Sep 01, 2023 This Week in Cardiology Podcast

John M. Mandrola, MD


September 01, 2023

Please note that the text below is not a full transcript and has not been copyedited. For more insight and commentary on these stories, subscribe to the This Week in Cardiology podcast, download the Medscape app or subscribe on Apple Podcasts, Spotify, or your preferred podcast provider. This podcast is intended for healthcare professionals only.

In This Week’s Podcast

For the week ending September 1, 2023, John Mandrola, MD comments on the following news and features stories.

European Society of Cardiology (ESC) Meeting: Part 1

Wow. What a meeting in Amsterdam. I just love that city. I only got in a car twice — to and from the airport. Otherwise, I used a rental bike. Even when I wore my suit. Imagine riding a bike with an Italian suit. Perhaps my favorite was riding at night home from dinner.

I also want to thank Irish cardiologist Dr Daniel O’Hare for reminding me about the ParkRun 5k on Saturday. He and Soren Diedricksen shredded me, but I loved it. On the meeting itself, either that was a small congress hall, or the attendance was amazing. Because there were a lot of people. Meetings are back, I guess. ESC did a great job of attracting a lot of great studies too. I will probably take three podcasts to cover them all.

Finally, to all of those who mentioned the podcasts, I say thank you times 1000. It means a lot to me. It motivates me to keep trying to get better. A super special thanks to Dr. Robert Tielman from the Netherlands. You know why. 


I am going to start my review with the most important study presented at ESC. No, it wasn’t on semaglutide, or atrial fibrillation (AF) ablation, or anything about percutaneous coronary interventions (PCI). This study should send shock waves through you every time you move to apply evidence from trials to a patient, especially if you think you are following guidelines or quality measures.

This study dealt with oral anticoagulants (OAC) for stroke prevention in patients with AF. Set in the Netherlands, in multiple general practice offices, associate Professor Dr. Geert-Jan Geersing, and his PhD candidate Dr. Linda Joosten, designed a study to test the safety of switching patients from vitamin K antagonists to direct-acting OACs (DOACs).

Side bar – in the Netherlands, the vitamin K antagonists acenocoumarol or phenprocoumon are much more commonly used than warfarin. They all work in the same way and elevate the INR.

The switch sounds like a no-brainer, right? Ruff and colleagues’ meta-analysis of the four seminal warfarin vs DOAC trials found that DOACs are better at stroke prevention and associated with numerically less major bleeding. Here is why FRAIL AF tops the list: Because they studied the switch in elderly patients with frailty.

I’ve got a column coming out soon on FRAIL AF, so I won’t give it all away. The short story is that FRAIL AF investigators designed the study out of a bit of frustration that there is really no evidence on the issue of using DOACs in the frail and elderly because so few of these patients were included in the trials. And since, by definition, frailty increases vulnerability to adverse effects and it comes with many competing causes of morbidity, we can’t be sure that DOACs would prove superior in these sorts of patients.

  • FRAIL AF randomly assigned patients with a mean age of 83 years with serious frailty and who were managed on vitamin K antagonists to either stay on the vitamin K antagonist or switch to a DOAC.

  • The question was safety, that is, major bleeding.

  • They powered the study to show DOACs were superior to vitamin K antagonists for major bleeding.

  • Boom. That is not what they found. Not at all.

  • The teaser is that after the first interim analysis the trial was halted for futility and harm in the switch to DOACs group, bleeding was higher in the DOAC group.

  • Imagine designing a trial for superiority and finding not only no superiority but the active arm was worse.

This is huge. In my column I make three arguments why, if you lose evidence to treat patients, FRAIL AF is the most important study from the meeting. Taking average effect sizes from trials and applying them to the patient in front of you will become the greatest challenge facing the doctors of the future.

The surprising finding from FRAIL AF should infuse us all with super high doses of humility in the use of evidence. I hope this study finds its way to guideline writers and anyone silly enough to think quality measures are a positive.


FRAIL AF should have led the meeting, but it did not. A trial called STEP-HFpEF led it off. Made the New England Journal of Medicine and earned a ton of press coverage and praise. STEP-HFpEF studied the use of the glucagon-like-peptide 1 (GLP-1) agonist semaglutide in patients with heart failure with preserved ejection fraction (HFpEF). HFpEF is a serious condition, it causes a lot of hospitalizations and leads to cardiovascular (CV) death. We need better therapies for HFpEF.

So, you probably figured the lead hot-line trial at the biggest heart meeting in the world would enroll many thousands of patients and address hospitalizations for HF (HHF) or CV death. Well, you would be wrong. STEP-HFpEF is a classic marketing masquerading as science-type study.

  • The industry-funded study enrolled only 500 or so patients who had obesity and HFpEF and measured — sit down for this — weight loss and quality of life (QOL) measured on the Kansas City Cardiomyopathy Questionnaire.

