Jun 06 2025 This Week in Cardiology

Medscape

15 hours ago

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 June 6, 2025, John Mandrola, MD, comments on the following topics: Listener feedback on cardiac sarcoidosis, out-of-hospital cardiac arrest, less is more when it comes to post-stent antiplatelets, lipoprotein(a), and atrial fibrillation in HFpEF.
Dr Riina Kandolin from Helsinki, Finland, writes via email that I should clarify some of my comments on the cardiac sarcoidosis (CS) paper I made last week. The paper I spoke about last week from a primarily Dutch and Minnesota group compared societal recommendations for an ICD vs cardiovascular magnetic resonance imaging (CMR) phenotyping in patients with suspected CS. The key word was 'suspected' CS.
In that study, in the European Heart Journal , CMR phenotyping by (1) no late gadolinium enhancement (LGE) and normal EF, (2) no LGE and abnormal EF, (3) pathology-frequent LGE, and (4) pathology-rare LGE, performed better than professional society recommendations for ventricular tachycardia (VT).
Dr Kandolin pointed me to an incredible Circulation EP paper from their group in Helsinki on 305 patients with either biopsy-proven CS (just under half) or highly suggestive by criteria. Right away, this is a different population because the first study looked at patients with distant sarcoidosis (say the lungs) and were evaluating for cardiac sarcoidosis while the second study, in Circulation EP this year, studied patients with proven CS.
Two main points of difference arise. In patients with true CS, all have pathology-frequent LGE. So it does not distinguish risk. Rather, the Finnish group showed that it was the degree (and somewhat the distribution) of LGE that predicts sudden death and VT.
They found that myocardial LGE making up >9.9% of left ventricular (LV) mass or affecting >6 LV segments may suggest prognostically significant LV involvement and a high risk of sudden cardiac death (SCD).
This analysis and its implications bear similarities to what we do for hypertrophic cardiomyopathy (HCM). Namely, in HCM, there is a risk score that corresponds to a yearly risk of VT. If th 5-year threshold reaches >6% for sudden cardiac death, then an ICD is recommended. ESC recommends a similar risk score prediction for laminopathies as well. ICDs for criteria above a risk threshold.
The Finnish authors therefore suggest in their discussion that their CMR findings could have clinical relevance. Quote:
In patients who have no other guideline-based class I or IIa indications, implantation could be considered if LGE mass is >9.9% by the full-width at half-maximum method or the LGE involves >6 LV segments. Less extensive LGE predicted a 5-year SCD/VT rate <7% with an SCD risk of 0.6 per 100 patient-years, and in these patients other EP studies or close surveillance with repeated risk assessments could be discussed at shared decision-making .
If you care about CS and risk stratification, this is an excellent paper. It's worthy of a bookmark. I appreciate the feedback because I learned a lot.
My one comment — and this is not at all a criticism of the paper (or papers) — but as a doctor I struggle with these risk thresholds. Say in HCM, what if the 5 year-risk is 5% not 6%. Or in this case of CS, the LGE mass is 7% not 9.9%?
The dichotomization of implantable cardioverter-defibrillator (ICD) or no ICD of continuous variables makes me nervous when the outcome is surviving sudden death because you have an ICD. I realize it would be unwise to place ICDs in all patients, including low-risk people, because of ICD harm, but not having an ICD is also a risk.
I don't know the answer and perhaps there never will be an answer to finding the ideal patient for an ICD, but HCM and CS are clearly different from RCT-based heart failure (HF) criteria. In HF, we have entry criteria from trials — albeit old trials. Sort of. We used to be able to say that if you are like a MADIT-2 patient, you have a 30% lower mortality with an ICD compared to no ICD. The modern problem is that medical therapy is a lot better, and that 30% benefit may be a lot less. By the way — that is why if you practice in Europe, you should be participating in PROFID EHRA trial of ICD vs no ICD in ICM.
One practical comment from my experience is that if a CS patient has heart block and you are implanting hardware, it seems wise to strongly consider an ICD if the LGE is even close to the threshold.
The listener feedback is great. I learn a ton from it. Thank you.
The American Heart Association sent me two emails yesterday regarding their efforts to increase awareness of bystander CPR. One was a National Football League player named Justin Reid who is leading an effort; the other email said the Atlanta Falcons are teaming up with the AHA to equip players and coaches with the lifesaving CPR (cardiopulmonary) skills during National CPR and AED (automated external defibrillator) Week.
