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 January 24, 2025, John Mandrola, MD, comments on the following news and features stories. Familial hypercholesterolemia (FH), cardiac arrest therapy, Factor XI inhibition, and more on tricuspid valve interventions.
I first want to say thank you to the University of Michigan fellows for inviting me to give cardiovascular (CV) rounds. It is an honor to be asked to speak, but it is an especially big honor to get a fellows’ invite. I spoke on atrial fibrillation (AF) treatment, with a focus on areas of uncertainty. It was also great to see my friend Venk Murthy — a true academic.
Familial Hypercholesterolemia
When I started in cardiology many years ago, getting the diagnosis of FH was terrible. Early coronary artery (CAD) and CV death were highly likely.
Well, things have changed for the better. So suggests yet another neat observational comparison study using the national Danish Health Registry.
First author, Jacob Reeh, and co-authors from the Gentoffe hospital in Copenhagen, used the registry to compare about 11,000 patients with and 9 million without FH over a 45-year period between 1978 and 2021.
During follow-up, 27% of those with and without FH died.
For individuals with FH compared with those without, the mean age at death was 22 years younger in 1978 (50 years vs. 72 years, P < .001). But in 2021, the corresponding difference in mean age at death was only 1 year (78 years vs. 79 years, P = .16).
When you look at a graph with age at death on the y-axis and years on the x-axis, the two curves are separate in 1978 but gradually come together by 2021. The relationship ship was the same in men or women.
However, the age of CAD diagnosis did not completely come together.
For individuals with FH compared with those without, the mean age at CAD diagnosis was 20 years younger in 1978 (49 years vs. 70 years, P < .001). But in 2021, the corresponding difference in mean age was reduced to 7 years (61 years vs. 68 years, P < .001).
Similar patterns were seen in acute coronary syndrome (ACS) and angina.
Comments. Here we have another example of a good use of observational data; that is, describing trends over time. Or as we say temporal trends.
It also describes good news. A once terrible disease has now been normalized.
What’s more, it is fun to speculate on why the diagnosis of CAD, angina, and ACS have not completely normalized, while age at death has. The authors do something I haven’t seen before but really like.
In the discussion section where they speculate, they write, “Importantly, the ideas presented below should be interpreted cautiously as we are not presenting data to support such speculations.”
They actually write that sentence twice — for emphasis. I don’t know why they think it is controversial, because lots of diseases have declining incidence. For example, I covered declining stroke rates in patients with AF last week. And I have mentioned the declining incidence of sudden death in patients with heart failure (HF).
Here, I see it as likely that the diagnosis of FH will increase, but as treatments increase, like statin use, we would see death rates decline.
It’s a great example of how domains affect treatment effect. What do I mean? Well, one domain is the risk of the event (coronary heart event), which is very high in FH. Opposing the risk of the primary event — ACS — is the competing risk of other things. Here patients with FH are diagnosed younger and have few competing risks. Another domain is the treatment benefit of a therapy. In this case, we know that statins are effective and have little harm.
The authors tell us in the discussion section that use of statins increased over time, from 46% in 1995 to 77% in 2021, whereas the rates of statin use went from 0.7% to 12% over the same time.
Another factor in normalizing the age of death is that the treatment of CAD surely improved. Stents replaced plain old balloon angioplasty, second generation stents were better than first generation.
Finally a reason for diagnoses not to completely normalize is that CAD diagnosis is much improved in recent times. High-sensitivity troponin is more sensitive than other biomarkers. We now have CT scans that pick up CAD. And, if more and more patients are diagnosed by cascade screening of a relative, than having the diagnosis of FH surely will make it more likely to be diagnosed.
I wanted to start the podcast with good news. Since so often I tell you all a story of dubious data or dubious data interpretation.
Route of Drug Administration in Out-of-Hospital Cardiac Arrest
Few areas of medicine seem better at generating proper evidence than the resuscitation space. I wish cardiology did this many proper randomized controlled trials (RCTs).