  • Yes, of course, the first of the dual primary endpoints, weight loss was positive. Patients lost weight. We already know, from multiple studies, that this drug induces weight loss.

  • For the second primary endpoint, QOL, semaglutide treatment was also associated with better scores. But again, which patient who is obese and then loses a ton of weight does not report feeling better?

  • They also looked at 6-minute walk tests and a host of biomarkers. All better with weight loss.

  • If this is the metric for GLP-1 agonists, you could literally use them in any condition in which patients are obese.

I am not against this drug class. The ability to induce weight loss could lead to a massive gain in health. Right now, I feel like obesity is the one of gravest threats to health in the Western world. I see it every day.

Yet, if you are looking to show that your drug specifically helps patients with a condition as serious as HFpEF, design a proper study that looks at clinical outcomes, such as HHF or CV death. I read somewhere that the valuation of Novo Nordisk, the maker of semaglutide, is close to or surpassed the gross domestic product of Denmark. There is plenty of money.


Not every day, but most days I get a notification from our pacemaker center that a patient with oodles of stroke risk factors has had some atrial high-rate episodes (AHRE). AHRE are really AF, but short in duration. Maybe minutes, maybe hours, but not long enough to visit a doctor to get a 12-lead ECG.

Should we start an OAC? That is the question that comes up often, but before ESC, no one had any answers. Smart watches have made this problem worse because they, too, pick up short duration AF, albeit with a lot lower specificity.

At ESC, we heard the details of the The NOAH AFNET 6 trial, or NOAH for short.  It answered some questions. How many depends on your point of view. Let’s do the details:

  • 2500+ patients with AF > 6 minutes captured on a cardiac device (pacer or implanted cardioverter defibrillator [ICD]), mean age 78 years, were randomly assigned to either a DOAC (edoxaban) or placebo. About half the placebo group were on aspirin for aspirin indications, such as coronary artery disease (CAD).

  • The primary efficacy endpoint was a composite of stroke, systemic embolism, or CV death. Primary safety endpoint was death or major bleeding.

  • We knew before ESC that NOAH was stopped for futility. But at ESC we also learned that it was stopped for harm.

  • The safety outcome — death or bleeding — occurred in 5.9% of those in the edoxaban vs 4.5% in placebo. A statistically significant 1.31%, or 31% higher risk. The higher risk was driven by 28 more major bleeding events in the edoxaban arm.

  • As for efficacy, a primary outcome – stroke, systemic embolism, or CV death —occurred in 3.2% vs 4.0% in the edoxaban vs placebo arm. This hazard ratio (HR) was 0.81 but the confidence interval (CI) went from 0.60 to 1.08 and thus is not considered statistically significant.

Here are some of the take-homes that I wrote about in a column on|Medscape Cardiology

  • First was that this was strong evidence against treating short duration AF with DOACs. Consider that these were high-risk older patients; mean CHADSVASC score was 4.  — And bleeding rates were substantially higher than stroke reduction.

  • Second was that there is another trial, called ARTESIA, which is studying similar patients, and testing apixaban vs aspirin. It will recruit more patients, follow them longer, and it has not been stopped for futility. The problem of course is that having aspirin in the control arm — which I think is a bad idea — will decrease the difference in bleeding. Nonetheless, rumors have ARTESIA reporting at the American Heart Association meeting in 2 months.

  • Third, NOAH clearly advances knowledge about the prognosis of AF and how it differs depending on the diagnosis. In the past, AF had to last long enough (or create enough symptoms) to trigger a clinician to order a standard ECG. Continuous monitoring requires neither symptoms nor long-duration AF.

  • NOAH proves that short-duration AF is a lower-risk condition than a 12-lead ECG-defined AF. (I am beginning to believe that short-duration AF might be a condition of aging — akin to gray hair and wrinkles.)

  • Fourth, NOAH is like FRAIL AF in that it demonstrates the power of the randomized trials. I constantly hear how trials are too expensive to run. And they are pricey. But now consider the costs of inappropriately treating the millions of people with short duration AF. In the end, many trials pay for themselves and then some.

  • Finally, NOAH further decreases my enthusiasm for AF screening. Screening programs often pick up short duration AF. That’s fine, but picking up AF, is not enough. You have to show that the treatment of picking up a disease early is beneficial. NOAH suggests it is not.

One caveat: an astute reader sent me a message that there may be a sweet spot of picking up AF, that is, one of the reasons NOAH was a null trial, was that older patients have too many competing co-morbidities. Screening, and picking up AF, in younger patients, may still be worthy, as younger patients have fewer competing causes of stroke. (I don’t agree, but the commenter makes a good point thinking about competing risks of the primary outcome). 