The second note is that Circulation: Outcomes published a paper today from a group at Mid-America finding that resuscitation practices for out-of-hospital cardiac arrest (OHCA) differ by geographies. Namely, compared to White catchment areas, communities serving Black and Hispanic have lower rates of OHCA survival.
Of course, the causes could be a lot of things, but the research team noted that in Black/Hispanic areas, first responders were less likely to recognize cardiac arrest, police were less likely to respond to a cardiac arrest, and patients were less likely to be defibrillated.
Basically, community-level quality of OHCA was less than what it is White catchment areas. And this may explain the differences in survival.
I think it explains some of the differences. But OHCA survival rates are complicated, and differences could also be due to severity of disease.
My point in highlighting OHCA is that it's one of the highest value interventions in all of medicine. I have spent many hours on this podcast discussing new treatments with marginal benefit. Bystander CPR and an adequate EMS system is hugely valuable. Why?
Because only 10% of patients with OHCA in the US survive. Ventricular fibrillation (VF) without CPR or an automated external defibrillator (AED) occurring out of the hospital is nearly 100% deadly. A VF patient depends on the knowledge and availability of his or her neighbors. Since survival is so bad, any tick up is likely highly cost effective. Bystander CPR and public AEDs are nearly free of risk — it's all benefit.
I know OHCA care isn't as exciting as the newest ablation catheter, or stent, or new drug, but it's surely a lot higher in value than either of these things.
A new tricuspid valve clip or more AEDs in your city? I'd go with the latter every time. So good on the AHA — and the NFL.
The other thing I love about promoting OHCA is that it brings communities together for an obviously great and totally uncontroversial cause.
Is a Year of Dual Antiplatelet Therapy Magical Thinking?
I hesitate to delve into this topic because it is perhaps the most confusing in all of cardiology. But I will, because yet another study finds that 'less is more.'
This seems too simple to state, but the idea is that having a metal cage propping open a coronary lesion is neither a fix nor is it free. Metal in the coronary attracts platelets and then clots can form. Stent thrombosis is a medicalized jargon, but it's a terrible outcome because you go from 0% blockage to 100% blockage. The good news is that the body eventually (and usually) forms a layer around the stent protecting it from the circulating platelets. This we call, and, jargon-wise it is a whopper, endothelialization.
The problem of course is that antiplatelet drugs are also not free — blocking platelets increases the risk of major bleeding in the gut, kidneys, and brain.
So, the quest — and it is a quest — is to find the Goldilocks recipe for preventing thrombotic or clotting events (such as myocardial infarction) with the lowest rate of bleeding. The usual regimen is to use two antiplatelet drugs (aspirin and either clopidogrel or ticagrelor or prasugrel) for a period before switching to single antiplatelet.
Perhaps you can see the complexity already. There are four antiplatelet drugs of varying intensity; and there are near-infinite ways to break up time after the stent, and different types of patients. There's acute coronary syndrome (ACS) vs stable coronary artery disease (CAD) patients; there's high-bleeding risk and not-high-bleeding risk patients; and then there are different stents. You start multiplying and you get about a quadrillion different ways to do antiplatelet drugs.
Today I will tell you about one recent trial — but know that there are many.
The South Korean trial was called 4D-ACS. It was a comparison of two prasugrel-based strategies in about 650 patients who had PCI and stent during an ACS: one group gets one-month of dual antiplatelet therapy (DAPT) (aspirin 100 mg + prasugrel 10 mg, except the prasugrel dose is adjusted to 5 mg for age>75 or body weight < 60 kg vs the control arm: 12 month of DAPT with aspiring 100 mg and prasugrel 5 mg.
The primary endpoint is called NACE. Not MACE , but NACE — or net adverse clinical events: death, myocardial infarction (MI), stroke, ischemia-driven revascularization and bleeding. Bleeding also has 5 different grades. This trial measured types 2-5, so minor bleeding was not included.
The results: NACE occurred in 4.9% of the 1-month DAPT group and 8.8% of the 12-month DAPT group. The trial was a non-inferiority comparison. This is a good use of non-inferiority. Why? Because the short duration is very much less intense than the standard 12-month DAPT regimen. If it was just as good, it would be a win.
Indeed the 4.9% vs 8.8% easily made non-inferiority and in fact with a hazard ratio [HR] of 0.51; 95% CI: 0.27-0.95; P = .034), it also made superiority.
