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 Aug 30, 2024, John Mandrola, MD, comments on the following news and features stories: Four Friday studies from the European Society of Cardiology (ESC) meeting, including post-myocardial infarction (MI) beta blockers, a new therapy for transthyretin amyloidosis with cardiomyopathy (ATTR-CM), and the strategy of holding RAS inhibitors before non-cardiac surgery; good news in heart failure with preserved ejection fraction (HFpEF), and more previews from ESC.
ABYSS: Post-MI Beta-Blockers
This morning, Prof Johanne Silvain presented results of the ABYSS trial comparing beta-blocker interruption vs continuation at 1-year after an MI. Stopping beta-blockers after MI is a hot topic in cardiology.
This spring, at the American College of Cardiology (ACC) meeting, we learned results of the REDUCE-AMI trial, a Swedish randomized controlled trial (RCT), that compared beta-blocker use right after MI vs no beta-blocker use. These were patients with normal left ventricular EF (LVEF). They reported no difference in the primary endpoint of death or new MI, though the confidence intervals (CI) were wide, ranging from a 21% reduction to a 16% increase in events.
Before I say anything about the details of the French ABYSS trial, I should set out that the data supporting post-MI beta-blockers is extremely old (way before percutaneous coronary intervention [PCI]) and quite weak. So, our pre-trial priors of ABYSS should be that discontinuation of beta-blockers at 1 year makes little difference. Remember friends, trials never give an answer, rather, they modify our pretrial beliefs, much as medical tests do.
ABYSS had a few differences from REDUCE AMI. First was that you could be randomly assigned with an LVEF > 40%, though the actual mean LVEF was 60% and the number with an EF between 40% to 50% was 23%.
Another difference was that interruption came at 1 year not right away.
But the biggest difference was the primary endpoint. In REDUCE AMI it was simple: death and MI. In ABYSS, it was not simple. It was a composite of a lot of things — death, MI, stroke, or hospitalization for any cardiovascular (CV) reason. (I’ll come back to that.)
ABYSS, like REDUCE AMI, was open-label, so patients and caregivers knew the treatment group.
The topline results seemed to contradict REDUCE AMI. Over a 3-year follow-up (pretty strong), a primary endpoint occurred in 23.8% of the interruption group vs 21.1% in the continuation group. The upper bound of 95% CI here was 5.5%. That 2.8% higher rate of bad things corresponded to a relative risk (RR) that was 16% higher. The hazard ratio (HR) was 1.16 with a CI of 1.01 to 1.33.
The authors also did a quality-of-life (QoL) assessment and found no differences. Which is surprising.
The study authors chose a non-inferiority (NI) assessment with 3% as the NI margin. Since they estimated the event rate in the continuation arm of 12%, that corresponded to a 25% RR margin.
The upper bound of the rate of events in both absolute terms and relative (the worst-case scenario) exceeded the NI margin. In other words, there were more events in the interruption arm and ABYSS did not reach NI.
ABYSS seems to contradict REDUCE AMI.
But. Let’s look closer at the results. A look at the actual primary outcome events tells a different story. Recall that REDUCE AMI looked only at death and MI. But ABYSS had four components — death, MI, stroke, or hospitalization for a cardiac reason.
In ABYSS:
-
The rates of death were 4.1% and 4.0%, so no difference;
-
The rates of MI were 2.5% and 2.4%, so no difference;
-
Stroke was 1.0% in both arms:
-
CV hospitalization rates were 18.9% in the interruption arm vs 16.6%.
-
Had ABYSS chosen the same endpoint as REDUCE AMI, it would have gotten the same result.
The most common cause of hospitalization was “coronary-related” (angina and angiography were the major drivers of hospitalization). Recall that ABYSS was an open-label trial.
By email, my friend Dr. Bogdan Enache, a person who you should be following on X wrote to me that this being a French trial may explain the higher rates of angiography in the interruption group. Bogdan practiced in Monaco for a few years. He wrote that:
French patients are followed-up by out-patient cardiologists who are traditionally seen as cardiologists who refer their patients for a lot of angiographies.
