Feb 10, 2023 This Week in Cardiology Podcast

John M. Mandrola, MD


February 10, 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 February 10, 2023, John Mandrola, MD comments on the following news and features stories.

SGLT2 Inhibitors in HFpEF

The European Commission (EC) has expanded the indication for dapagliflozin to include heart failure (HF) across the full spectrum of left ventricular (LV) ejection fraction (EF) — including HF with mildly reduced and preserved ejection fraction (HFpEF).

The decision was largely made on the basis of the DELIVER trial, which compared dapagliflozin vs placebo in patients with HFpEF. I discussed this on the September 2, 2022, podcast, but it’s worth a brief re-look, because my take is that SGLT2 inhibitor drugs can provide substantial benefit but that benefit turns on the underlying condition. In diabetes especially, in chronic kidney disease (CKD), and in HF with reduced EF (HFrEF), the drugs are beneficial, but I find the results in HFpEF underwhelming. I’d be less worried about it if the drugs were not so costly.

DELIVER randomly assigned about 6000 patients with HFpEF to dapagliflozin or placebo; the mean EF was 54%. The first point to make, and I think it is a huge teaching point when it comes to evidence translation is the difference between inclusion/exclusion criteria of a trial and the actual characteristics of enrolled patients.

  • For DELIVER, you had to have an EF ≥ 40% to be enrolled. But the mean EF was far greater than that. The mean LVEF was 54%. Only about one-third of patients had EFs between 40% and 49%. That’s a lot higher than in the HFrEF trials. These are different patients.

  • A primary endpoint of cardiovascular (CV) death or HF event —hospitalization/urgent visit for HF — occurred in 16.4% of the dapagliflozin group vs 19.5% in the placebo arm. The absolute risk reduction was 3.1% and the hazard ratio or relative risk reduction was 0.82 — or an 18% relative risk reduction.

  • The driver of the primary endpoint, as it was in EMPEROR-Preserved, was HF events. In a trial of more than 6000 patients that lasted more than 2 years there were only 87 fewer HF events in the dapagliflozin arm.

  • There were no significant differences in CV death or all-cause death.

I’ve spoken many times about hospitalizations for HF being only a fraction of total hospitalizations, so it’s a quite weak endpoint. You have to measure total hospitalizations.

For the Kansas City Cardiomyopathy Questionnaire, the mean placebo-corrected difference between baseline and month 8 among survivors was 2.4 points (95% confidence interval, 1.5 to 3.4). Recall that an increase by 5 points is considered meaningful.

Comments. From a regulatory standpoint, DELIVER was a positive trial. But when used in patients with preserved EF, SGLT2 inhibitors are low-value agents. They wouldn’t be low value if they cost no more than furosemide. But at the current cost, they could not possibly be high value drugs.

  • Some doctors are emotional about this class of drugs. And if I had a patient with HFpEF who had a creatinine of 1.6 and type 2 diabetes, I would absolutely use an SGLT2 inhibitor.

  • But in the absence of diabetes or CKD, how can you justify the cost of a tablet that has been shown to reduce only one of many types of hospitalizations?

  • What is more, it is totally normal in medicine for a drug class to have differential benefit/harm/value properties depending on the condition. Aspirin, for instance, is going to be more net beneficial in patients who have had lacunar strokes and stents than it would in a primary prevention patient.

When regulatory authorities approve a drug for an indication, it provides a tailwind for its use. And I know some of the purists who listen won’t like this, but I believe part of the job of a modern-day clinician is the parsimonious use of healthcare resources.

Just this morning I saw a Tweet pointing to a draft document from the UK National Institute for Health and Care Excellence (NICE) saying that they would not recommend either empagliflozin or dapagliflozin for patients with preserved or mildly reduced EF.

In the document, which is still open for comment, the authors wrote that it is not clear whether the drugs reduce the likelihood of dying from either any cause or from CV causes in this setting, and there are "uncertainties" in the modelling of the economic benefits the drugs deliver, making cost-effectiveness hard to determine.