The driver of the lower NACE was a 77% lower rate of bleeding in the short duration DAPT arm: 1.2% vs 5.5%. There were almost no differences in thrombotic events such as MI, stroke, ischemia-driven revascularization.
I cover this study because it is such a clear result. If using prasugrel, a one-month DAPT regimen is surely better than 12-month DAPT.
You would think it's clear now, but it is surely not. First of all, all patients were Korean, and Asians may respond differently to antiplatelets. Second, the study used a type of stent that is not used in the US. Third, what about 1-month DAPT vs 3 months? What about combos of different antiplatelets. Like aspirin/clopidogrel vs clopidogrel.
In general, the trend of late is for shorter courses of DAPT. But my solution in real life is to use one of the most important tools in all of medicine: phone-a-friend. I call my interventional cardiology colleague and ask what they think. They've done the stent and a bifurcation stent in the proximal LAD is going to be different from a mid-circumflex lesion.
Perhaps I can make these general statements:
ACS patients generally require longer DAPT duration (12 months) due to higher thrombotic risk
Non-ACS patients most likely can safely use shorter DAPT durations (6 months for DES.
Patients with high bleeding risk may benefit from abbreviated DAPT (1-3 months) regardless of presentation
P2Y12 monotherapy after initial DAPT period shows promise for reducing bleeding while maintaining efficacy
An individual risk assessment may prove helpful in guiding duration and intensity of DAPT.
Final comment: if you are having an ST-elevation myocardial infarction (STEMI), you want a stent done fast. If you have chronic CAD, medical therapy first avoids the entire issue of deciding on the combination of antiplatelets. Sadly, this doesn't happen often enough in many places in the US.
JAMA Cardiology has published an interesting study looking at the value (or lack of value) in adding Lp(a) into the new PREVENT equation.
The first thing to say is that I covered the AHA's PREVENT equation in August of last year. Proponents of PREVENT say it's better than the standard PCE — which outputs 10-year risk of atherosclerotic events.
Of course, the proponents of PREVENT don't just say it's better. PREVENT has been validated in observational datasets.
PREVENT differs from the pooled cohort equation because it incorporates kidney function, statin use, social determinants of health and removes the race category. It also starts at a younger age.
The most provocative aspect of PREVENT is that it simultaneously is felt to be more accurate, but it lowers estimates of CV risk and lowers the number of statin-eligible people. Gulp. Double gulp.
Anyways, the question of the JAMA Cardiology study was whether the PREVENT equation can be made better by adding Lp(a) — which as you probably know is mostly genetically determined and, when elevated, is strongly associated with higher risk of CV events.
Since race is removed from PREVENT and race can bear on Lp(a), adding it could be very important.
Here is that they did. Two databases — UK biobank and MESA. This was big, big data. Both databases have long-term follow-up.
Individuals in the database were put into 4 categories: low risk (5%), borderline (5%-7.5%), intermediate (7.5%-20%), and high risk (> 20%)
Then they measured 10-year event rates and correlated it with Lp(a) levels overall and by risk category. Of course they do adjustments for age and sex.
The main endpoint is the net reclassification improvement or NRI, which is complicated. Let me try to explain:
NRI measures how well a new risk prediction model reclassifies people compared to an old model, but it does this by tracking movements in both directions and netting them out.
The NRI splits reclassification into events (people who actually had the outcome) and non-events (people who didn't). For each group, it calculates:
Proportion who moved up in risk categories minus the proportion who moved down
This gives you a net movement figure for each group.
The problem comes in relative vs absolute differences. You see this in the NRI studies of coronary artery calcium (CAC) where, when you consider absolute numbers, more people don't have events and are actually misclassified using NRI.
I asked the AI tool Claude about this — because Claude helps — and Claude calls this 'The Conceptual Trap,' and I think it's worth talking about.
Claude says that people often interpret NRI as "20% improvement in classification" when it's really "20% net improvement after accounting for movements in both directions." The absolute amount of reclassification — which tells you how much the models actually disagree — gets hidden in this netting process.
In this study, they measured NRI both categorically — that is, how many moved into a different risk category based on Lp(a) (ie, going from borderline risk to intermediate risk) — and they also measured the NRI category-free (which is simply whether the patient moved up or down in risk at all).
They differed — a lot.
The category-free NRI for atherosclerotic events using Lp(a) was about .06 (or 6%). Now, the categorical NRI was 10 times lower — at 0.6%
The way I would try to translate that is to say that adding Lp(a) reclassifies about 6% of people on a net basis for CV risk, but it is far less valuable for moving people between low, borderline, intermediate and high-risk categories. Here it was less than 1%.
