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 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 September 16, 2022, John Mandrola, MD comments on the following news and features stories.
At ESC, Prof Michelle O’Donoghue, from Harvard, presented results of an extension trial of FOURIER called FOURIER-OLE.
Recall that in FOURIER, a little shy of 28K patients with established CAD and an LDL-C > 70 mg/dL on statins were randomized 1:1 to Evoculomab or to placebo.
Evolocumab resulted in a 15% relative risk reduction in the primary endpoint of CVD, MI, stroke, UA, or revascularization. Nonfatal outcomes drove the modest 1.5% ARR. There were no significant differences in CVD or overall death.
My take of the trial FOURIER and ODYSSEY trials of PCKS9i is that for young patients (62 years) who have established disease there is a modest effect in reducing nonfatal outcomes. The argument is not whether PCKS9i work, it’s whether the effect size on nonfatal outcomes warrants the cost. FOURIER-OLE set out to establish safety over a longer time. They also looked at outcomes as well, but these were secondary endpoints.
FOURIER-OLE was a non-randomized post-hoc comparison. About 3300 of the (14k patients or 25%) in each group stayed in the extension study. 75% of the two groups did not stay in the extension study.
Both groups in FOURIER OLE got the PCSK9 inhibitor.
The main finding was that over about 5 years in this small and selected sample of patients, no safety issues (muscle-related, hemorrhagic stroke, diabetes, neurocognitive) were identified—this despite getting median LDL-C to 30!.
The finding that garnered most attention was that the primary outcome of CVD, MI, stroke, UA and revascularization, and a key secondary outcome of CVD, MI, Stroke favored the evolocumab from the start group over placebo then evolocumab group.
The KM curves kept diverging—for the first 3 years. Though in a landmark analysis looking only at years 1-5 of FOURIER-OLE, the legacy benefit of being on evolocumab from that start waned after 3 years – the placebo-evolocumab group caught up.
Comments: I have come to believe that long-term exposure to low LDL-C is where the most benefit lies. It makes sense, is seen in statin trials and confirmed in Mendelian randomizations studies.
So, the legacy effect where in being randomized to evolocumab continues for 2-3 years even when the comparator arm also gets the drug makes sense.
I agree with the authors’ conclusions about this data supporting earlier use of LDL-C therapies in young patients with bad vascular disease.
Here are the caveats though:
The median follow up was 5 years. That’s hardly very long for a 62-year-old. The worry with LDL-C of 30 is not 5 years, it is 10-20 years. The other reason I am not swayed by this analysis is that it represents a fraction of the main trial, it’s a highly selective non-random sample, and there is no longer a comparison arm.
Again – to re-iterate the probability of a safety signal is low.
As for the legacy effect, it is there. But the effect was transient and driven mostly by nonfatal events.
PCSK9i would make a lot more sense at a lesser cost. But that is a debate for policy and health economists not cardiology podcast.
RE-Analysis of Paradise MI of sacubitril/valsartan in Post MI patients
At ESC, investigators presented a re-analysis of PARADISE MI using a statistical technique called the win-ratio. It’s a complicated endpoint, but the take home is that its proponents say it incorporates more of the data collected.
Recall that PARADISE MI randomized patients within days of an AMI. These patients had LVEF ≤ 40% and/or pulmonary congestion, and at least one other risk enhancer, like older age, CKD, AF, or EF < 30%.
Over a median follow up of 23 months, there was a 10% reduction of primary endpoint with sac/val with a hazard ration of 0.90 (95% CI, 0.78-1.04) and a p-value of 0.17.
It was definitely a neutral trial. The investigators then published another post-hoc analysis that focused only on investigator-reported events (rather than clinic events committee [CEC] adjudicated). There were more events in this analysis. And the HR was slightly better at 0.85 and now the CI were tighter and it met statistical significance.
This gives one the clue that had the trial used more of the data, there would be an advantage of sac/val. The other point to make is that time to first events are usually dominated by nonfatal events. But you also want to know about fatal events.
The objective of this new analysis was to provide an additional perspective into understanding the effects of sac.val in patients with AMI and HF.
They used a win-ratio analysis which considers the hierarchy of outcomes—that is the more serious events are given higher priority and are analyzed first. They also investigated the difference in analysis when looking at events that were reported by investigators vs events that were confirmed by the CEC committee.
Win-ratio considers wins vs losses in each of the three components of the primary endpoint—CVD, hospitalization for HF (HHF), outpt HF.
For each matched pair, the new treatment patient is labelled a 'winner' or a 'loser' depending on who had a CV death first.
The win ratio is the total number of winners divided by the total numbers of losers. The other thing they did was tell us the main contributors to wins. CEC confirmed CV death contributed more than a third of the wins. This is important b/c a CEC confirmed CV death is the “hardest” endpoint.
The advantage of this technique over a time to first event analysis is that it incorporates more information. In time to first event, if you get a HHF, that’s it. In a win-ratio, the same patient can contribute a HHF and CV death.
Professor Otavio Berwanger, MD, from Brazil was very careful in saying this analysis was directed at trialists. He started by saying PARADISE MI remains a neutral trial. This analysis does not change guidelines nor should it change clinical practice. Good on him.
