Sep 06 2024 This Week in Cardiology

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 Sept 6, 2024, John Mandrola, MD, comments on the following news and features stories: Brief feedback on ABYSS and beta-blockers (BBs), more European Society of Cardiology (ESC) meeting recap including SHAM-PVI, FINEARTS-HF, and RESHAPE HF.

Feedback on Beta-Bockers

Feedback on my ABYSS discussion: Recall that was a trial looking at BB discontinuation at 1 year after myocardial infarction (MI) and preserved ejection fraction (EF). The trial was positive-ish for the continuation group, but it was driven by increased hospitalizations mostly for angiography in the BB discontinuation group. However, I opined that it was a positive trial for discontinuation because death and MI were not different.

Multiple emailers noted that there was no difference in quality of life (QOL) scores in ABYSS. That is interesting, and worthy of a comment, because we all hear how bad BB are for QOL. Yet, discontinuation did not cause an improvement in QOL.

Also, one of the senior authors of the REDUCE-AMI trial emailed me to remind me that they also looked at QOL and presented and published a QOL paper in the European Haert Journal – Cardiovascular Pharmacology, that found that the EQ5D, EQ-VAS, and WHO-5 QOL scores were similar between the randomized groups.

This lack of difference in QOL, seen now in two trials, reminds me to mention that BBs may be one of the more nocebo-effecting drugs we use. Patient after patient tell you in clinic that they feel terrible on BBs. Clearly, you cannot win the Tour de France on BBs, and some patients have symptoms (probably more on metoprolol than other BBs), but still, these sorts of data suggest that prior studies on placebo effect are probably real. Why, for instance, if BBs make people feel so bad, was there no difference in ABYSS when patients dropped the drugs in open-label fashion?

Why did the no BB arm of REDUCE-AMI randomized controlled trial (RCT) not report better QOL?

I am going to link to a fascinating study that I use in my clinic regarding the nocebo effect of BBs and erectile dysfunction (ED); first author Antonello Silvestri.

The study included 96 patients with heart disease, not suffering from ED, were placed on atenolol.

Group A: 32 patients were told nothing about the drug;
Group B: 32 patients were told about the drug but not about side effects;
Group C: 32 patients were told the drug could cause ED.

• The incidence of ED in the told nothing group was 3.1%;

• The incidence of ED in the told something group was 15.6%;

• The incidence of ED in the told about ED group was 31%.

My friends, words can harm and words can heal. Use words wisely. Especially around the use of drugs that have proven benefit.

After MI with preserved left ventricular (LV) function, I think we can safely withhold BB in the absence of some other indication.

SHAM-PVI

I will start the ESC meeting recap with SHAM-PVI.

This is a remarkable trial as it is the first proper sham-controlled trial for atrial fibrillation (AF) ablation, despite the fact that we have done this procedure for 20 years in close to a million patients.

I would call those who said we did not need such a trial incurious.  Of course, there is a placebo effect. When a patient with terribly symptomatic AF pre-ablation returns to the post-ablation clinic declaring that you changed his life because he feels so much better, but is in AF, then ablation is nearly 100% placebo. This situation is not common, but it is also not rare.

Led by Rajdip Dulai, from the United Kingdom, and published in JAMA, SHAM-PVI randomly assigned 126 symptomatic patients with persistent or paroxysmal AF to either cryoballoon pulmonary vein isolation (PVI) or sham with phrenic nerve pacing.

• Notably, 80% of the 67-year-old patients had persistent AF. Cardioversion was done for all patients in both groups who were in AF at the time of the procedure. Also notable was that the time to ablation from diagnosis was about 2 years.

• All patients had an implantable loop recorder (ILR) before randomization and this allowed the team to measure pre-ablation AF burden as well as AF burden at 6 months. They also measured QOL and well as time to event.

• The results were quite positive. Ablation vs sham resulted in an absolute change in AF burden from baseline of 60% in the ablation arm vs 35% in the placebo arm, which was a highly significant.

• All three QOL scores improved significantly at 6 months in the procedure arm.

• The blinding index results were interesting. After the procedure there was near complete blinding, but after 6 months, the guessing was much better, indicating the possibility that symptoms are “just’ the way the patient feels the PVI or sham.

Comments.  The first thing to say is congratulations to the UK team. This is a remarkable achievement.

