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 February 16, 2024, John Mandrola, MD, comments on the following news and features stories: An update in trial interpretation, confidence intervals (CI), and P-values; percutaneous tricuspid Interventions; and news on HDL cholesterol and humility in Medicine.
Evidence of Absence Is not Absence of Evidence
First, I need to update my discussion of the ARCADIA trial last week. Recall that ARCADIA tested apixaban vs aspirin (ASA) in post-stroke patients who had evidence of atrial cardiopathy. Atrial cardiopathy could be diagnosed by abnormal ECG, BNP, or left atrium (LA) size on echocardiogram.
My discussion centered on the fact that the hazard ratio (HR) was 1.00 and there were equal numbers of primary outcome events (stroke) in the two groups.
This past week, I was told by groups of trial experts and statistics professors that I made a classic error. I lead the podcast with my mistake because it is such a core teaching point for interpreting medical evidence.
My error, experts emphasized, was that I had equated the absence of evidence with evidence of absence. Let’s translate that.
ARCADIA had the same number of stroke events in the apixaban group as in the ASA group. The HR was 1.00. The P-value was 0.99. Indeed, there was no signal of difference.
The clue that I should have noted right away was the 95% confidence interval (CI) surrounding that HR of 1.00. The CI went from 0.64 to 1.55. Roughly, this means apixaban, relative to ASA, could have reduced stroke by up to 36% or increased it by 55%. Both of these are clinically important effects.
The experts point, therefore, is that ARCADIA was underpowered to tell us that there was evidence of absence of effect of apixaban vs ASA.
But there are even more teaching points. You may wonder why the trial was underpowered. How did this happen?
Trialists have to make estimates before a trial as to how many patients to enroll. The main variables that go into this estimate are the expected event rates in the respective arms of the study, as well as the minimal clinically important difference you would like to detect.
The authors estimated a 7% annual stroke rate in patients with atrial cardiopathy who had just had a stroke.
To determine the minimal clinically important difference, they used a previous trial of apixaban vs ASA (AVERROES), which found a 55% reduction in stroke with apixaban, to declare a 40% reduction as their minimal clinically important difference.
More translation: they “powered” ARCADIA to find a very large effect size. You can use a rough sample size calculator to estimate how many patients this would require. They got to about 1100 patients.
But. There is another factor in these calculations that does not appear in any calculator. And that is pragmatism. Before I started reviewing trials, I did not know about pragmatism in the design of trials.
What I mean by pragmatism is that trials cost money and effort. So, if you try to detect smaller differences in the two groups, you need many more patients. Precision, like money, does not grow on trees.
Dr Sanjay Kaul calls this sample size samba: A dance (or tension) between getting enough patients enrolled to detect differences but doing so within the pragmatic budget constraints.
Many cardiology trials are funded by industry. Industry has plenty of money. These tend to be mega trials that can detect small differences between groups. (Look at the FOURIER trial of evolocumab vs placebo; N ≈ 27,000)
I played around with the online sample size calculator and entered a 25% reduction in stroke with apixaban in ARCADIA. This would have required triple the number of patients and that would have required a lot more funding. ARCADIA was not funded primarily by industry.
You can see now how ARCADIA got to that wide CI. They declared as their minimal clinically important difference a 40% reduction.
The problem with that is we would all agree that a 39% reduction in stroke would be quite important.
Because of these revelations, I would amend my take of ARCADIA as saying we simply do not know whether or not apixaban is better than ASA in the cohort of atrial cardiopathy patients (as defined in this study).
Further, we have also not settled the issue of whether or not markers of atrial disease or cardiopathy is an important predictor of success with oral anticoagulation therapy.
The final point, however, affects my views on randomized controlled trials (RCTs). While I still believe RCTs are the way forward in knowing what to do in Medicine, we must be careful in how we interpret the uncertainties in trial results.
Tricuspid Interventions
Structural interventions for tricuspid regurgitation (TR) had a big 2 weeks at US Food and Drug Administration (FDA). Two terms – percutaneous valve replacement and repair.
Before I tell you about the developments, let’s set out some quick background on TR. TR is super common. Most is secondary TR, which means the valve leaks because of pressure or volume overload. TR is easily seen on echocardiogram.
In other words, secondary TR means that it almost always occurs because of something else--heart failure with reduced ejection fraction (HFrEF), mitral regurgitation (MR), etc. TR almost always gets better when you treat the underlying cause. The patient who has TR as a primary source of their problem is quite uncommon.
A nice trial from the Cardiothoracic Surgical Trials Network (CTSN) investigators, published in the New England Journal of Medicine in 2022, tested concomitant tricuspid valve (TCV) repair at the time of surgery for MR.
The trialists reported that concomitant repair reduced the degree of TR, had no effect on mortality, and increased the rate of pacing by 6 times. Clearly this was not strong evidence for adding TCV repair.
The trial illustrates my point wherein TR commonly accompanies other pathology.