HFpEF and Heart Rate

A few comments about last week’s discussion on the myPACE trial and the issue of heart rate and beta-blockers in patients with HF. I have a column forthcoming. In it, I try to capture the excitement of the moment.

  • This week, I had the opportunity to set pacing rates in a couple of patients with sinus node dysfunction and HFpEF. It made me think.

  • I also saw a handful of patients with HFpEF who were on high dose beta-blockers. Again, for the past two decades, I did not have any thoughts about the choices I had in those two scenarios.

  • The issue of beta blockers and heart rate choices comes up nearly every day. And that is what makes the topic so huge.

  • Also, I received a fair number of direct messages and texts from clinicians who have been avoiding beta blockers for years in these patients. So not everyone has had this blind spot.

I hope you read the column on | Medscape Cardiology when it comes out. And if any HFpEF academics accidently happen on to this podcast, I hope you are planning trials on beta-blocker withdrawal and trials replicating myPACE.

Stroke and AF

The electrophysiology service is often asked to see patients post stroke to weigh in on anti-thrombotic and monitoring decisions. Before seeing the patient, we always, and I mean always, visit the neuroradiologists. Not by Zoom, but in person, like in the old days.

There was an important presentation made at the International Stroke Conference (ISC) in Dallas this week. Journalist Sue Hughes, as always, has excellent coverage of the latest data from the STROKE AF trial.

As I tell you about this trial, remember Mandrola’s first rule of critical appraisals: is the purpose of the study marketing of a product or is the study truly asking an important scientific question?

  • STROKE AF, first published in JAMA in 2021, was a randomized trial of nearly 500 patients who had stroke from large or small vessel atherosclerosis. These are patients who normally would get antiplatelet and statin therapy.

  • One group was randomly assigned to an implantable loop recorder (ILR) made by Medtronic; the other group got usual care.

  • Guess what the primary endpoint was: no, it was not recurrent stroke, or death, or bleeding. The primary endpoint was detection of atrial fibrillation (AF) lasting more than 30 seconds.

  • With absolutely zero surprise, AF was detected in 12% in the ILR arm vs 2% in the control arm. The hazard ratio (HR) was 7.7.

This data was quite similar to the CRYSTAL AF study which also found much more AF with loop recorders vs standard Holter monitors in patients with stroke of unknown causes. But at least in this cohort, you needed to find a source of stroke.

At this week’s ISC congress, we learned the 3-year follow-up data from STROKE AF.

  • Recall that STROKE AF enrolled patients who were in their late 60s, had a mean CHADSVASC scire of 5, and had a vascular stroke.

  • After 3 years of monitoring, the rate of detected AF had risen to 21.7% in the continuous monitoring arm, vs 2.4% in the control arm (HR, 10.0; P < .001).

The presenting author, who disclosed financial relationships with Medtronic, said, "This shows that 30 days of monitoring with an external patch is not sufficient to exclude the presence of AF. And this really argues for a strategy of immediate insertion of cardiac monitor placement if your goal is to look for AF."

Sue Hughes remains one of the smartest journalists I’ve ever met. Right after that quote, her next section is subtitled: Is this clinically relevant AF?

  • The senior author, who recommended immediate insertion of the $10,000 monitor in stroke patients, had to acknowledge the truth: No one knows whether treating device-detected AF is warranted.

  • In fact, only two-thirds of the patients who had AF detected had an episode of AF that lasted more than 1 hour.

He then went on to say, "AF lasting more than 1 hour crosses the threshold for most practitioners I know to feel confident in treating the patient with anticoagulation. If it was symptomatic AF, this wouldn't even be a question."

In the EP world, that is not at all the consensuses. Our consensus is that no one knows, but we like to cite a substudy of the ASSERT trial, first author van Gelder, which found that nearly all the risk of device-detected atrial tachy-arrhythmias occurred in patients with more than 24 hours of AF.

Comments. Before I tell you about the tension I feel in this situation, let’s also set out the added complexity brought up by STROKE AF.

  • Here, the authors propose searching for AF in patients who have had an atherosclerotic source of stroke. These patients will be taking either aspirin or clopidogrel.