One of the most exciting things in the AF world is a new technology to perform AF ablation, which is basically electrical isolation of the pulmonary veins (PV). It’s called PFA, or pulsed field ablation. The idea is to send electricity (think shocks) through a catheter and it creates pores in the myocyte membrane. We call this electroporation. The neat thing is that it is supposed to affect only cardiac tissue, which is important because the most feared complication is thermal damage to the esophagus which can be deadly. Phrenic nerve injury is also a big issue with cryoballoon (CB) ablation.

At ESC, Vivek Reddy presented results of the ADVENT trial which was essentially an FDA regulatory-type trial. PFA is approved in Europe and widely used in some countries there.

  • 300 patients with paroxysmal AF, at many US centers; PFA vs thermal ablation. Thermal could be with radiofrequency (RF) or cryoballoon (CB).

  • The primary endpoint was a composite of freedom from initial procedure failure, any atrial tachycardia or AF, antiarrhythmic drug use, cardioversion, or repeat ablation.

  • The results were very similar. Kaplan Meier curves were pretty much superimposed. Success rates were in the 71% to 73% range. Procedure times were shorter with PFA.

  • Safety overall was also similar. Not better, but similar, although the only two tamponade cases were in the PFA arm, and one of these patients died from this procedural complication.

  • No atrial esophageal fistulas, which you would expect; since the incidence of this is super low, a trial of 300 patients would be unlikely to pick that up.

  • Phrenic nerve injury (PNI) did occur in the PFA arm. I thought it was cardioselective. All four cases of the PNI in the PFA group resolved. But it occurred.

I also wrote a column about this, and I am underwhelmed. Proponents and some people I have a lot respect for disagree. They think I should be excited. They say that this will ultimately prove safer, especially in the hands of low-volume ablation operators.

My reasons for being underwhelmed:

  • PFA is clearly not more effective, although it is a first generation device.

  • I worry a lot about brain lesions. There are intense bubbles formed in the left atrium. A brain MRI sub-study of Advent showed zero asymptomatic cerebral lesions in the thermal group vs three in the PFA group. We should have more study of post PFA brain scans.

  • PFA is faster but how much faster than an hour do we need to be?

  • If you must have additional lesions, like for atrial flutter, then you have to open a new catheter, then costs go up. PFA will surely be more costly. Future iterations may fix that.

  • Finally, though, this is not a breakthrough because a new way to destroy atrial tissue gets us no closer to finding the causes of AF.


The last two trials I will cover deal with the necessary collaboration between an electrophysiology (EP) doc and a HF doctor. I really like the BUDAPEST trial because it addressed a decision that most assume is the correct one. The history of medicine is full up with examples where this reasoning has been disastrous, so good on the investigators.

BUDAPEST looked at two options for the common scenario wherein patients have cardiomyopathy and high amounts of right ventricular (RV) pacing. Some brief background.

  • If you have severe bradycardia, then RV pacing is better than no pacing. But RV pacing causes a dyssynchronous RV-LV contraction. That’s because RV pacing causes the RV to contract first, then conduction goes muscle to muscle and then the LV contracts. This dyssynchrony leads to LV dysfunction, or pacing induced cardiomyopathy (CM), in 10% to maybe 30% of cases. Pacing induced LV dysfunction is quite similar to left bundle branch block-induced LV dysfunction.

  • It can be corrected by adding a coronary sinus (CS)/LV lead, because now you can bi-ventricularly (biV) pace and activate the RV and LV simultaneously, narrow the QRS, and improve LV function.

  • We know from seminal trials in patients with HF with reduced EF (HFrEF) that de novo implantation of devices capable of biV pacing, now called cardiac resynchronization devices or CRT, outperform standard ICDs.

But we did not know if “upgrading” devices to a CRT would be equally effective. It was not a sure thing, because an upgrade procedure is more complicated. Consider a person with a dual chamber pacer, HFrEF and what this procedure entails. You would have to add a new RV defibrillator lead (pacing leads cannot be used for defib), a new CS/LV lead, and possibly extract the existing RV lead. That’s a big deal.

  • BUDAPEST CRT randomly assigned these patients who had high pacing amounts (≈85%) to either upgrade to CRT-D (CRT with defibrillator) or ICD. It was similar to MADIT CRT: active arm was CRT-D, control arm was ICD.

  • The primary outcome was a composite of death, HHF, <15% reduction of LV end systolic volume.

  • The results were markedly positive for the CRT-D arm. After 1 year of follow-up, a primary outcome occurred in 32% of patients in the CRT-D arm vs 79% in the ICD arm (odds ratio, 0.11; 95% CI, 0.06 - 0.19; P < .001).

  • The key secondary outcome of all-cause mortality and HHF occurred in 10% of the CRT-D arm vs 32% in the ICD arm (HR, 0.27; 95% CI, 0.16 - 0.47; P < .001).

