not powered for safety. And that is true, of course, so we should not extrapolate to anything safety-wise. For instance, much was made in their presentation at the European Heart Rhythm about how these results run counter to the cast trial. But I I don't think that is exactly correct, is it? This study measured the effects of one IV dose of phleconide with a half life certainly less than twelve hours.

The CAS trial measured oral phleconide use over days and weeks. Fleconide proarrhythmia is surely not acute. You'd not expect it after a single dose of IV medicine and a totally stable, well perfused heart. That said, I also think CAST has been widely overtranslated and the blocks black box warning prohibiting use in patients with any amount of structural heart disease is really a blemish on intelligent evidence-based practice.

I mean perhaps CAS was such a shocker to people at the time that it caused this overreaction. Oh, I wish we had something this good for chemical cardioversion in the US because I like the idea of giving a painless 10-minute injection to restore sinus rhythm. It's nice. As well as the fact that phleconide is around for hours after the cardioversion, which would of course help prevent ERAF, capital E R A F or early recurrence of AFib after a shot.

And let it be known that I hate shocking people. It's an ugly procedure, it's aggressive, and it's inelegant. Which brings me to the matter of doing cardioversion in the emergency department in the first place. Uh this breaks one of my number one rules of AF management, and that is to give peace a chance.

Now, if you leave acute onset AFib alone, it often corrects itself. One of my favorite trials is the RACE 7 ACWAS trial published in New England Journal of Medicine in 2019, and I'll link to it. In that Dutch trial, 400 some patients with recent onset AFib in the E E D, one group gets early cardioversion and the other group gets delayed cardioversion. The wait and see approach meant giving rate control meds and delayed cardioversion if the AFib did not resolve within forty eight hours.

The primary endpoint of sinus rhythm at four weeks occurred in ninety one percent of the delayed cardioversion and ninety-four percent in the early cardioversion group, and this was the non inferiority analysis and clearly that met non inferiority. The best though was that sixty nine percent of patients

in the delayed cardioversion group spontaneously converted to sinus rhythm in forty eight hours. These one hundred and fifty of two hundred eighteen patients avoided a rash on their skin where the pads go, they avoided IV sedation, and they simply waited. And to me, this is elegant. Even in the early cardioversion group, sixteen percent, almost one in five patients converted before they can even set up the cardioversion group.

Now, the other reason I don't like cardioverting stable patients in the emergency department is that it creates the wrong frame for atrial fibrillation. Shocking a patient creates the AFib scary frame, as in oh crap, AFib must be seriously dangerous if a doctor puts me asleep with an anesthetic and then shocks my heart. This is the opposite of what I tried to achieve in the AFib clinic.

The majority of patients with AFib are stable, but but they are fearful. So I spend a great deal of time reassuring patients. I say we will find a solution, but there is no immediate danger. There is time to work on things. Things like diet, exercise, lifestyle changes. Creating this calming frame is often harder because patients are often told they need to see a cardiologist immediately. Less experienced clinicians often scare the patient with nuance at AFib.

But shocking someone in the ED for a stable rhythm totally creates the Aphibis scary frame. So I'm against it in the vast majority of cases.

No, we can't say much about safety due to the small numbers in the study, but there really wasn't a Bayesian or prior expectation that there would be a safety issue. But please, my friends, help me spread the message for AFib. Unless this patient has hemodynamic compromise, please give peace a chance.

All right, next topic is heart failure monitoring, the alleviate HF trial. Of course, one of the many goals of heart failure therapy is prevention of hospitalizations, which are usually due to volume overload. Readmission rates with heart failure are high and no one likes it and the costs are high. Now, before there were fancy devices, doctors and patients had to use clinical clues. Symptoms such as dysmia orthopnea edema, signs like increase in weight, abdominal girth or edema.

These are old school and rare was the patient and doctor team that could head off an admission by boosting diuretics simply by using these old school measures. So technology has come, right? First in the form of a little paperclip like device that goes in the pulmonary artery. This is called the Cardio MEMS device, which was supposed to be able to provide real-time wireless feedback on PA pressures and how could you miss with that?

and this would allow adjustment of meds, but the data for that is extremely dubious, and I will review it briefly later because it's instructive. Now another way to monitor heart failure and potentially influence outcomes. is with an insertable cardiac monitor that is equipped with multiple sensors that could then trigger interventions such as a temporary increase in diuretics.

Well last week Jack published the Alleviate HF trial, which tested an algorithm from a Medtronic branded implantable cardiac monitor for efficacy and safety in patients with hephra. Now this is a nifty device, right? It collects data continuously, including ECG stuff like AFib, ventricular rate, heart rate variability. It records accelerometer data for activity, bioimpedance measures for fluid status.