NEJM published an RCT comparing intraosseous vs intravenous (IV) access during an out of hospital cardiac arrest. Acute care of cardiac arrest is a bit out of my lane, but I cover this trial not only for the specific question but also because of its evidence-based medicine (EBM) lessons.
First author Michel Vallentin from Aarhus Denmark led the study.
As background, and from my reading, guidelines recommend initial attempts at IV access but “despite this recommendation, the use of intraosseous access is increasing.” A systematic review of observational studies showed that patients who received medication by an intraosseous (IO) route during cardiac arrest had worse outcomes than those who receive medication by IV. But that could be confounded.
Well, the way to sort out this question is not with more observational studies or trial emulation, but to randomize. And so they did.
About 1500 patients were randomly assigned to the two routes of initial IV access. Their primary endpoint was return of spontaneous circulation (ROSC).
Results:
The first finding was that 92% of the IO group had successful access vs 80% in the IV group.
Sustained ROSC occurred in 30% in IO group vs 29% in the IV group. Risk ratio 1.06; confidence interval (CI) 0.9 to 1.24; and P = 0.5. Obviously, no statistical difference.
Secondary outcomes, including survival at 30 days, favored neurologic outcome were also not different.
They also randomized humeral vs tibial IO access. There was no difference in ROSC, but CT scan at the hospital showed that 100% of IO catheters were correctly placed in the tibia vs only 71% of humeral catheters.
The authors concluded simply that there was no significant difference in ROSC with either of the two methods.
Comments. While you and I are not out in the field resuscitating people, as most of our patients in cardiac arrest are in the hospital and patients have IVs, this is a really nice example of how to answer questions in medicine. You randomize.
Now, the resuscitation people know that there is unlikely to be a difference in outcomes with either access point. Just get access and give epinephrine ASAP. In this trial the times to first dose of epinephrine were similar in both arms, and the authors tell us that this is likely the reason for equivalent outcomes.
That the trial was conducted in Denmark is also not a surprise, as the medical community and perhaps the general population accept equipoise and randomized trials.
When I visited the University of Michigan, we talked a lot about trying to change the culture of US medicine to be more accepting of randomization. If we had such a culture, we would know a lot more about what works and what does not.
The lack of knowledge in the left atrial appendage occlusion (LAAO) space comes immediately to mind. Hundreds of thousands of these devices have been implanted in patients who were excluded from the seminal trials, and so, we have zero idea if they work, all because we have not done what Danes seem to be so good at. RCTs.
Also notable was the data-sharing statement: “Will the data collected for your study be made available to others?” Answer: Yes.”
That is the right answer. Congratulations.
Factor XI Inhibitors May Not Be Dead
NEJM has published the results of AZALEA-TIMI-71, a phase 2 trial of the monoclonal antibody abelacimab vs a direct oral anticoagulant (DOAC) in patients with AF.
First, some background, which most of you probably know. DOACs are quite good at preventing stroke; are non-inferior to warfarin in efficacy and superior at preventing intracerebral hemorrhage (ICH). But DOACs still have a higher bleeding risk.
Factor XI inhibition holds promise to be an effective OAC (prevention of stroke) and have lower bleeding risk. There is mounting evidence that factor XI is essential for thrombosis but nonessential in most cases for hemostasis. Persons with genetically mediated factor XI deficiency have fewer embolic events without an appreciable increase in the occurrence of spontaneous bleeding.
A previous factor XI inhibitor, asundexian, also showed lower bleeding rates vs DOACs in phase 2 trials, but failed to be an effective stroke preventer. The OCEANIC trial of asundexian vs apixaban was stopped early because of a nearly 4-fold higher rate of stroke in the asundexian arm.
Abelacimab looked strong in a phase 2 trial of patients having knee surgery where a single IV dose after surgery reduced the rate of venous thromboembolism (VTE) by 80% compared with enoxaparin, without increasing bleeds.