  • Overall death occurred in 5.6% vs 11% of the CRT-D vs ICD arms. Interesting was that the rate of serious ventricular arrhythmia was much lower in the CRT-D arm (0.5% vs 14.5% for ICD).

  • Complications were similar.

Comments. In this case, common wisdom, plausibility was correct. Some will look at an 80% reduction in the primary endpoint and think that’s too big. They would be right to worry about large effect sizes, see the next trial I discuss, but in this case, I think it fits the clinical scenario.

Why? Because I liken CRT to the Lazarus procedure. When you choose patients well, which is important, and relevant to BUDAPEST, because they took 7 years to find these perfect patients, CRT can often massively and rapidly improve heart function. I’ve had people look and feel better in the recovery room after establishing biV pacing.

This is because pacing induced cardiomyopathy is one of those times you can reverse the LV dysfunction. And if that happens, outcomes improve.

One weakness of the study and one I would love to see studied is the authors’ choice to use only CRT-D devices. I think you could accomplish the same with CRT-pacemaker (CRT-P). Upgrading only to a pacemaker would simplify the upgrade procedure, as you would only need to insert one LV lead.

Witness the fact that only one patient in the CRT group had a serious ventricular arrhythmia. Ongoing trials are happening now for CRT-P and CRT-D and I am putting my money down on non-inferiority.

Think about it: if CRT corrects the LV dysfunction, there is no longer HF, and if there is no longer HF, there is no longer a need for an ICD.

Great job to the BUDAPEST CRT team.


AF and HF often occur together. Sometimes AF is the cause of deterioration. There is this select group of patients who have HFrEF, then they have AF, and boom things get bad.

  • The first CASTLE AF trial found that in patients with AF and HFrEF and ICDs, AF ablation reduced a composite endpoint of death and HHF compared with medical therapy alone.

  • The CASTLE HTx screened about 900 patients and randomly assigned 200 patients with HFrEF who had AF and who were referred to their center in Bad Oeynhausen, Germany, to either AF ablation or medical therapy.

  • These patients were 62 to 65 years old, mostly male, and had LVEF 26% to 29%. NT-proBNP levels were high at > 3800 and 6-minute walk test was around 300 meters.

  • The results were positive. The trial had to be prematurely stopped after 18 months. 

  • Of the 97 patients randomly assigned to ablation, 16 did not undergo the procedure. Of the 97 patients in the medical therapy arm, 16 crossed over to receive ablation.

  • After a median follow-up of 18 months, a primary endpoint occurred in 8% of patients in the ablation arm vs 30% in the medical therapy arm (HR, 0.24; 95% CI, 0.11 - 0.52; P <.001).

  • There were no primary endpoint events in any of the 16 patients in the ablation group who did not undergo ablation.

  • Death from any cause occurred in 6% of the ablation group vs 20% of the medical therapy arm, (HR, 0.29; 95% CI, 0.12 - 0.72).

  • LV function, AF burden and less use of amiodarone were all better in the ablation arm.

Comments. These results really made a stir. On the one hand, ablation proponents rejoiced. See, we can help a lot in the treatment of patients with HF and AF.

And since I have seen such patients, done AF ablation on them, and watched them do great, with less AF and  better LVEF, I can understand that if you select these patients well enough, AF ablation can be great procedure.

But. Critical appraisal people found a lot not to like. Many criticized me for being too soft in my column, which I hope you read.

Here are the criticisms:

  • It’s a single center study and was unblinded. There is a risk of performance bias because ablation proponents are doing the study.

  • Sanjay Kaul and Milton Packer  both pointed to the huge effect size, and the small number of events.

  • Then there is the matter of super-early separation of the Kaplan Meier curves. The authors explain this by saying that these patients were super sick and some were prioritized for AF ablation.

  • But that isn’t an easy explanation because their EFs weren’t much different from the first CASTLE study (wherein the Kaplan Meier curves separated after 3 years not 3 weeks).

  • Also problematic is that the average duration of AF was years. So it is hard to posit acute decline due to new AF.

  • Another problem is that many of the major baseline characteristic favored the ablation arm—patients were younger, had better EFs, and lower nt-proBNP, for instance.

  • Some of my colleagues called for a large-scale multicenter trial. Another friend said, hey, why didn’t the AMICA trial of AF ablation vs meds in patients with persistent AF and HF show positive results?

I wonder about this too. The best conclusion I can come up with now is that it is patient selection. If you take patients whose meds have been failing, and randomly assign them to AF ablation vs more of the failing meds, then ablation will win.

The best take home that I can apply to my practice next week is that at minimum this data supports the early involvement of the EP team in patients with HF and AF. We may be able to help in selected patients.


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.