The background is important and I think largely ignored by FDA because when these interventions are approved, there will be a tremendous incentive for doctors to overuse them.
Percutaneous TCV Replacement. The first percutaneous TCV replacement, called the EVOQUE, was recently cleared for use without an advisory committee meeting.
The valve replacement was studied in a trial called TRISCEND II that, as far as I know, is not published yet. It was presented at the Transcatheter Cardiovascular Therapeutics meeting in October 2023.
Edwards Lifesciences makes the valve. The approval process began with FDA declaring the valve as a “breakthrough” therapy. This is important because part of the designation is that FDA may accept greater uncertainty if probable benefits outweigh harms. I try to keep cynicism at bay, but getting this designation sounds somewhat dodgy.
TRISCEND had a two-part design: The first 150 patients were like a safety feasibility group. The total cohort would be 400 patients.
The report below is the first 150 patients. Here we get only safety and soft endpoints. The total cohort study will add clinical outcomes like death, TCV surgery, hospitalizations for HF. But this first report does not include any clinical endpoints.
Enrolled patients had to have signs and symptoms of TR, and a local heart team determined the patient to be appropriate for TCV interventions.
Exclusion criteria included anatomy that would not work, low life expectancy, EF < 25, bad right ventricular (RV) function, and pacing leads — but not all pacing leads. Jailing a pacing lead was allowed if the patient was not pacer-dependent or had a primary prevention implantable cardioverter defibrillator (ICD).
The primary endpoints of this trial are seriously problematic. There was a safety endpoint that was a composite of major adverse events (MAE) at 30 days, and a primary efficacy endpoint at 6 months of a composite of TR reduction (an echo parameter), and then a combination of the Kansas City Cardiomyopathy Questionnaire (KCCQ), New York Heart Association (NTHA) functional class, and 6-minute walk distance (6mWD).
The safety endpoint was derived from real world data from TCV surgery with a benchmark set at 44%. I kid you not, If you they had less than a 40% major adverse event complication rate at 30 days, the new valve could be declared safe.
The first efficacy endpoint was TR grade reduction or the proportion of patients with TR < moderate; and the second efficacy endpoint was KCCQ improvement > 10 points, or NYHA improvement > 1 class, or 6mWD > 30 meters.
The trial procedures were valve group vs optimal medical therapy (primarily oral diuretics).
It took about 2 hours to place the valve. Hospital stay was about 4 days.
And the results – I know you will be shocked. The valve group met its safety endpoint. Of the 96 patients randomly assigned to the valve group, 26 had safety events. The 27% incidence of safety events easily passed the bar of safety.
These safety events were things like 3 deaths, 10 severe bleeds, 2 device complications, and 14 new pacers (keep in mind that pacing with a TCV replacement is no small thing. It will require coronary sinus pacing most likely).
The first efficacy endpoint showed marked reduction in TR. And, of course, the hierarchical endpoint was positive. Because when you compare subjective endpoints like KCCQ and NYHA functional class in patients who had a procedure vs those who did not have a procedure, there will be a massive placebo effect. Same for 6mWD. Patients with surgery will be motivated to go that extra 30 meters.
The authors’ final slides declare the benefits of the new valve. They write:
Treatment of severe TR with the EVOQUE system resulted in meaningful improvements in functional status and symptoms at 6 months.
And this:
The unique trial design of TRISCEND II, based on FDA breakthrough designation, provides an early look at the safety and effectiveness of the EVOQUE system in the first 150 patients. Important clinical and echocardiographic endpoints including mortality and heart failure hospitalization from the full cohort of 400 patients will be presented in the future.
My Comments. I hope they do finish the study and tell us about hard clinical outcomes like death, rates of TCV surgery, and hospitalizations for HF. I find it shocking that FDA would approve this valve without an advisory committee based on this data.
A 27% rate of MAE. That is nearly 1 in 3 patients have a bad outcome. And, one of the efficacy endpoints is an echo parameter (TR). And of course it is not blinded.
The other two efficacy endpoints are subjective and there is no placebo-arm. Online, a few doctors called me out for dissing the subjective endpoints. They said it’s important to make patients feel better.
My response is of course it is. But to know whether TCV interventions actually do that, you need a proper sham control. You can’t do a procedure on one group and give the other group simple tablets and make any conclusion about quality of life.
The fact that colleagues of mine — those with MD after their names — are so easily bamboozled shocks me.
Sanjay Kaul had a nice comparison on X. He compared the approval of this new valve with the rejection of a new drug, patisiran, for ATTR (transthyretin amyloid) cardiomyopathy.
In the APOLLO-B trial (N = 359). patisiran had a win ratio vs placebo of 1.27 but a CI of 0.99 – 1.61. The advisory committee voted 9 to 3 in favor, but FDA rejected patisiran.
He noted that TRISCEND had fewer patients, shorter follow-up, and all components of the efficacy endpoint were susceptible to placebo effect. Patisiran had a proper placebo.