Then it analyzes these data points using a proprietary algorithm to generate a daily heart failure risk status classified as low, medium, or high risk for worsening heart failure within the short term. The trial was conducted in fifty-two sites in the US. Patients in both groups had the device implanted. Active arm patients used the algorithm. The control arm had basic care. Now these trial procedures were quite specific.

When the device made note of a high risk alert, centralized nurses evaluated the patient before any treatment was initiated. If no rule out conditions were identified, the patient self-administered an individualized four-day increased diuretic prescription.

Now nurses could withhold the intervention if any of the rule out criteria occurred, and these were things like A new atrial fibrillation, recent C O P D treatment, febrile disease, recent heart fire medication or diuretic changes, low blood pressure. A lot of these things. These rule out criteria blocked intervention in more than half, 53% of high-risk notifications. And only about a third of first high risk onset uh notifications receive treatment within the first two weeks.

The primary endpoint was a five component hierarchical composite analyzed by win ratio. And of course, win ratios analyze these things in order of clinical priority. First CV death, then hospitalization for heart failure, then outpatient heart failure event requiring invasive therapy, then the KCCQ, and then a six-minute walk test.

The primary efficacy endpoint was clearly null in this trial. The hierarchical composite win ratio was 0.79 with conference intervals going from 0.62 to 1.01. That p-value was 0.06. But we need to pause here because you might think the zero point seven nine hazard ratio is in favor of the intervention arm. But since this is a win ratio, this there were substantially fewer wins, twenty-one percent fewer wins in the

implantable cardiac monitor arm. It just missed significance for the control arm being better. So the authors were very careful about not emphasizing this strong trend favoring the control arm. Instead, they focused on the fact that the KCCU quality of life scores drove nearly half of all pairwise comparisons. And the intervention arm started with a higher baseline KCCQ scores, an imbalance that hurt them on change from baseline anal analysis.

Also, though, I would have to say cumulative CV death and heart failure events were numerically higher in the intervention arm, hazard ratio 1.43, though this was not statistically significant. When the analysis was corrected for the baseline KCCQ imbalances, the win ratio became essentially neutral at a ha at a ratio of one point zero two.

They'll say that for an alert tool to affect outcomes, it has to cause treatment. And in this trial, more than half the algorithm detected high-risk notifications were, quote, ruled out. They call this an execution gap and that the rule out framework was overly cautious. Quote Future studies should stop treating the alert as the intervention and instead define the alert to action pathway as the true intervention.

For instance, in heart failure care, the barrier is often not knowledge, but workflow. And indeed, I agree with that because there's too many clicks, too much ambiguity, and too little ownership, and all that leads to inertia for treatment of these patients. But given that, I take a more pessimistic view of technology's ability to help us treat patients with heart failure. My take isn't that there is a problem with the device or with action, rather it's with the concept.

that technology driven intervention would ever work. This idea, I think, assumes too much Now, while there are some similarities between heart failure patients, there's also a lot of differences, such that an algorithm, no matter how sophisticated, is unlikely to work.

And I have evidence for this negativity. there are trials and technology hasn't turned out very well. And it starts with the cardiomems device. In two thousand eleven The Lancet published the Champion trial, not the Champion AF trial, just called the Champion Trial, and this was of PA monitor versus standard care in five hundred and fifty patients with class three heart failure.

Now the top line result of that trial was a statistically significant twenty-eight percent reduction in hospitalizations for heart failure, which was the primary endpoint, and that sounds good, doesn't it? But The FDA advisory committee rejected the claim of efficacy. On two occasions it did. Why? Well, I couldn't remember and I had to do some digging. I had forgot. Now, former hard.org um journalist Steve Stiles has some great reporting on the issues, and I'll link to it.

It turned out that an audit by FDA inspectors found that the sponsor or principal investigators routinely contacted investigational sites and made specific therapeutic recommendations for some treatment study subjects. Specifically, nurses employed by Cardiomem's group made therapy recommendations on 193 occasions for treatment group patients. None were permitted or occurred for control group patients.

So in short, the treatment arm received not just the device, but also intensive sponsor employee nursing support, a potentially confounded intervention. For sure a potentially confounded intervention. Of course, another problem with this trial was that while patients were blinded, physicians were not. And of of course you know that most MDs that are doing this study are likely proponents and

Hospitalization for heart failure is a subjective endpoint, meaning a doctor has to decide. So it's no wonder that there were fewer h hospitalizations for heart failure in a treatment arm. A decade later, the Cardiomem's team was back with publication of the Guide HF trial, wherein a thousand patients were randomized to device based management versus standard care.