  • Now you search for AF and if you find it, you aren’t just talking about adding an oral anticoagulant (OAC) alone, say as you are in a patient with stroke of unknown cause. You are talking about adding an OAC to an antiplatelet drug. And that combination substantially increases the rate of major bleeding.

  • A net benefit is far from certain, so this is no easy call.

  • A better trial, if the authors and sponsor Medtronic wanted to answer the real question, would have been to behave like Richard Whitlock in LAAOS-3. In that trial of appendage closure vs no closure, they measured stroke.

  • STROKE AF is a much better trial if you randomly assign patients to ILR or standard care and measure stroke or bleeding. Then we’d see if this lucrative strategy of placing ILRs in post-stroke patients benefits patients as much as it does doctors, hospitals, and ILR makers.

  • Now to the tension I feel. Here is the thing; I am not an industry scold. Industry makes great products. The ILR, for instance, can be a great device. I’ve used it. It has changed care in some patients for the good. This is true of many things industry makes for us. Better EP catheters, better pacemakers, and implantable cardioverter defibrillators.

  • But it can go too far. Way too far. In the same way I rail against low-value use of SGLT2 inhibitors, I rail against low-value use of ILRs. In this device’s case, it could even be worse. Did you know, for instance, that in addition to the $10,000 price tag, patients are billed every 30 days for a download of data? These little things are money printing machines for doctors’ practices.

I’d be fine with all that if we had trial-level data showing that, in some group of patients, the device led to better outcomes. Consider what you would find if you put ILRs on patients in their late 60s who had elevated CHADSVASC scores who were in the hospital for a urinary tract infection or orthopedic or gall bladder surgery. You’d also find AF. Then what?

I think we have enough detection studies. We desperately need data from treatment trials like NOAH and ARTESIA.

Exercise as Medicine — Again

In January, JAMA-Internal Medicine published a research letter from a group in Shanghai. They present 10-year follow-up of an exercise randomized controlled trial. Ten-year follow-up is good. The trial is simple:

  • Patients with central obesity.

  • Three arms – vigorous exercise, moderate exercise, and standard care.

  • Primary endpoint – incident diabetes.

  • At the end of the 12-month active exercise intervention, all participants were encouraged to continue with healthy lifestyle and moderate intensity aerobic exercise.

  • Participants were followed up at 2-year and 10-year visits to assess the incidence of type 2 diabetes and the changes in body weight, waist circumference, and metabolic risk factors.

  • The risk of diabetes was reduced by 49% (RR, 0.51; 95% CI, 0.27-0.94; P = .01) in the vigorous aerobic exercise group and by 53% (RR, 0.47; 95% CI, 0.25-0.89; P = .01) in the moderate aerobic exercise group compared with the non-exercise group

  • Noteworthy in this trial was the focus only on exercise. Unlike other intervention trials, they did not combine diet or lifestyle intervention.

Comments. Okay, I know, a small unblinded trial with surrogate endpoints isn’t changing the paradigm of medicine. But if you listen to this podcast, you know I am going to cover these sorts of trials because one of my strongest biases is that doctors under-use exercise as medicine.

  • Many of my patients are retired. Many don’t exercise. I look them square in the eyes and tell patients that a daily regimen of some sort of exercise is a like a heart pill, or diabetes pill, or a blood pressure pill, and surely an AF pill.

  • I tell patients that if they care about health, they should carve out 15 to 30 minutes to do some sort of exercise. Todos los dias. Every day. Or every day they that eat. And Golf doesn’t count. Exercise before golf.

  • I specify that I am not talking about training. They don’t need to go to a gym, though they can. They don’t need to do crossfit, or speed training. Or get shelled on Zwift, like I do.

  • Patients just need to do something physical and sustained. Every day.

Here’s the thing, some people flip. Some people do it. They exercise as a pill frame sometimes works. And as it was in this trial from Shanghai, the habits can stick.

America College of Cardiology Meeting

ACC is coming up in a few weeks. This weekend I will work on my ACC preview. If you have any great studies, you think I might want to see, let me know. As journalists, we focus on the late-breakers and featured science, but we often miss things off the main stage.

Don’t be afraid to promote your work.


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.