It was a good contrast between different standards for drug and devices.
Now to Tricuspid Valve Transcatheter Edge to Edge Repair or TriClip. This week FDA convened an advisory committee meeting to evaluate the application for the Abbott TriClip transcatheter edge to edge repair system (TEER) for patients with severe TR.
I’ve served on an FDA advisory committee (for the drug vernakalant) and what happens is that FDA presents data, the sponsor presents data, and the committee even hears from the public.
Then they vote on questions. The big three for the TriClip were 1) was the device safe; 2) was it effective, and 3) were benefits greater than the risks.
The committee relied mostly on the TRILUMINATE trial, which I have covered several times.
Before I tell you the results of the vote and before I recap TRILUMINATE, I want to set aside that there may be a niche for a percutaneous approach to patients with severe TR, because a) these patients can be quite symptomatic with edema and right HF and b) surgical TCV approaches are risky.
The problem of course is as I stated, most of TR is caused by other issues in the heart.
TRILUMINATE may have been the most disappointing trial I have covered in my critical appraisal career. Disappointing specifically in its design.
TRILUMINATE tested TEER vs medical therapy in patients with symptomatic severe TR.
The trial delivered a positive result for TEER, but authors chose a three-component hierarchical primary endpoint of all-cause death or tricuspid surgery, hospitalization for heart failure, or an improvement in quality of life as measured with the KCCQ. Note the two objective and one subjective endpoint.
The results favored TEER with a win-ratio of 1.48 (95% CI, 1.06 to 2.13; P = 0.02).
Positive results occurred solely due to the KCCQ aspect of the endpoint. The hard endpoints did not differ.
Another curious finding: If the device worked so well in improving quality of life, why didn’t the 6mWD improve? There were no significant differences in this endpoint.
The core problem was that TRILUMINATE did not include a placebo (sham) procedure.
It’s highly likely therefore that a placebo effect drove the positive results. And the thing of it is, everyone involved with this design knows this. This week, the principal investigator of ORBITA 1 and 2, which tested severe coronary lesions with a placebo (they randomly assigned people with proximal LADs to a sham PCI) tweeted the following:
Of course a placebo-controlled trial of TriClip is possible. It is time for these trials to become entry standard for interventional procedures aimed at improving subjective endpoints. We should not accept substandard evidence of efficacy for procedures with both risk and cost.
The committee voted positive on all three questions, and it seems likely that FDA will also grant approval to TEER of the TCV.
Summary Comments. I am not against helping patients with TR. But this is a mess. There are proper ways to study devices. You need hard endpoints. You need proper controls.
Doing a sham control for tricuspid TEER would be so simple. But it was not done. Even though the scientists charged with studying this device knew it was necessary.
The FDA was the last line of defense. I have no idea how regulatory guidance could be this weak. FDA could have demanded proper studies.
In a perfect world, where doctors are wise adjudicators of evidence, these devices would be used sparingly, in very select patients. But in a profit- and productivity-driven system I doubt this will happen.
I am sad for our profession because proper studies were not done. I am sad about FDA. I hold out some hope that my colleagues will not fall gullible to marketing pressure.
HDL and Cholesterol Efflux after MI
This is a story about how good stories often don’t pan out.
HDL is known as the “good” cholesterol. Because it takes cholesterol from the periphery back to the liver for excretion. HDL levels inversely relate to cardiovascular (CV) outcomes. High HDLs, to a point, are associated with less CV disease.
Yet, despite a massive effort by lots of people and companies, modification of HDL levels has not proven successful in improving CV outcomes.
The first step in the cholesterol clearing process is called cholesterol efflux. And, for this, you need a big protein called apolipoprotein A-I (apoA-I) which is the chief protein of HDL.
This clearing process may be especially important in the few days after acute coronary syndrome (ACS) and related myocardial infarction (MI). But, in the days after MI, it seems cholesterol efflux is diminished, which is, obviously, bad and may contribute to recurrent events early.
Enter a novel, infusible human plasma-derived apoA-I that robustly elevates cholesterol efflux with relatively modest increases in apoA. This compound may be less susceptible to such acute phase modification and may retain functionality during this time period.
The story goes that this agent can elevate cholesterol efflux capacity and may be beneficial among post-acute MI patients.
But this week we learned that the AEGIS II trial, a placebo-controlled RCT that randomly assigned more than 18,000 patients after an ACS did not meet its primary efficacy endpoint of major adverse coronary event reduction at 90 days. As a result, there are no plans for a near-term regulatory filing.
There were no major safety or tolerability concerns with this agent, CSL112.
The results will be presented at the upcoming American College of Cardiology meeting.
I can keep this as a short story but, once again, the message is humility. Humility because stories that make sense (raising HDL or increasing cholesterol efflux after MI) usually don’t pan out.
© 2024 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: Feb 16, 2024 This Week in Cardiology Podcast - Medscape - Feb 16, 2024.
Comments