And in Guide HF, it's in the Lancet. I'll link to it. There was a non-significant 12% reduction in the primary endpoint of all-cause death. Hospitalization for heart failure and urgent heart failure visit. The hazard ratio was 0.88. Conference interval 0.74 to 1.05. So now maybe you're confused. You're like, John, uh wait a minute, we use cardiomems. How did it get approved with the negative thousand patient study?

Now, old listeners to the podcast r may remember, but the way the company and the K opinion leaders salvaged this device. And when they sliced up the data in that way, Pre COVID and post COVID, the pre COVID hazard ratio was 0.81 with listen to these conference intervals and p value. Zero point six six to one point zero zero p value, zero point zero four nine. So boom.

The monitoring device was improved was approved based on this uh slicing up uh of the subgroups. But I think this is one of the greatest examples of post hoc spin I have ever seen in any trial. Now looking at the cardiomemps data fairly, there is little doubt declaring that this device is unhelpful. So putting it together from a Bayesian view where we consider prior to the

If an invasive wireless monitor in the PA was not effective, you would have expected the ICM with the proprietary algorithm to be negative as well. And it was strongly so. We can keep trying to study this technology in heart failure, but I remain pessimistic that it will ever be better in simple education of the patients and close follow-up with clinicians who know the person. All right.

Third topic today, a weight loss study in atrial fibrillation. JAM has published the Lose AF trial of weight loss in older patients with persistent AFib. This was an unblinded RCT done at two UK hospitals from 2018 to 2025. Nearly fifteen hundred patients, one thousand five hundred.

who had cardioversion were screened for eligibility. And about five hundred were aged sixty to eighty five years, had a BMI greater than twenty-seven, and they were invited. And three hundred of those five hundred said no thanks. In the end, about 118 patients were randomized to either an eight month low calorie diet with behavioral support versus usual care.

In the intervention group, each person was assigned a local counselor who provided regular, individually tailored appointments over the eight month period. Patients in this arm were also supplied formula meal products as well. The trial was pragmatic, so all AF decisions were left up to the treating doctors, including AF ablation or cardioversion. The primary outcome was AF symptom score at eight months, assessed using a standard AFSS instrument. There were also many secondary outcomes.

Now these highly, highly, highly selected patients were sixty-eight years old. About a third female average BMI, thirty-four, most had hypertension, and importantly about fifty percent of patients had longstanding persistina fib. The intervention arm did produce weight loss. The intervention group lost on average 9.7% of their body weight, which was statistically significantly more than the 3.1% of weight loss and control participants.

So the delta in kilograms was minus 6.9. But despite these differences in weight, no significant between group differences were observed in the AFSS symptom severity score at eight months. There were no other endpoints differed. Health care, related quality of life, freedom from AFib, any CMR measured parameter of atrial or ventricular function, Uh was unchanged in the two arms. There were also no differences in blood pressure, LDL, triglycerides, CRPs, nothing, NATO.

At long-term follow-up of 3.5 years, the delta in weight was decreased to 5 kilograms, but again, there were no differences in any AF symptoms. Now, uh my comments. I think this is also an important effort on the part of the John Radcliffe and Milton Keynes hospital group.

Guidelines have weight loss as a class 1A recommendation, but honestly, there is only one RCT. It's from many years ago. Most of the evidence underpinning weight loss for AF treatment stems from one group, the Adelaide group, and it's observational. It's compelling observational work, especially since it's coupled with corroborating biochemical, physic physiologic, and anatomic work.

So in other words, yes, the group of patients in Adelaide who can lose ten percent of their body weight may do other things that reduce A fifth. But these patients also show lower blood pressure, better glycemic indexes, and favorable cardiac remodeling. So it makes sense to consider the weight loss causal and not just correlational. But technically, technically, it's still observational and weight loss programs are neither easy nor inexpensive.

In this trial, for instance, each patient had their own counselor, and counselors don't come free of charge in the real world. So these results were interesting. The program was mildly successful in causing weight loss, a delta of seven kilograms or fifteen pounds. That isn't bad. Yet that amount of weight loss didn't move any metric, none. Aphib symptoms persisted, and no important measure of health were improved.