My final conclusion: The results of ABYSS may appear different from REDUCE AMI, but clinically speaking, the two trials are similar.
If you count only strong clinical endpoints (death, MI, stroke), both trials show no harm from either not using beta-blockers or interrupting the drugs at 1 year.
The driver of beta-blocker-interruption missing NI compared with continuation, was a bias-prone soft endpoint of minimal clinical significance.
I respect the idea that trials should be interpreted strictly given their primary endpoint and trial procedures. But that strict scientific interpretation need not affect clinical translation at the bedside.
I will continue to stop beta-blockers in post-MI patients who have no other important reason to be on the drugs.
ATTR CM – HELIOS B
Another ESC trial presented this morning is the HELIOS-B trial.
This was a regulatory trial studying the drug vutrisiran, a subcutaneously administered RNA interference therapeutic agent, which inhibits the production of hepatic transthyretin. It’s given as an injection once every 3 months.
The company that makes the drug issued a press release earlier this summer saying that HELIOS B delivered positive results.
Sue Hughes has nice coverage of the press release from July 6. She tells us that vutrisiran is already available in many countries for the treatment of amyloidogenic mutated transthyretin-related polyneuropathy, which is not as common as the cardiomyopathy (CM) complication.
She also reminds us that transthyretin-mediated amyloidosis with CM results from the aggregation and deposition of transthyretin amyloid fibrils in the heart and various tissues. It’s a progressive disease that is ultimately fatal.
In fact, vutrisiran is approved for ATTR amyloidosis with polyneuropathy on the basis of the HELIOS A trial. A substudy of that small trial suggested potential cardiac benefits (BNP).
One more background point: ATTR CM occurs as a result of inherited TTR gene variants (hereditary, or variant, ATTR amyloidosis), often with a mixed phenotype that also includes polyneuropathy, or it occurs with aging in the absence of predisposing TTR variants (wildtype ATTR amyloidosis).
HELIOS B was placebo-controlled, vutrisiran vs placebo, in 655 patients with ATTR CM who had a history of HF hospitalization (HHF). About 40% of both groups were receiving tafamidis. Average age of patients, 77 years; 92% male.
The primary end point was a composite of death and recurrent CV events (hospitalization for cardiac cause or urgent HF visit). Secondary end points included death, the change from baseline in the distance covered on the 6-minute walk test, and the change from baseline in the Kansas City Cardiomyopathy Questionnaire (KCCQ) score.
The trialists broke up the groups into vutrisiran vs placebo in overall population and vutrisiran vs placebo in the monotherapy population (those not on tafamidis).
The main results over a trial period of 36 months:
-
In the overall population, a primary outcome event (death or CV event) occurred in 38% vs 48% with placebo. An ARR of 10%; HR 0.72 (CI 0.56 to 0.93). Death was 31% lower, recurrent CV events 28% lower, so both drove the primary endpoint.
-
Results were similar in the monotherapy arm as well. Very similar hazard ratios and CI. The Kaplan Meier curves started diverging after about a year.
-
Both the 6 minute walk test and KCCQ declined over the trial period in both groups (patients progressed); the median changes were less in the active arm. Similar patterns were seen in both the overall population and the group on tadamidis.
-
There were no differences in adverse effects.
Comments. I am not an amyloid expert but this seems like a clear win for the drug: A death reduction along with recurrent cardiac events, as well as a slower decrease in functional and QOL measures. No difference in adverse effects.
It’s impressive that there was benefit on top of tafamidis. The company plans to seek approval for the drug.
One specific issue noted by the authors was that tafamidis was an allowed background therapy at baseline to which patients were not randomly assigned; therefore, this trial does not allow for a randomized comparison of vutrisiran alone with tafamidis alone.