And I believe they have shown that, yes, AF ablation has a placebo-resistant effect. They claim there is no clinically relevant placebo effect, but I don’t completely agree. I will come back to that.

The first nuanced factor was the choice to measure AF burden. We know that ablation will reduce AF burden vs sham or cardioversion. That’s not a question. This would be like measuring degree of stenosis reduction in the ORBITA PCI trial. The question at hand is whether symptom relief will be better in the active vs sham arm.

In fact, it was better in the active arm. But there are a few buts.

These were highly symptomatic patients — the  mean AFNET score of 53 at baseline is quite low. My friend Dihraj Gupta wrote to me that the mean score in CABANA was 63 (higher is better).

Gupta’s group in Liverpool have shown that QOL improvement following PVI is inversely proportional to the baseline AFEQT score. So, by starting with highly symptomatic patients with longstanding AF, they set the stage for large improvements in QOL. This is fine. I am not criticizing the choice, but merely pointing out that if we ablate early in the course of a patient’s AF, wherein they are less sure of symptoms, there may be less delta between sham and procedure. SHAM-PVI reinforces the value in not ablating too early.

I point everyone to the improvement in the sham arm. Some of this is due to cardioversion and antiarrhythmic drugs as it is well known that AF stays away in a small but substantial number of patients with simple shock. There is also a well-known regression to the mean and the natural history of AF.

What I am about to say is the most important message from this week’s podcast:

AF often goes away on its own. There is value in letting it happen if it will. In the absence of heart failure (HF) or decompensation, please my friends, give peace a chance to heal AF. Educate, reassure, treat metabolic risk factors, cardiovert, and see what happens.

Soren Diederichsen emailed that when you look at individual symptom scores say by the European Heart Rhythm Association classification (1 no symptoms, 2 mild, 3 severe symptoms, 4 disabling symptoms) then there was only modest improvement in the scores (Tables e9-11).

The final criticism is the lack of a transseptal (TS) in the sham arm. That they did not do this is totally understandable. It makes it ethically easier to do a sham trial.  TS increases risk to the sham arm, but there are clear effects on intra-atrial hemodynamics to the atrium with a TS. And we should recall that the heart has its own intrinsic nervous system and TS can perturb that, possibly in favorable ways. I realize that this argument sounds soft, and you might be right, but then please explain to me why many patients with reconnected PVs are cured of their AF.

A TS in the sham arm may have narrowed the difference in the two groups.

My conclusion: This is large step forward. It was bold, scientifically strong, and the best trial of the year in electrophysiology (EP). But I don’t think it closes the chapter on “quantifying” the placebo effect from ablation.

There are two ongoing sham-controlled trials and we will learn more when we take them together as a whole. I worry that many of these have too cautious a sham arm.

And a final note to the ethical naysayers. I mention the counter factual of what if SHAM-PVI had been an ORBITA or SYMPLICITY III HTN. In other words, what are the ethics of doing a QOL procedure in millions of patients and it turns out to be mostly placebo? SHAM-PVI-like trials should have been done sooner. They weren’t because EP, as a field, has too much confidence, and too little curiosity.

FINEARTS-HF Trial of Finerenone

Cardiologists are about to see more free donuts and burritos in their offices. The nonsteroidal mineralocorticoid receptor antagonist (MRA) finerenone has passed muster in a large cardiovascular (CV) outcomes trial in patients with HF with preserved EF (HFpEF).

The title of the FINEARTS -HF trial includes the words “mildly reduced or preserved EF” but the mean LVEF was 53% and only one-third of patients had an EF between 40% to 50%.

Nonetheless, HFpEF is a difficult condition to treat. Proponents say we have some available treatments, but technically speaking, the sacubitril/valsartan PARAGON HF trial was negative, as was the TOPCAT trial of spironolactone, and in EMPEROR-Preserved, the reductions in hospitalizations for HF (HHF) were modest and only a small fraction of total hospitalizations.

My point is that, unlike HF with reduced EF (HFrEF), where the four classes of medical therapy have clear additive benefits, patients with HFpEF, garner far less benefit on average.

That’s why FINEARTS is so important. HFpEF is not only harder to treat, but also more prevalent than HFrEF. 

