Oct 16 2026 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, Apple Podcasts, Spotify, or your preferred podcast provider. This podcast is intended only for healthcare professionals.Light Chain Amyloidosis — A Potentially Helpful DrugAnselamimab Does Not Reduce Light-Chain Amyloidosis DeathsJust to briefly review, Light Chain Amyloidosis is a progressive systemic disease caused by plasma cell dyscrasia in the bone marrow, resulting in overproduction of immunoglobulin light chains that aggregate to form insoluble fibrils, which deposit systemically, leading to clinically significant morbidity and mortality.The treatment has always been multipronged with chemotherapy to suppress plasma cells expansion. These drugs have proven somewhat helpful, though the heart is slow to respond because it takes time to remove the kappa and lambda fibrils that get deposited. Anselamimab is a monoclonal antibody directed against human kappa light chains when they misfold.At ESC and in Circulation is a research letter describing results of trials called the CARES trial, which is actually placebo-controlled phase 3 trials in 406 patients enrolled between 2020 and 2023. Treatment with anselamimab did not improve all-cause mortality (ACM) and cardiovascular hospitalizations in the overall cohort; however, there was a significant isotype-by-treatment interaction, with a 62% reduction in all-cause mortality and a 71% reduction in cardiovascular hospitalizations in the kappa isotype group compared with placebo, with no benefit observed in patients with the lambda isotype.The problem was that the kappa subgroup represented only 72 of 400 patients, or 18%. In addition to better outcomes, patients who had Kappa amyloid also showed better echo parameters — which strengthens the subgroup analysis. In addition, there is strong biologic plausibility, given that anselamimab was generated using a kappa isotype immunogen and thus may be most effective against kappa fibril aggregates.I looked up the regulatory issues of anselamimab and it is not yet approved anywhere. The FDA granted a Fast Track Designation in June 2021, and it also has Orphan Drug Designation from the FDA and European Commission. It’s going to be tough road regulatory wise because it’s small subgroup in an overall negative trial, though there are biologic reasons to think it would work in the kappa subgroups of AL amyloidosis.I enjoyed reading the study and thinking about drug development. More than a decade ago, I lost a cycling teammate to light chain amyloidosis. If my friend had the condition now, he might have survived. That’s neat to consider.Aficamten in Nonobstructive Hypertrophic Cardiomyopathy: The ACACIA trial Before I say one word, I had no idea what acacia was. It turns out that Cytogenetics, the company that makes aficamten, names their trials after trees: REDWOOD, SEQUOIA, MAPLE, CEDAR, and now the ACACIA-HCM trial. That has nothing to do with critical appraisal but it is something I did not know. Names have no bearing on anything as evidenced by the best name ever for a pacemaker, the VIGOR, which was literally the worst pacemaker I have ever used. Now to the medical stuff: Myosin inhibitors have shown benefit in patients with obstructive hypertrophic cardiomyopathy (HCM), but a large number of patients with HCM have no left ventricular outflow tract (LVOT) obstruction. These patients also have symptoms, but the first myosin inhibiter mavacamten failed to reach statistical significance in the placebo-controlled ODYSSEY-HCM trial of patients with HCM and no obstruction, though the two components of the primary endpoint peak VO2 max and Kansas City Cardiomyopathy Questionnaire (KCCQ) were lower in the mavacamten arm, but the P-values were.06 and.07 respectively. So it was close. At ESC, we heard results of the aficamten placebo-controlled trial called ACACIA-HCM of 258 patients with symptomatic nonobstructive HCM. Patients were 55 years old, slightly more than half female, and two-thirds with a likely pathogenic variant. At 36 weeks, the change in the KCCQ was 11.4 points in the aficamten group and 8.4 points in the placebo group (the 3.0 point difference just met statistical significance at P =.02). The mean change in the peak oxygen uptake at week 36 was 0.64 mL/kg of body weight per minute in the aficamten group and basically zero change in the placebo group. The small