AZALEA-TIMI Results:
About 1300 patients underwent 1-1-1 randomization of abelacimab 150mg , abelacimab 90 mg, or rivaroxabban. These were 74 year-old patients with mean CHADSVASC of 5.
Since this was a phase 2 trial, the primary endpoint was major bleeding.
The trial was stopped early because of a greater-than-anticipated reduction in bleeding events with abelacimab.
The incidence rate of major or clinically relevant nonmajor bleeding was 3.2 events per 100 person-years with 150-mg abelacimab and 2.6 events per 100 person-years with 90-mg abelacimab, compared with 8.4 events per 100 person-years with rivaroxaban.
The hazard ratio (HR) for 150-mg abelacimab vs. rivaroxaban was 0.38 (CI, 0.24 to 0.60); HR for 90-mg abelacimab vs. rivaroxaban, was 0.31 [CI, 0.19 to 0.51]; P < 0.001 for both comparisons).
The trial was not powered for stroke events, which are much lower. Note the DOAC trials had more than 10,000 patients and the previous asundexian vs apixaban trial had 14,000.
There were only 28 stroke or systemic embolism events in the trial. But they were slightly lower in the rivaroxaban group vs the A150 and 190. 7 vs 10 vs 11.
For ischemic stroke it was 5 (rivaroxaban) vs 10 (150 mg abelacimab) vs 10 (90 abelacimab).
Comments. Obviously, the question is whether abelacimab will face the same fate as asundexian — great in phase 2 and flop in phase 3 because it is fails to reduce thrombotic events.
The authors spend a paragraph telling us that there are important and favorable differences between abelacimab compared with asundexian. I won’t list them because there is now only one question:
Does it work in phase 3 trials vs DOAC? For this we have to wait.
A side note from an EBM perspective: I am struck by how much more rigorous drug development is than devices. I realize that device trials can never enroll as many patients as drug trials, but when a phase 3 trial of abelacimab comes out, we will have a very good idea if it works.
This sort of knowledge is less common with devices. But it doesn’t have to be. The answer is to approve a device for use if the regulatory trial is reasonably assuring, but then require proper phase 4 trials soon after approval. Such an approach would not penalize first-generation devices and would allow for continuing knowledge generation. We don’t have that with LAAO. And we went 15 years without it with IMPELLA.
JACC and the Tricuspid Valve
Last week, JACC had a focus issue on one expensive and new drug — finerenone. This week, JACC focuses on tricuspid valve interventions. As many of you know, I have been highly critical of the tricuspid valve evidence generators because of their highly soft trial design. Namely, not having a proper surgical placebo arm, and then using subjective patient measures to assess efficacy.
In the TRILUMINATE Full Cohort Study, nearly 600 patients were enrolled and the primary endpoint choice tells us a lot about trial design.
The primary outcome was a hierarchical composite of death or tricuspid valve surgery, heart failure hospitalization (HHFs), and quality-of-life (QOL) improvement measured by Kansas City Cardiomyopathy Questionnaire (KCCQ) at 1 year.
It was analyzed by the win-ratio. And of course the 1-year results were positive, just like the main trial results. The report in JACC this week included the full randomized cohort whereas the previous paper comprised part of the cohort.
And the results were the same. A positive win-ratio. But no difference in death, valve surgery, or HHF. All the difference was in change in KCCQ. But again. No placebo arm. It’s mind-boggling.
Now, though, my view has support from inside the gates of the institution. Drs Levy, Singh, and Stone wrote an editorial that dared to question the value of this intervention.
They cite the fact that in COAPT, the mitra-clip trial, large improvements in KCCQ correlated with hard outcomes, whereas it did not in TRILUMINATE.
They conclude that the discordance in QOL measures and hard outcome causes uncertainty of benefit as to the “true magnitude of benefit of this promising therapy.”
