The 2025 European Society of Cardiology congress (ESC) will be in Madrid, Spain. Here are five topics from six trials (among many) that piqued my interest.
DIGIT-HF
I have long believed that cardiac glycosides have been unfairly maligned. Observational studies have shown an association between digoxin use and higher mortality. Despite the statistical adjustments, these nonrandom comparisons probably are marred by selection bias wherein sicker patients receive digoxin.
People forget that in the DIG trial in patients with heart failure with reduced ejection fraction (HFrEF), digoxin did not reduce the primary endpoint of mortality, but it did lower hospitalizations due to heart failure — to a similar degree as SGLT2 inhibitors.
This is why I am excited to hear the results of the DIGIT-HF trial, a double-blind, placebo-controlled trial of digitoxin in patients with HFrEF. The primary endpoint is a composite of overall mortality and first heart failure hospitalization. Notably, the trial uses digitoxin, which has more favorable pharmacologic properties than digoxin. Digitoxin is no longer available in the US; maybe that will change after the trial.
The authors have published a rationale, baseline characteristics, and drug dosing papers, as well as an amendment letter. It looks to be a well-run trial.
But it will be hard to show a 20% reduction in the primary endpoint on top of other therapies. Yet, even if not significant for the primary, I predict that there will be many secondary endpoints and subgroup analyses that may re-ignite enthusiasm for a mostly forgotten drug class. If nothing else, a lack of harm signal should stop the dig-as-poison thinking.
Two Trials on Beta-Blockers After MI
Use of beta-blockers after myocardial infarction (MI) remains a quality measure. This is curious because the underlying evidence stems from trials in the 1980s, well before urgent revascularization transformed the care of patients with MI. The contemporary REDUCE-AMI trial found no difference in death or MI with long-term use of beta-blockers after MI in patients who have preserved ejection fraction. The ABYSS trial of continuing or interrupting beta-blockers after MI favored the continuation arm, but rates of death and MI were similar in the two groups.
ESC will feature two large trials testing beta-blockers in patients after MI. The Spanish REBOOT-CNIC trial enrolled more than 8000 MI patients with a left ventricular ejection fraction (LVEF) > 40% to beta-blocker vs no beta-blocker. The pragmatic trial will be powered to detect a 20% relative risk reduction in the composite primary endpoint all-cause death, nonfatal reinfarction, or HF hospitalization over a median follow-up period of 2.75 years.
BETAMI-DANBLOCK began as two separate trials in Norway and Denmark, respectively. The trial procedures were similar, and from the beginning, the authors planned on pooling data, so after the COVID pandemic slowed enrollment, they combined the trials into one. A design-and-rationale publication last year noted that more than 4000 patients had been enrolled thus far. The composite primary endpoint differs from REBOOT-CNIC in that it includes more outcomes: MI, heart failure, coronary revascularization, ischemic stroke, all-cause mortality, malignant ventricular arrhythmia, or resuscitated cardiac arrest.
Interpretation of these trials could be easy or difficult — easy if they both return null primary endpoints. In that case, we would have three trials with null results, leading to a reversal in routine use of beta-blockers after MI. Interpretation will be a lot harder if one trial is null and one is positive. If that happens, we will be forced to delve deep into the nuances and then contemplate the notion of a positive trial meta-analyzed with null trials to output a null result. Which to believe: a single well-done trial or a meta-analysis?
No matter the result, the meta-lesson here is that most evidence requires an expiration date.
Oral Anticoagulation and Aspirin
I am pretty sure we cause substantial harm when using aspirin and oral anticoagulation (OAC) together. The scenario is common: Either a patient with atrial fibrillation (AF) on OAC gets diagnosed with atherosclerotic vascular disease, or a patient with chronic coronary disease on aspirin develops AF. Little evidence informs combined use of the two antithrombotic agents.
AQUATIC is a French placebo-controlled trial of OAC plus aspirin vs OAC plus placebo in patients with an indication for long-term anticoagulation who also have stable coronary artery disease. The primary endpoint is a composite of cardiovascular death, MI, stroke, systemic embolism, coronary revascularization, or acute limb ischemia. Major bleeding will be the safety endpoint.
The previous trial in this area was AFIRE, which enrolled more than 2200 patients from Japan and randomly assigned them to either rivaroxaban alone or rivaroxaban plus an antiplatelet. Rivaroxaban alone met criteria for noninferiority for the primary efficacy endpoint and superiority for the safety endpoint of bleeding. AFIRE had limited external validity because rivaroxaban was used at a reduced dose and about a third of patients with coronary disease did not receive a coronary stent.
I celebrate having more evidence in this area, but my worry is low event rates. The authors have powered the superiority trial for a 25% reduction in the primary endpoint. That seems like a big difference for adding aspirin to OAC in patients with stable disease. I worry that if the results are null and the confidence intervals wide, we will be left with uncertainty. That said, I predict that we see a significant safety signal (increased bleeding) for the combination therapy.
Stopping Anticoagulation After AF Ablation: The ALONE-AF Trial
Not once but multiple times per day, patients ask me whether they can stop taking OAC after successful AF ablation — because the reason for OAC was AF, which is gone.
This may be the biggest question in everyday electrophysiology practice. I believe we overtreat many patients with anticoagulation after AF ablation.
That is why I am excited to hear results of the ALONE-AF trial, comparing continuing or stopping OAC in patients who have no documented AF recurrence for a year after ablation. The key inclusion criteria are that patients have to have at least one risk factor for stroke and no documented AF on (at least two) 24- to 72-hour Holter monitors. Some may argue about the fact that intermittent monitoring misses more AF than implantable loop recorders. I would counter that intermittent monitoring is a positive because it enhances external validity of the trial.
The primary outcome is a composite of bad things: stroke systemic embolism and major bleeding. The worry with this trial is low event rates and wide confidence intervals. The thing about postablation patients is that not much bad happens to them, either thrombotic or bleeding wise. If the trial is null but uncertainty is high, the danger would be assuming that it supports stopping anticoagulation.
Potential Surprise of ESC: The REFINE-ICD Trial
When we place a primary prevention implantable cardioverter defibrillator (ICD), the most likely outcome is no benefit. Modern therapeutics have exacerbated the problem of patients with ICDs not needing therapy. What’s more, the vast majority of cardiac arrest occurs in patients with higher than “ICD-eligible” LVEFs.
One of my great career disappointments has been an utter failure to risk-stratify patients into groups that may benefit more from an ICD. Our ICD problem can be analogized to treating every sore red throat with antibiotics, rather than sorting out which ones are due to bacterial infection.
The potential surprise of the ESC meeting is the REFINE-ICD study, set to present in the third hotline session. Primary investigator Derek Exner, MD, MPH, from Calgary, has been publishing for nearly 20 years on the use of ECG parameters, heart rate turbulence, and t-wave alternans (TWA) to risk-stratify patients after MI.
REFINE-ICD will enroll post-MI patients with abnormal heart rate turbulence and TWA and an EF between 35% and 50% to an ICD or medical therapy. The primary outcome is mortality.
Wouldn’t this be something: Two noninvasive, inexpensive, and easy-to-measure parameters might be able to identify a high-risk substrate in patients who could benefit from an ICD. Medical science is hard, so my Bayesian priors are pessimistic. But what a shocker it would be. And even if it is negative, we should all buy Derek an espresso for a great effort.
John Mandrola practices cardiac electrophysiology in Louisville, Kentucky, and is a writer and podcaster for Medscape. He espouses a conservative approach to medical practice. He participates in clinical research and writes often about the state of medical evidence.