absolute difference met statistical significance (P =.003).Serious adverse events were reported in 20.2% in the aficamten group and 14.7% in the placebo group.Early discontinuation of aficamten or placebo because of adverse events occurred 7% vs 1.9% in the placebo group. Heart failure occurred in 4.3% vs 0.4% in the placebo group.This ACACIA trial was published in NEJM, but in a second simultaneous publication in Circulation, the authors of the trial published a series of secondary outcomes and responder analysis. In this paper, the notable findings were:a greater proportion of patients treated with aficamten vs placebo experienced improvements in symptom burden (71% vs 53%), peak exercise capacity (45% vs 37%), left atrial volume index (31% vs 23%), > 10% improvement in septal early diastolic mitral annular velocity (septal e’) (51% vs 26%) 50% decrease in NT-proBNP (63% vs 5%) A partial or complete clinical response (at least 3 outcomes) was achieved in 53% of patients on aficamten vs 14% on placebo (P < .001)My Comments and Status of Aficamten In December 2025, the FDA approved aficamten (brand name Myqorzo) for adults with symptomatic obstructive hypertrophic cardiomyopathy. The approval based on the SEQUOIA-HCM trial.As with mavacamten (Camzyos), the label carries a boxed warning about heart failure risk, and the drug is only distributed through a risk evaluation and mitigation strategy, or REMS program. A REMS program means that prescribers and pharmacies have to be certified and echo monitoring has to be done. In other words, a lot more paperwork. A separate application to add the MAPLE-HCM data (aficamten vs. metoprolol) to the label for obstructive HCM was accepted by the FDA with a decision date next month. The very short term MAPLE-HCM trial found aficamten superior to metoprolol for a primary endpoint of change in peak VO2 max at week 24. In an October 12 press release this week, Cytokinetics said it plans to submit a supplemental New Drug Application to the FDA this quarter for nonobstructive HCM. An approval decision would normally follow 6 to 10 months after a filing is accepted, depending on whether it gets priority or standard review.You might wonder why one myosin inhibitor was effective and the other was not in nonobstructive HCM. I did — until I looked at effects in the trials. Upon inspection, it seems the effects are pretty similar — just on opposite sides of the 0.05 significance level. In ODYSSEY-HCM (mavacamten): neither peak VO2 (difference 0.47 mL/kg/min; P = 0.07) nor KCCQ-CSS (difference 2.7 points; P =.06) improved significantly compared with placebo. In ACACIA-HCM (aficamten): 3.0 points of improvement on the KCCQ and 0.67 mL/kg/min of peak oxygen uptake vs placebo, and both primary endpoints were met. A 2.7- vs. 3.0-point KCCQ difference and a 0.47 vs. 0.67 peak VO ₂ difference are not very different. Any real advantage for aficamten is probably modest, and some of the “fail vs win” split could reflect trial-to-trial variability. Consider that in the Circulation paper in the binary responder analysis, the peak VO ₂ responder endpoint was not significant (45% vs. 37%, P =.1), even though the continuous peak VO ₂ analysis was.So this could be that both drugs were effective and fell on either side of that significance level.Or aficamten could be a slightly better drug based on its shorter half-life and better ability to up-titrate and stop if left ventricular ejection fraction (LVEF) drops. There are also fewer drug–drug interactions with aficamten. The thing that seems reasonable to conclude is that both drugs affect the biology of HCM, as echo and biomarkers do go in the right direction. The effect size vs placebo, which also had a remarkably positive effect in both trials, was less than in obstructive HCM.I am not an HCM expert, just a Neutral Martian looking in, but I am struck by the short-term nature of these trials. Both drugs work by creating negative inotropy.I mean, LVEF fell below 50% in 10.5% of aficamten patients vs. 0.8% on placebo. And serious heart-failure events occurred in 4.3% vs. 0.4% in the Circulation paper. While there is no evidence that myosin inhibitors are proarrhythmic, I do worry about harder outcomes in a year or 2 years or 3 years. We don’t know, and I don’t think we should assume a drug that makes people feel better translates to better or even neutral survival. Open