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 5, 2024, John Mandrola, MD, comments on the following news and features stories: Neuro-modulation in HF, private equity as hospital owners, HCM, SCD and ICD re-evaluated, and peer review.
Barostimulation for HF
About a year ago, a vascular surgeon friend asked me what I thought of this carotid stimulation device. I’m like, a) why are you asking, and b) what are you talking about?
Apparently, our hospital system was considering doing barostimulation and it required a vascular surgeon. He told me he had looked at the study, and it’s not that impressive.
At the time I briefly looked at the BEAT HF study, and thought not much here, but, I paid it little attention. Mainly because this therapy didn’t have much traction.
Well, over the holidays, the FDA approved an expanded label for the Barostim neuromodulation system in patients with heart failure (HF) that incorporates longer-term, postmarketing clinical data from the BeAT-HF randomized clinical trial, the manufacturer, CVRx Inc, has announced.
The updated labeling for Barostim now states that the device is indicated for patients with NYHA class III or class II HF with a recent history of class III disease despite treatment with guideline directed medical therapy (GDMT), and who have a left ventricular ejection fraction (LVEF) ≤ 35% and an N-terminal pro B-type natriuretic peptide level < 1600 pg/mL.
I did not really know there was ANY indication. Now we have an expansion of an indication.
The grand idea is that there are groups of patients with HF with reduced EF (HFrEF) who still have symptoms and bad LVEF, despite medical therapy. And sadly, their QRS duration precludes benefit from cardiac resynchronization therapy (CRT).
This is the patient who might improve with modulation of the autonomic nervous system.
A company called CVRx came up with a subcutaneously implanted device, placed in the upper chest. A lead is tunneled from the carotid bulb area to the device which delivers stimulation to the carotid area with the goal of decreasing sympathetic tone (like beta-blockade). Basically, it counters the excessive sympathetic tone in HF.
The company and academics worked with FDA to design an adaptive trial. The main publication, (that I could find) is JACC 2020, first author, Michael Zile.
BEAT HF had four cohorts. Some designed for pre-market approval.
Each cohort was a small randomized controlled trial (RCT) comparing one group with the device and the other with medical therapy.
The main publication focuses on the last cohort of 260 patients randomized 1-1 to BAT+ medical therapy vs medical therapy. This so-called cohort D was the “intended use” indication.
The primary endpoint was a composite of three things—quality of life (QOL) scores, 6 min walk, and NT-pro-BNP.
At 6 months, the BAT group had better QOL, longer 6 min walk tests and slightly better NT-proBNP. All were statistically significant.
But I hope you all note the limitations. A) this is a surgical procedure that requires indwelling hardware and a vascular surgeon. B) there were no clinical outcomes, such as hospitalization for HF (HHF) or CV death, and C) and this is huge: there was no placebo or sham procedure. Patients in the device arm knew they had had surgery and patients in the medical arm knew that they did not.
What are people thinking? Of course, there will be a placebo effect. The authors tell us the lack of placebo is why they used NT-proBNP, an objective measure. But come on, BP is an objective measure and look what happened with the early renal denervation trials.
Patients who had surgery may have done other things that helped lower their BNP—better diet, more exercise, better adherence to meds, etc.
I had no idea how weak the evidentiary data for barostimulation was. And then reading that the FDA expanded the indication last month stimulated me to look harder.
I found that Zile presented longer term data from BEAT HF looking at clinical outcomes in March 2023 at the technology and HF therapeutics conference (THT). There was no improvement in the composite of CVD and HF events.
We refer most of our advanced HF patients to transplant centers. We don’t use this device. I’ve never seen a patient with one.
But it is FDA approved. With uncontrolled data and no clinical outcome data. How could this be? I am not against innovation.
But why would there not be a proper placebo-controlled trial? Or one powered for clinical endpoints?
If you are a HF person and know of stronger data, please alert me and I will update this critical appraisal.
Private Equity Owned Hospitals
Hospital Adverse Events Rise After Private Equity Acquisition
Changes in Hospital Adverse Events and Patient Outcomes Associated With Private Equity Acquisition
JAMA published a study looking at changes in hospital adverse events and patient outcomes associated with private equity acquisition.
I am not sure what’s happening outside of the US, but here, many hospitals are being acquired by private equity groups.
From the paper: “private equity has an often-distinct business model, in which the acquired entity typically assumes debt in the initial acquisition and is sold within a short time frame (often within 3-7 years).”
When I asked ChatGPT what do private equity firms seek to gain when acquiring hospitals, it said there can be many goals, but first listed profitability and financial returns.
I get the sense that profit in the healthcare sector is baddy. Indeed, the Biden administration is scrutinizing these takeovers.
The Bayesian priors before an observational study looking at outcomes after takeovers, may be that the health scientists expect to find worse outcomes. Because profit in healthcare is bad.
Here’s a brief summary of the study.
Harvard authors used Medicare claims data (adults > 65) to study outcomes at 52 private equity acquired hospitals vs 259 matched control hospitals over a decade. This was more than 4 million hospitalizations.