I have two hats here: the hat that loves methods and trials and evidence really appreciates this analysis.
That said, I am not so persuaded by their emphasis on investigator reported events. A clinical events adjudication committee is there to adjudicate and make sure the events are real. They don’t tell us the win ratio for only CEC adjudicated events. I wonder if that win-ratio would have been positive.
When results are close—as they were in PARADISE MI—an analytic method can toggle the significance/nonsignificance line. The thing is you have to prespecify what methods you are going to use BEFORE the experiment. Post-hoc analyses once you know the results can only be used to generate ideas for other trials.
The other hat I am wearing is my skeptical or maybe even cynical hat. If this were for trialists more than clinicians, why not present it at a statistical or epi conference? Yes, ESC has basic science but the thrust is for clinicians.
And here you have a hot-line presentation that could be interpreted like this: hey, docs, PARADISE MI was almost positive for sac/val. But when we use this better statistical technique, it is positive. Professor Berwanger was clear in his words that this wasn’t the right interpretation, but still, it’s a favorable analysis for an expensive drug at a hotline at ESC.
SGLT2i Post MI
EMMY was a double-blind placebo-controlled trial of the sglt2 inhibitor empagliflozin started in the hospital within 3 days of an acute PCI for MI.
This is an important space to study SGLT2i. The drugs are proven beneficial in DM, CKD, HFrEF and there is an arguable degree of benefit in HFpEF.
The post-MI space is another area they could help. However, there are hurdles—because our treatments are so good. Acute PCI now preserves LV function, and prevents heart failure so it’s hard to improve on that. EMMY used BNP as a primary endpoint. It measured EF and diastolic parameters by echo as a secondary endpoint.
The results were positive. But modest. There was a 15% lower BNP at 26 weeks in the Empa arm. There was a statistically significant but very small change in EF at 26 weeks. Same with diastolic parameter. They were a bit better too in the E arm.
Again, the mean EFs were normal, so I am not sure how you improve on that. The trial had too few outcome events to make any comment on.
I think this class of drugs has real value in selected patients. That said, before we start using them in a totally new indication--post-MI patients, we are going to need big outcome trials. Remember –no one knows how SGLT2i work. So anything short of clinical outcomes in this drug class is not very helpful. EMMY is a great effort, but a 15% decrease in BNP at 26 weeks and modest changes in echo parameters are like mile 1 of a marathon.
I am not sure you can make AMI patients who have a PCI and a normal EF much better with ANYTHING in the short term. Yes, of course, statins, antiplatelets, lifestyle changes, ACE- ARB, but even beta-blockers in this setting are questionable.
The ongoing DAPA-MI and EMPACT-MI trials, will test whether dapagliflozin and empagliflozin, respectively, can lower the risk for HHF and death in patients with new cardiac dysfunction after MI. EMPACT-MI will additionally evaluate patients with preserved EF but new and acute signs of HF in the setting of MI.
If they show an important and statistically robust decrease in clinical outcomes, I will change my mind.
LAAO patient selection:
JACC EP has published an important observational study of patients who received percutaneous left atrial appendage occlusion (LAAO). First Author Jules Mesnier.
The study included more than 800 patients from centers in Spain, Italy and Canada from 2009 to 2019.
Their goal was to investigate the causes of early death after LAAO. The mean age of patients was 76.
Here is why it is an important question:
If you believe there is a net benefit from percutaneous LAAO, I don’t, but let’s say you do, it takes time for this to accumulate. Proponents believe that the combination of plugging the appendage and removing anticoagulation will lead to fewer major bleeds and no increase in strokes.
But these sort of probabilistic benefits take years to outweigh the probabilistic harms of the procedure. Even the Watchman authors themselves published a paper saying it took 5 years to reach cost-efficacy.
If patients die soon after the procedure, they can’t possibly garner any probabilistic gain.
What these authors showed was that 1 in 6 or 15.5% of patients in this series died within the first year.
They also did multivariate regression and found typical risk factors for early death—older age, underweight, diabetes, HF, and CKD. Patients with 3 or more of these had a 30% chance of dying.
This really worries me. In my first debate on LAAO—at ESC against Horst Sievert from Germany, even he admitted that older patients with multimorbidity were least likely to benefit.
My experience locally and among US colleagues is that older patients with co-morbidities are commonly referred for and implanted with a closure device.
In fact, I was part of a poster at ACC that found Medicare beneficiaries undergoing pLAAO have higher comorbidity burden, thromboembolic risk, and prior bleeding events compared to patients enrolled in the pivotal trials that led to FDA approval.
The authors chose strong words in their closing. They write that patients who have early death after LAAO have had a futile procedure. Some may disagree, but I don’t.
The whole point of this procedure is to provide a probabilistic benefit in the distant future. If there is no significant future, b/c of early death, there can be no benefit. Worse than futile, the patient with early death still incurs the risk of harm from the procedure.
Again, I don’t think the seminal trials suggest benefit from this procedure in any patient. But obviously, many of you disagree, so, I say, at least choose patients wisely.
© 2022 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: Sep 16, 2022 This Week in Cardiology Podcast - Medscape - Sep 16, 2022.