label trials like the FOREST-HCM extension study in European Heart Journal will not answer the question. It’s not an easy question to answer because how do you keep randomization blinded if you know one arm is doing better on quality-of-life measures? Perhaps you could argue a 3-point KCCQ difference and 0.6 diff in peak VO2 is marginal enough to stay blinded for 1-2 years. I am just not sure. The other issue is cost. Is the 3 point KCCQ and 0.6 delta in peak V02 worth the huge price tag? Aficamten launched with a wholesale price of about $108,000 per year. Gosh that is a lot for small differences. A back-of-the-envelope calculation from the Circulation paper follows:New York Heart Association (NYHA) improvement was 41.9% vs. 27.8%, which works out to treating about 7 patients for one extra person to improve a functional class. At list price, that’s roughly $750,000 a year per additional patient whose NYHA class improves, before monitoring costs. All told, I am not sold, are you? Microaxial Flow Pump Cost Analysis of the DanGer Shock Trial JACC has published a research letter from the DanGer Shock trial researchers on hospital costs of the microaxial flow pump in patients with cardiogenic shock. This was a prespecified secondary analysis of all 355 randomized patients in the DanGer Shock trial analyzed by intention to treat (ITT). Patient level resource use was collected at 14 centers in Denmark, Germany and the UK. Prices were calculated in Danish units.Before I tell you about costs, let’s review DanGer Shock, one of the most important trials of 2024. Recall that before DanGer Shock, there was zero compelling RCT-level evidence that the microaxial flow pump (brand name Impella) improved outcomes in myocardial infarction (MI)-related cardiogenic shock (CS). Danger shock trialists were careful about who got into this trial; patients had to be Goldilocks-level sick to be randomized. The exclusion took out patients too sick to benefit and too well benefit. It took a decade to randomize 355 patients. The primary endpoint of death from any cause occurred in 45.8% in the flow pump arm vs 58.5% in the standard therapy arm. The 12.7% absolute risk reduction translated to a 26% relative risk reduction (hazard ratio, 0.74; 95% CI, 0.55-0.99; P =.04). Pause there, because these are truly remarkable results. The absolute risk reduction (ARR) is 12.7 %, which is a number needed to treat (NNT) to save a life of 8 patients. But because the number of events were so close, the 95% confidence intervals come shocking close to 1.0 or no effect. Thus, the P value was just.04, meaning if just one or two outcomes switched groups the trial would have been nonsignificant. It’s an extremely fragile result, but the way I digest it, and it may be wrong, is that if you spend a decade and actively select patients with just the right amount of illness, such that LV support during CS might help, you could get benefit. But they screened 1200 patients to enroll 360. Yet the highly selective nature of what I call “Goldilocks-like” CS patients is not a top-line finding. The top-line finding when the trial is promoted — sometimes over free burritos in the cath lab — is that the IMPELLA saves lives. Now to the cost analysis: In 2025 US dollars, the cost of Impella + hospitalization was $96,000 vs $52,000 in the standard care arm. The increase was about $44,000. That was about the same difference — about $43,000 higher at 180 days. The higher costs break on three aspects: Mechanical circulatory support devices and procedures: +$23,468. This is about 53% of the gap, and it is mostly the pump itself.Hospital and ICU stay: +$18,217, or about 42%. Median ICU stay was 6 days with Impella vs 3 days with standard care. Median ward stay was 3 days vs 1 day.Complications (+$1,409), ICU drugs (+$704) and respiratory/monitoring care (+$89) account for the small remainder.If you consider the DanGer Shock trial a true positive and not false positive the ARR of 12% and NNT of 8 patients, this works out to approximately $330,000 in extra costs per death averted at 6 months. The 12.7% ARR is a point estimate. The 95% confidence intervals go from a 2.4% difference to 23% difference. That’s an NNT of 42 to 4.3.If the extra costs are about $41,000, then the costs per life saved would be 42 x $41,000 or $1.7 million to $176,000.A couple positives were that long-term benefit, presented