I would substantially change that wording. I would say that having a primary endpoint of a subjective nature without a proper placebo doomed the trial to fail. You can’t compare a big procedure to tablets. It belies basic science 101 principles. That it was not done makes me cynical that the industry-sponsored trial was not actually a science exercise but a marketing exercise. That academic doctors and the US Food and Drug Administration (FDA) went along with it is sad.
The wording I would use is that the discordance in QOL measures and hard outcomes causes uncertainty of benefit as to whether there is any benefit. If I were at FDA, I would make the company do a proper trial. A sham arm would have been so easy to do with this procedure.
My libertarian friends would say that FDA should approve devices like this and let the market decide. I am sensitive to that idea, because in a proper world, where doctors could adjudicate evidence, TRILUMINATE would convince no one. The device would be considered low in value and not used until a proper trial showed it to be effective.
But that doesn’t seem to happen. Incentives to do this procedure include financial rewards, as well as becoming a Top Person in the field. It’s sad. The whole thing is sad.
TRISCEND Trial Reports QOL
Also in the feature issue, is a report from the TRISCEND investigators. TRISCEND was a similarly weak RCT comparing the tricuspid valve replacement called EVOQUE vs tablets. This open label, non-placebo-controlled trial had a similar composite endpoint.
And once again the win ratio favored the valve (of course it did) because of the favorable QOL data. TRISCEND was even worse than TRILUMINATE because not only were there no differences in hard endpoints of death or HHF, there was a massive nearly 3 times higher rate of bleeding in the valve arm as well as a 9 times higher rate of pacemakers needed, not a small thing because this requires special pacers, either a coronary sinus lead or leadless.
The new report focuses on only QOL. It’s positive and similar to the main report.
My comments are the same. It is sad.
German Observational Study Called the PASTE Registry
JACC also included a report from a registry called PASTE. This included about 1000 patients who were treated with the PASCAL transcatheter edge to edge repair (TEER) system.
The good news, and you should always check for this, the registry included consecutive patients. That’s important because voluntary registries are about useless as can be, as investigators might not include ‘bad cases.’
The results show that the device performs reasonably well in reducing tricuspid regurgitation (TR). The second -generation device was better. Good.
They also report clinical outcomes, but these are useless without a control arm.
I may have missed it but I did not see a list of complications. The authors write that there were no major complications, but I am not sure if there were no complications.
As I have said, registry studies are fine to describe what happens. And so this is fine. But 1000 patients were in this registry. Imagine what we could have learned if these were randomly assigned with a proper placebo arm. We would know how effective this device is.
There is also an imaging substudy of TRILUMINATE. I don’t think it helps. We need outcomes and proper placebo controls.
Finally, there is an excellent editorial, from Drs Gupta, Yang, and Al-Lamee. They address in quite soft language the lack of benefit in hard outcomes, the problem with no placebo-control or sham arm, as well as the problem of selecting patients — especially the issue of causing right ventricular failure because of afterload mismatch.
They write and I quote:
Future trials must incorporate placebo controls to quantify the true benefits of these interventions and use longer-term follow-up to assess device durability and survival. Additionally, head-to-head comparisons of T-TEER and TTVR [transcatheter valve replacement] are needed to better inform patient selection and refine treatment strategies.
The problem I see, at least in the United States, is that once a device is on the market, and Centers for Medicare and Medicaid Services agrees to pay for it, equipoise will be shredded. That, too, is sad.
Finally, I want to add my disclaimer on devices. I suspect, as I did with IMPELLA, that there may be highly selected, even rare patients who could benefit from a TEER, and if the device is not approved, these few patients can’t benefit.
But the flipside of this is the marketing forces and incentives for doctors to do procedures combined with the high prevalence of TR. The risks of letting this device on the market is huge. If only the FDA and clinical scientists had followed the advice of Prof Al-Lamee and done proper placebo-controlled trials. Then we could know the true improvement with these devices.
© 2025 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: John M. Mandrola. Jan 24, 2025 This Week in Cardiology Podcast - Medscape - Jan 24, 2025.
Comments