That’s an important thing to say because I can imagine when a new patient is diagnosed the question will be which drug to start. Given that they both are likely to be quite costly, there will be tension not to use both.
There’s more to learn with this condition, but I can’t seem to find any major issues with this trial.
STOP or NOT Trial
The third trial presented this morning in London, in the first (of 12) hot line trials, was the STOP or NOT trial comparing stopping renin-angiotensin system inhibitors (RASi) 48 hours before a noncardiac surgery vs continuation of the drugs.
It turns out that French guidelines recommend holding RASi before surgery while American Heart Association guidelines say continuation is reasonable and give it IIa level of recommendation.
What’s more, there is an observational study of 4800 patients who had noncardiac surgery, the authors found a strong association of benefit (lower rates of death, stroke, or MI) from holding RASi 24 hours before surgery. The HR was 0.82 and the P = 0.001.
They also found a higher rate of intra-operative hypotension in the continuation group. The authors concluded that there was a strong association of benefit but a large RCT is needed to confirm this finding, because there may be confounding in this non-random comparison. Good on them.
I often hold someone’s blood pressure (BP) meds the day before an atrial fibrillation (AF) ablation so as to avoid a soft BP during and after general anesthesia.
STOP or NOT is a nice trial. Conducted in more than 30 French centers, one group gets RASi held for 48 hours before and the other group continues the drug.
Inclusion criteria included serious surgery lasting more than 2 hours and having an expected hospital stay of at least 3 days.
The primary endpoint was death or major post-op complications within 28 days. The list of post-op complications was quite long, including death, cardiac event, sepsis, respiratory issue, ICU admission, acute kidney injury, or surgical complication.
The results were straightforward.
-
They randomly assigned about 2200 patients in the trial. Half discontinued, half continued.
-
The rate of the primary outcome (death or surgical complication) was 22% in both groups. The CI were pretty tight at 0.87 to 1.19, so, we can say pretty confidently that there was no significant difference.
-
The authors did note a lower rate of intra-operative hypotension, 41% vs 54%, but this did not translate into any serious complications.
The authors concluded that, since there were no differences, both strategies were acceptable. Thus, clinicians have greater flexibility in managing pre-operative RASi based on individual characteristics.
Comments. I would like to commend the authors of this trial, first author Mathieu Legrand. JAMA published the manuscript and the trial answered an important question.
We need more of this kind of work. Everyday practices ought to be tested. The nice thing about the results is that it gives clinicians evidence regarding choices regarding RASI before surgery. In general, there is no compelling need to hold RASi, and in the average patient, I would tend to do the simplest thing and continue the drugs.
But if you have particular concerns, there is no reason not to hold the drugs. There is no harm.
HFpEF
HFpEF is tough to treat. Yes, there are approved medications, sacubitril/valsartan and SGLT2 inhibitors, but the strength of the data is not like it is in HF with reduced EF.
Many patients with HFpEF also have obesity. In recent years, the GLP-1 agonists have shown great promise in inducing weight loss.
Might the GLP-1 agonist drugs also be HF drugs? Last year at ESC, I covered such a trial called STEP-HFpEF. It was a small, semaglutide vs placebo trial of only 529 patients with HFpEF and obesity. The primary endpoints were soft and included KCCQ and weight loss. My column said that semaglutide cleared a low bar. It did because these were subjective endpoints and surely those taking the drug came to know their treatment assignment.
If we were going to use GLP-1 agonists in patients with HFpEF we had to have outcomes data from a big trial.
We don’t quite have that yet, but at ESC, authors from four trials involving semaglutide have published a patient-level meta-analysis taking HFpEF patients with obesity and measuring the standard HF endpoint — CV death and HHF
The four trials included:
-
STEP-HFpEF – semaglutide vs placebo in HFpEF patients with obesity;
-
STEP-HFpEF-DM – semaglutide vs placebo in HFpEF patients with DM and obesity;
-
SELECT – semalutide vs placebo in patients with established atherosclerotic CVD and obesity;
-
FLOW – semaglutide vs placebo in patients with CKD and DM.