• The trial enrolled 6000 patients with class 2 to 4 HF and an EF more than 40%. One group got placebo and the other finerenone.

• Over 32 months, a primary endpoint of total HF events (first or recurrent hospitalization or urgent visit) and CV death occurred in 14.9 per 100 patient years in the finerenone arm vs 17.7 per 100 patient years in the placebo arm. This was a 16% reduction with a hazard ration (HR) of 0.84 (95% confidence interval [CI] 0.75-0.95).

• The composite endpoint was driven by an 18% reduction (HR 0.82; CI 0.71-0.94) in HF events, but the rate of CV death was not significantly different, 8.1% vs 8.7%; not statistically significant.

• The results for the primary outcome were consistent across all prespecified subgroups, including those defined according to baseline LVEF (<60% or ≥60%) and baseline use of SGLT2 inhibitors (yes or no).

• As for secondary endpoints, the Kansas City Cardiomyopathy Questionnaire (KCCQ) QOL score improved in both groups but the delta was tiny. So, little effect in my opinion. There was also no difference in the number of patients who improved by one class of the New York Heart Association functional class.

• Death from any cause was not different.

• Total adverse events were similar but in the finerenone arm, 3% vs 1.4% of placebo patients had potassium levels > 6, though no episodes of hyperkalemia led to death.

Comments. There were several other papers on MRAs and finerenone published around the time of the ESC. We already know that finerenone has two positive trials.

FIDELIO-DKD. We know that finerenone improves kidney outcomes in patients with chronic kidney disease (CKD) and diabetes. FIDELIO-DKD  showed an HR 0.82 (CI 0.73-0.93) for composite of kidney failure, sustained decrease of eGFR by 40%, and death from renal causes. There was also secondary outcome of major adverse cardiac events (MACE) and it was lower in the finerenone group by 1.8% in absolute terms.

FIGARO-DKD. This study looked at finerenone vs placebo in patients with CKD and diabetes with a CV MACE as primary endpoint. Composite primary was CV death, MI, stroke or HHF. HR was 0.87 (CI 0.76-0.98). absolute relative risk was 1.8%

I am a huge fan of MRAs. I call them my secret weapon. The RALES trial of spirinolactone for HFrEF shows one of the largest effect sizes in mortality reduction in all of cardiology.

FINEARTs was positive, though the risk reduction is modest in absolute terms and is driven solely be a HF events, not CV death. JACC published a very nice paper this weekend from HF trialists, titled “Why Have We Not Been Able to Demonstrate Reduced Mortality in Patients With HFmrEF/HFpEF?”

It’s a long essay, but the gist of it is simple. Patients with HF with decent EFs are older and have more competing causes of death. A patient with HFrEF is sick because of a weak heart. A patient with HFpEF is sick with lots of things and the diastolic dysfunction is only one issue.

HFpEF, in fact, is a great condition for thinking about the domains that affect treatment effects.

When we treat a patient with HFpEF, we have to think about two opposing forces.

One is the risk of the primary outcome vs the risk of a competing cause of morbidity. The other opposing force is the chance of treatment benefit vs treatment harm.

In a young male patient with HFrEF, there is a high risk of a primary outcome, little chance of a treatment- related harm, and minimal competing causes.

It’s the opposite in an 80-year-old woman with hypertension, diabetes, CKD, arthritis, immobility, poor social supports, and a high salt diet who has recurrent HF admissions. Finerenone treats one of these many issues. Of course it cannot reduce CV death, because there are a blue million ways this woman could die.

With FINEARTS-HF, the elephant in the room is not how finerenone does against placebo, but how it goes against spironolactone. The design of FINEARTS-HF, with a comparison against placebo, seems to assume that the TOPCAT trial of spironolactone in HFpEF was totally negative. Yet everyone knows the TOPCAT story, that if two countries were excluded,spironolactone has a beneficial effect.

So what a clinician needs is a finerenone vs spironolactone trial, because the question is not whether MRA’s reduce HHF in HFpEF but is finerenone superior to the far less expensive spironolactone? I suspect that it will be marketed as having less steroidal side effects, and less hyperkalemia, but should we not go first with spironolactone and see how it is tolerated? I am asking but I think the answer is yes.