Now, why was this data so negative? I can think of three reasons, maybe you can think of more. First, it's not enough weight loss, and eight months isn't long enough. Well 9.7% body weight loss is good. It's not great. A 250-pound person who loses 10% is still 225 pounds. Now, 225 pounds is kind of normal on a US scale, but you don't see many 225-year-old Western Europeans running around Belgium or France or Denmark. For them, 225 pounds is still quite large.

Second, to move AF measures, you need more than weight loss, right? You need to address sleep apnea, alcohol intake, and cardiorespiratory fitness. And if you believe the observational data are causal, then all of these are additive to the weight loss. And third, close to a majority of these patients had the worst kind of AFib, long-standing persistent AFib, and this might be the toughest AFib to change with any intervention, weight loss included.

But despite this negative study, I do think there is hope for weight loss as an AF intervention, though, and it comes in the form of the GLP1 agonism. These drugs are amazing. Not only do they induce weight loss, I've seen them improve every metabolic measure, such as. Triglycerides, blood pressure, glucose, even sleep apnea. And if you improve all this, surely AFib will improve.

And finally, I really reiterate my call to my colleagues, my friends, just because it is hard, we need to keep pressing our patients for their under metabolic health. Honestly. I'm afraid uh the ease of PFA ablation has made us soft when it comes to working on hard things like lifestyle. If a patient with AFib loses weight, if they come off a blood pressure drug, if their LDL cholesterol goes down, triglyceride goes down, that cardiovascular risk plummets, and that is a good thing for patients.

I think it's important to consider an AF diagnosis more than just a rhythm issue, but an opportunity to help them lower their heart disease risk. Plus, as cardiologists, as doctors even, we have our patients attention. We have standing with our patients. We have an opportunity. We should use it wisely to try and help them with their metabolic health.

All right, last topic today is surgical left atrial appendage excision. The European Heart Journal has published the opinion trial of surgical left atrial appendage occlusion at the time of valvular heart surgery in patients without without AFib, but high stroke risk. A trial was conducted in three Chinese centers and the question was, does preventive appendage excision reduce thrombombolic events in patients with valvular heart disease?

It's an important question because I think some doctors do think that appendage is just bad in general. And if if you're in there doing heart surgery, it should be excluded. And we know from LayOS III, I think one of the best clinical trials in all of cardiology, we know that surgical closure of appendage at the time of surgery in patients who do have AFib reduces stroke by 33%.

and that's a lot, and it convinced me that at least the left atrial appendage is an important target for stroke prevention. Note also that the surgical treated patients in LayOS three continued anticoagulation. Opinion was a large trial Approximately twenty one hundred patients were randomized to either having their appendage closed mostly by cutting and sewing versus not having it closed.

The primary endpoint was the composite of ischemic stroke, TIA, or CV death in one year. Now, trialists powered to study to find a forty percent reduction in the primary endpoint. which is very, very optimistic. Too optimistic in my opinion, given that it is more than it more than LaOS, which was done in patients with Aphib.

And the opinion trial was null. The one-year primary endpoint occurred in 6.9% of patients in the closure group. And an 8.2% in a control arm, the hazard ratio was 0.83, but conference intervals were wide and went from 0.61 to 1.14, a p-value not even close to significant. The incidence of ischemic stroke was also very low and identical in both arms, two point five percent.

The CV death was even lower at zero point eight and zero point nine percent. Uh there were only fifty-two stroke events in opinion total and contrast that with two hundred eighty-two primary stroke events in Laos three. Post op A fib was also not reduced with the appendage exclus exclusion. So my comments, I think this is another example of an RCT attempting to answer a good clinical question.

The problem here was low power. I mean ischemic stroke rate is expected to be low, and appendage closure is only going to affect stroke. It doesn't affect CV death. So if the procedure does have a stroke lowering effect, they'd need a lot more patience to show it. De Layos three authors wrote the accompanying editorial. It's quite good. They included a nice table of trials finished or ongoing looking at preventive appendage closure during cardiac surgery.

The largest of which these trials is called LEAPS, which finished enrolling more than 6,500 patients to an atrial clip appendage closure strategy or not in high-risk atrium without AFIP. The primary end point of leaps is perfect. It's a schemic stroke and systemic embolism and the follow up is five years, but it is an event driven trial, which I think is a good thing. You just Have enough events and then you have the power.

For now, I think, though, given this data, there's no compelling reason to close the appendage in patients without AFib. I would not recommend it, but data is coming and we shall see. So that's it for this week in cardiology. As always, if you like this podcast, please take the time, give it a rating, write us a review.

Uh that helps people find us. If you don't agree with something I said, let me know and we can address it in the listener feedback. Until next week, this is John Mandrola from the Heart.org Medscape Cardiology.