Now, the study did something else that is perhaps somewhat useful. They looked at subgroups of people who may get more reclassification. They found that adding Lp(a) resulted in slightly more reclassification for borderline risk and low risk people. But I have to say the differences are quite modest. For instance, the NRI for low-risk CHD prediction was 10% vs 7.5% for intermediate risk. 10% vs 7.5% — it doesn't seem like a big difference.
This is a nice effort. Adding Lp(a) helps a little with prediction. A low-risk person who has an Lp(a) may want to take statins. A borderline person with a low Lp(a) may pass.
But here's the thing, and I feel it's the same with CAC. I've seen patients for nearly 30 years in middle America, and the vast majority, perhaps 90% of people cannot conceptualize the difference between 7% and 9% vs 13% risk of 10-year events.
I've had lipids done recently and I am borderline risk. And I can't really conceptualize 10-year risk of a nonfatal event. Maybe you can.
Two reasons I struggle. One is that philosophically, I think much of disease and health is good and bad luck. As in…stuff happens. The second reason I struggle with 10-year risk prediction is relative to Ukraine attacking Russia's planes and the threat of World War III: how much worry should I place in whether my nonfatal CV risk is 7% without Lp(a) vs 9% with Lp(a)?
I come back to my bottom-line when it comes to prevention: there are maximizers who want to do everything. For them, Lp(a) seems useless because they are already on aggressive lipid lowering. There are minimizers who aren't taking pills regardless. And for these patients, Lp(a) is also useless.
So, for the few who sweat the details, fine — add Lp(a) to the mix.
Publishing in EuroPace , a group of prominent researchers re-analyzed the TOPCAT Americas trial looking at the role of AF in predicting bad outcomes in patients with heart failure with preserved ejection fraction (HFpEF).
Recall that TOPCAT is one of the most important trials in cardiology. Sadly, it was marred by data irregularities from Russia and Georgia. TOPCAT was spironolactone vs placebo in HFpEF. The hazard ratio (HR) was 0.89 for the composite primary of death, cardiac arrest, or heart failure hospitalization (HHF). The 11% reduction did not reach statistical significance. But all-cause death was 22% in the Americas and only 8% in Russia/Georgia, and the drug had little effect in that region. Further metabolic studies found that a third of patients supposedly randomized to spironolactone had undetectable metabolites. Excluding Russia and Georgia, TOPCAT was positive with a HR of 0.82 and a confidence interval of 0.69-0.98. I think this is a really important point.
Ok, now to the EuroPace paper. The problem with studying HFpEF is it is diverse condition — unlike HFrEF. The authors, mostly EP docs, were interested in AF effects in HFpEF. Specifically, whether AF is simply a marker for advanced CV disease or is it an independent risk factor.
They made two groups:
Study patients who had AF before or at study entry. This was the 'any AF' group.
Study patients with ongoing AF who had it at entry.
Then they propensity-matched patients with no AF and looked at outcomes.
I know, it's pretty easy to predict what will happen. Before I tell you the results, I hope you are thinking: small, hyperdynamic, noncompliant left ventricles are not going to do well with losing the atrial kick when AF occurs.
The primary outcome of these comparisons was CV mortality.
About 580 patients in TOPCAT who had any AF and 400 had ongoing AF. And they found that: Any AF was associated with a statistically significant increase in CV hospitalization, HHF, and progression of HF. Any AF, however, was not associated with an increased risk of sudden death.
Ongoing AF was associated significantly with CVD, pump failure death, CV hospitalizations, and HHF.
It was interesting to me that neither any AF or ongoing AF was not strongly associated with sudden death. AF seemed to associate mostly with progression of HF or pump failure.
This paper has a lot of complicated comparisons, but I think it can be summarized as AF is a bad thing to have with underlying HFpEF.
The question of course, the clinical question, is how to modify this problem with therapy. My first recommendation is to phone-a-friend: your EP colleague. I can't give you a generic right answer because there isn't one. HFpEF patients often have serious comorbidities that have to be considered, things like polypharmacy and CKD and valvular HD and frailty. Maybe antiarrhythmic drugs are the right answer. Maybe ablation. And underused in some places is the 'pace and ablate' strategy.
We see a lot more HFpEF these days, because people live longer and with more chronic illness. When AF occurs, it's serious. Take Mr. Rogers' advice and realize that we all need helpers in our lives, and please: call your EP friends.