last year at ESC 2025, showed a 16.3% better death rate at up to 10 years. So that makes the cost a little lower per life saved — a little lower. Some of the caveats in translating this evidence: US costs are likely going to be higher. The population is highly selected to do well. Impella use in less idea patients would surely raise the cost — by a lot. There is a real safety trade-off, with more bleeding, limb ischemia, and dialysis in the flow pump arm. They show up as only a modest $1400 dollars more, but the clinical implication may be substantial, since the trial counts only deaths as a primary endpoint and these could add morbidity. Though the counter is that you have to be alive to suffer safety events. My friend (and extremely thoughtful academic) Dr Sanket Dhruva and others wrote the editorial and focused on the problem of flow pump use in non–DanGer Shock-like patients. Not only would this increase costs, but outside of that trial, we have no reason to believe it would be money well spent. In sum, I love the DanGer Shock trial, because now at least we have evidence where there was none. We know that if you work as hard as Professor Møller to select just the right patient with CS from MI, there may be a lifesaving benefit. And with the cost analysis, clinicians and administrators alike are forced to make really hard cost decisions. More on Evolocumab at ESC: Numerous Vesalius Trial Substudies The main VESALIUS-CV trial was presented at AHA last year and published in NEJM. It was a first for the PCSK9 inhibitor class as the two big trials before, FOURIER and ODYSSEY, had enrolled patients after MI, and from a prevention standpoint, that is a bit late. VESALIUS tested the drug in 12,500 patients with qualifying atherosclerosis or high-risk diabetes, no but had no previous MI or stroke, and LDL-C of at least 90 mg/dL. In the main trial, after 4.6 years, evolocumab reduced the 3-point composite of first coronary heart disease death, MI, or ischemic stroke by 25% and the corresponding 4-point endpoint, which included those 3 things plus ischemia-driven revascularization, by 19%. I covered this trial three times on TWIC: after than main trial on Nov. 14, 2025 and again on April 24, 2026 and May 29, 2026. The issue with the main trial was a nuance in statistical reporting. That is the main 3-point major adverse cardiovascular events (MACE) endpoint of first coronary death, MI, stroke was positive as well as the 4-point endpoint adding ischemia-driven revascularization.However, the authors, in their effort to avoid false positive declarations from multiple comparisons of many secondary endpoints, required hierarchical testing wherein each test spends alpha, that is then passed down the chain of testing. Said another way, coronary death sat above CV death and all-cause death. But....but....coronary heart death was not significant (HR, 0.89; 95% CI, 0.68–1.16; P = 0.39). So the chain broke and you could not consider subsequent comparisons as significant. But the CV death number and all-cause death numbers were indeed favorable: CV death (HR, 0.79) and all-cause death (HR, 0.80) are formally exploratory, even though both were prespecified and both look strong taken together.The companion analysis presented at ESC, not as a hotline but as a late-breaking clinical trial, and now published in Circulation, reports specifically on death outcomes. As well as the timing of events, which turns out to be sort of interesting. Overall, the paper reports 434 deaths with evolocumab and 539 with placebo, with 5-year Kaplan-Meier estimates of 7.9% and 9.7%, respectively (HR, 0.80; 95% CI, 0.70–0.91; nominal P =.0005). The 1.8-percentage-point difference corresponds to an approximately 5-year number needed to treat of 56 patients to prevent 1 death.There was consistency of benefit: for CV death 2.8% vs 3.6% (HR, 0.79; 95% CI, 0.64–0.98), non-CV death 4.2% vs 5.0% (HR, 0.85; 95% CI, 0.71–1.01), and deaths of undetermined cause 1.1% vs 1.4% (HR, 0.64; 95% CI, 0.45–0.92) Results were consistent regardless of age, sex, region, race, qualifying atherosclerosis or high-risk diabetes, baseline LDL-C, or background lipid therapy. As for timing, the mortality curves were similar for 18 months, then there was a 27% lower hazard thereafter with a significant P for interaction, which makes sense in how the drug would work on plaque. The editorialists note