RESHAPE-HF

At the ESC we learned the results of the third RCT for transcatheter edge-to-edge repair (TEER, aka mitraclip) for functional mitral regurgitation (MR) in patients with HFrEF. As you all know, MITRA-FR was stone cold negative. No difference. In COAPT, TEER was almost as good as clean drinking water or antibiotics for bacterial infections.

These are two of the most opposite trials I have ever seen in cardiology. Of course, proponents of TEER say that COAPT recruited the perfect patient.

Anyway, finally, we get the results of RESHAPE-HF, the third trial in the space.

The trial randomly assigned 505 patients with functional MR and LV systolic dysfunction to TEER or medications. No sham so it was open label.

Primary endpoint was…wait, there were three primary endpoints.

1. The rate of the composite of first or recurrent HHF or CV death for 24 months;

2. The rate of first or recurrent HHF for 24 months;

3. The change from baseline to 12 months in the KCCQ; scores.

Before I tell you anything about the results, stop and think. Why three endpoints? Why a subjective endpoint in a trial with one group getting a big caring procedure and the other plain tablets?

I am sure you are thinking that the chance of RESHAPE being negative with these endpoints is pretty low. And indeed, RESHAPE HF was positive:

Results:

• At 2 years, the rate of CV death or HHF was 37 events per 100 patient-years in the device arm vs 59 events per 100 patient years in the control group (rate ratio, 0.64; 95% CI, 0.48-0.85; P = .002).

• First or recurrent HHF was 27 per 100 patient-years in the device arm vs 46.6 per 100 patient-years in the control group (rate ratio, 0.59; 95% CI, 0.42-0.82; P = .002).

• The mean difference in KCCQ score over 1 year was 10.9 points higher in the device arm. This too met statistical significance. 

• More impressive risk reductions in the first primary endpoint were notable in subgroups with more severe heart failure; worse MR, higher effective regurgitant orifice area (EROA), and especially in patients who had been hospitalized for heart failure within a year before randomization. 

These were the results that made headlines. Let’s move now to the endpoints less emphasized on social and news media. 

• Despite the 20% absolute risk reduction in heart failure events, the device did not produce a statistically significant reduction in all-cause mortality, CV death, or the rate of total hospitalizations for any cause. 

• Nearly one in five patients died from CV causes and the device, which reduced MR and reduced HF events, did not have an effect on CV death or all-cause death.

In my column I listed multiple reasons to be cautious about this trial.

• It took 8 years to recruit for the trial from 30 centers. So, these were highly, highly selected patients.

• RESHAPE was open label. The two drivers of a positive endpoint were HHF and KCCQ. Obviously, the KCCQ is susceptible to knowing your treatment assignment, but so are HF admissions, because the decision to put a person in the hospital requires a clinical decision. Knowing the treatment assignment may bias such a decision. This is not nefarious; rather it is clinical. A patient with a clipped mitral valve who has fluid overload may be more likely to receive an outpatient diuretic than a person without the bright white clip visible on echocardiogram.

• Third caution: Why doesn’t the MR treatment reduce hard outcomes? If the device is so great for people. If it shreds HHF, why doesn’t it prolong life in these sick patients. That it doesn’t affect hard endpoints strengthens the arguments for performance bias in the matter of the hospital admissions endpoint.

• There is a similar argument for all-cause hospitalizations. If the device is great, and it reduces HF, and these patients have serious HF, why don’t we see a reduction in all-cause hospitalizations? This, to me, speaks to the fact that many of these patients have competing causes of illness. Or, of course, that the effect from TEER is simply not that great.

• A technical concern was differential loss to follow-up. Premature termination of the trial occurred in 41 patients in the control arm vs 13 in the device arm.

RESHAPE enrolled patients with less severe functional MR. Some at the meeting and some online suggested that we should use this trial to expand indications for TEER. I disagree strongly. For the reasons listed above, plus the observation that the subgroups in RESHAPE showing the greatest benefit were those with more advanced MR (recent hospitalization, severity of MR, a larger EROA).

In sum, TEER is a nice innovation. Taking these trials together, I get the sense it helps some, perhaps few, patients with functional MR. The challenge seems to be choosing those most likely to benefit.

Sadly, there are other factors that go into these decisions — here in the US, it is a) doctors love doing procedures, b) doctors and hospitals are paid well for the procedures, and c) the more procedures you do and bill for, the more status you get.