SELECT and FLOW did not require obesity but due to the strong relationship between HFpEF and obesity, 99% of patients in all these trials were overweight or obese. Over 80% had obesity defined as body mass index (BMI) > 30.
The meta-analysis therefore includes all patients in the STEP HFpEF trials (about 1100), plus those with HFpEF from the SELECT and FLOW trials. The average age of patients was 64 years.
The primary outcome was CV death or first HF event.
-
The results were that about 17% of patients in the four trials (3700) had a history of HFpEF. About half got semaglutide and the other half got placebo.
-
Semaglutide reduced the risk of CV death or worsening HF events compared with placebo. 5.4% had events in the semaglutide group vs 7.5% in the placebo group; HR 0.69 [95% CI 0.53 to 0.89]; P=0·0045. The absolute ARR was 2.1% so, a number needed to treat of about 47.
-
Breaking down the composite, the authors saw no statistically significant difference in CV death though it was numerically lower in the semaglutide arm (3.1 vs 3.6%).
-
In subgroup analyses, semaglutide did very well in patients with BMI >35 (HR 0.49) vs BMI< 35 (HR 0.96). The P for the interaction was significant. There was no interaction based on use of non-use of the SGLT2 inhibitor.
-
Side effect profiles were similar except for hepatobiliary disorders, which were about double 1.4 vs 0.7 per 100 person years.
Comments. This is fairly strong finding of HF benefit. But first let’s go over the limitations, which the authors did well in the Lancet manuscript.
This is a post-hoc study. Trials had different entry criteria, different follow-up, different endpoints, and even different dosing (lower in the kidney study).
There were low numbers of CV death, which makes it hard to conclude much about the lack of CV death reduction.
A large trial would be better.
But these were similar patients. The HF reduction is clinically significant and statistically robust. It doesn’t matter why it occurs. It is likely through weight loss. The important point is that there were fewer HF events.
Whether this is enough data to warrant an approved indication for HFpEF, I am not sure, but it seems like a reasonable use of meta-analysis. The subgroup finding wherein those with highest BMI had the greater reduction in the primary outcome could be noise, but it also seems highly plausible.
I say this nearly every time I discuss these drugs, but I will say it again, I wish we could implement equally robust weight loss programs with diet and exercise. But the fact remains that few patients achieve GLP-1 agonist-level amounts of weight loss,
I would also remind everyone that when we treat HFpEF, we are usually treating older adults with multiple co-morbid conditions who have been admitted to the hospital for HF. That is a lot different than using the drugs for weight loss in younger otherwise healthy people.
ESC Preview
I wrote a preview column on five trials at ESC.
RESHAPE-HF will be the third trial for transcatheter-edge-to-edge repair of secondary mitral regurgitation in HF. Recall that Mitra FR was negative and COAPT was strongly positive. RESHAPE-HF will be a hugely important addition to the space, though the authors have selected not one but three primary endpoints.
SHAM-PVI will be the first proper placebo-controlled trial of pulmonary vein isolation (PVI) vs a placebo or sham procedure. The problem is the choice of endpoint. The authors chose AF burden. We know PVI reduces AF burden. The question is QOL. They will measure QOL only as a secondary endpoint. Even though I have long advocated for a placebo-controlled trial in AF ablation, I have been on record saying such a trial will likely be positive. The question I have is not whether AF ablation is a total placebo, but rather, how much of the effect is placebo-resistant.
Later this weekend we will also hear the results of the FINEHEART trial of the nonsteroidal mineralocorticoid receptor antagonist finerenone in patients with HF and mildly reduced to preserved EF. The company announced that the trial was positive. At ESC we learn the clinical relevance and statistical robustness of the results.
There are many many more trials coming. I will probably have to have two or three podcasts to cover them all. I will also write up some of these trials as well.