that this may be why no mortality benefit was seen in the 2-year FOURIER trial — for example, not enough time. The editorialists Drs Fatima Rodriguez and Michael Shapiro have a very interesting take on the nuance of VESALIUS. They cleverly write that the VESALIUS findings wherein nonsignificance of coronary death stops the process of assessing CV death or all-cause death, “require statistical discipline but not statistical amnesia.” They go on to write:VESALIUS-CV was designed and powered for cardiovascular composite end points, not mortality. Coronary heart disease death, which preceded the other mortality end points in the prespecified hierarchy, was not reduced (hazard ratio, 0.89; P = 0.39). Formal sequential testing therefore stopped, and the subsequent mortality P values are nominal and the findings nonconfirmatory. Although that limitation governs what can be claimed, it does not erase what was observed. There were 105 fewer deaths in a randomized, double-blind trial. Multiplicity rules properly protect against false-positive declarations but do not require clinicians to pretend that a large, directionally coherent mortality difference is clinically uninteresting. The appropriate conclusion is neither proof nor dismissal, but a compelling survival signal that deserves to influence how the parent trial is understood. They also add that the parent trial exposes a persistent flaw in clinical practice, namely that we wait for an MI or stroke before becoming aggressive about risk factor control. They write:The [MI] event transforms treatment urgency, insurance approval, clinician behavior, and patient acceptance. But it also marks the failure of prevention. They point out that VESALIUS patients were quite high risk, with the placebo mortality rate at 5 years at 10%. Which makes sense because the patients had established coronary disease or high-risk diabetes and a still high LDL despite therapy. The most provocative finding considered from the authors of the Circulation paper was the notion that nonfatal MI could lead to later non-CV death. This came from modeling where they purported to show that 78% of the effect on non-CV death appeared to be related to reduction in events such as MI. They ask us to think of the cascade where an MI event may not be deadly then at that moment, but could lead to other non-CV deaths later like falls, pneumonias, frailty.My Comments I don’t know about this analysis. As for the “MI prevention prevents non-CV death,” it is a nice story — but the numbers don’t add up. The non-CV gap was 39 deaths — 78% of that is roughly 30 deaths. If preventing a first event lowers later non-CV death risk by about 4 to 14 percentage points, explaining 30 deaths would take several hundred fewer first events. There were only 80 fewer MIs in the main trial. That accounts for about 6 fewer non-CV deaths, not 30. I think my original take on this trial stands. I think the drug works, it reduces LDL-C a lot and this results in a real but modest 1.8% reduction in a 3-point MACE endpoint, mostly driven by lower MI. Despite enrolling a group of people with established coronary disease or high-risk diabetes and a very high LDL-C, the trial was underpowered to detect a difference in coronary death. There were only 222 coronary deaths and if you wanted to show a reduction in death with sufficient power you would need more patients and/or longer follow-up. I am compelled by the effects over time in VESALIUS, as you’d expect LDL-C lowering to exert its effect slowly. As was seen in all the statin trials. But I don’t see any way to make much of the non-CV death and all-cause death when coronary deaths were not different. I am glad that it wasn’t numerically higher like in FOURIER, but this analysis does not convince me that the drug reduces mortality even in patients selected for benefit — because of those very high LDL-C numbers. To me, if we had a VESALIUS-like patient in our clinic with that LDL-C, why wouldn’t we change to a high-dose, high-potency statin (less than 70% in VESALIUS were on high-intensity statin) and/or add ezetimibe (only 20% in VESALIUS were on ezetimibe), and/or add bempedoic acid. These may be less costly easier ways to lower LDL-C.There was also another VESALIUS substudy from ESC, published in JACC, and it looked at total events rather than first events. This analysis largely upheld the primary analysis.