So this was a non-random comparison study looking back at claims data.
The main outcome measures were adverse events in the hospital—falls, infections. They also looked at mortality discharge disposition and readmissions.
You can guess the results. Profit motive in hospital care is BAD. The authors found a roughly 25% increase in hospital acquired conditions at acquired hospitals vs controls. This was driven by a 27% increase in falls, 37% increase in central line infections, and a doubling of surgical site infections.
And the news coverage was all bad. Headlines related bad outcomes after profiteers take over.
I am no fan of excessive profiteering in the care of people who are asking for our help. But I also think the use of dubious evidence to advance an idea is bad.
First, the relative increase of 25% sounds terrible. But the absolute increase in say falls, was equal to an extra 2 falls per 10,000 admissions. In fact, all of the adverse outcomes were measured in per 10,000 admissions. An extra 2 falls per 10,000 admissions. I am sorry, that may be statistically significant, but 9,998 other patient admissions had no difference.
Lesson number 1 – studies with millions of patients can have tiny differences that reach statistical significance.
Second issue, if private equity was so bad then why didn’t it also associate with an increase in mortality? In fact, mortality was lower in the private equity hospitals. And guess what: this lower mortality also met statistical significance.
The authors downplay this finding by saying a) it is explained by healthier case mix, and b) there were no differences at 30 days.
Hmm… except when they adjusted for observable variables, the mortality remained lower at those bad private-equity-owned hospitals.
Third, as the authors tell us in their limitations: “private equity acquisitions are not random and private equity firms may acquire hospitals for unobserved reasons, our findings remain susceptible to unmeasured confounding and do not imply causation.”
My take-home:
Always be wary of news headlines from observational studies. This one is from a smart group at Harvard, and it looks to be a very detailed careful analysis.
But big data like this is noisy. There are tiny absolute differences. Private equity proponents might have spun this study as we save lives. They might have easily explained away the higher rates of adverse events as noise, or confounding, or whatever.
I highlight this paper because these types of studies are common in our field of cardiology, but I just don’t think they are capable of answering big policy questions.
Look at the paper and tell me why I am wrong.
HCM News – A new Drug Coming Soon
Cytokinectics, the makers of the myosin inhibitor, aficamten, announced topline results of the SEQUOIA-HCM trial. This was a phase 3 placebo-controlled trial in patients with hypertrophic cardiomyopathy (HCM).
From the press release:
Aficamten significantly improved exercise capacity compared to placebo, increasing peak oxygen uptake (pVO 2) measured by cardiopulmonary exercise testing (CPET) by a least square mean difference of 1.74 (95% CI, 1.04 - 2.44) mL/kg/min (p=0.000002). The treatment effect with aficamten was consistent across all prespecified subgroups reflective of patient baseline characteristics and treatment strategies, including patients receiving or not receiving background beta-blocker therapy.
Statistically significant (p<0.0001) and clinically meaningful improvements were also observed in all 10 prespecified secondary endpoints, including (KCCQ-CSS) at weeks 12 and 24, the proportion of patients with ≥1 class improvement in (NYHA) functional class at weeks 12 and 24, change in provoked left ventricular outflow tract gradient (LVOT-G) and proportion <30 mmHg at weeks 12 and 24, as well as exercise workload and guideline-eligibility for septal reduction therapy.
LVEF was observed to be <50% in 5 patients (3.5%) on aficamten compared to 1 patient (0.7%) on placebo.
The market thinks this is going to be a big deal. The stock price of CYTK doubled after the announcement around Christmas.
It sure seems positive. The results will be announced at a meeting and published as well.
I don’t know about your practice, but use of the first generation myosin inhibotr mavacamten has been logistically challenging.
Perhaps this drug might be easier. The chance of lowering EF is a headwind. But a disease modifying drug for pts with HCM surely seems a welcome innovation.
Let’s wait and say more when we see the actual results.
The Primary Prevention ICD gets another look.
News came this month that the first patient has been randomized in a trial called PROFID EHRA.
This may be the boldest trial done in cardiology in decades.
German professors Gerhard Hindricks and Nikolaos Dagres will lead this modern day repeat of MADIT 2.
PROFID EHRA challenges the current guidelines that patients with ischemic, yes, ischemic CM, EF < 35% on GDMT, should have a primary prevention ICD.
The trial will randomize more than 3000 patients who are post MI with EF of < 35% to an ICD or GDMT and then measure death rate.
It will be a noninferior (NI) design. And that is fine, because not having surgically implanted hardware, and the potential for bacteremia, sepsis, and wide-awake shocks is clearly a less invasive option.
The trial will run for many years. I can’t say enough for the scientific thinking of Hindricks and Dagres here. This trial likely could not have happened in the US.
But the primary prevention ICD story is a classic one for evidence having expiration dates.
Think about it. MADIT 2 recruited patients around the turn of the century. HF therapy has advanced so much since then. The Kaplan Meier curves in MADIT 2 started to separate after 1.5-2 years. This was because the proportion of HF patients who died suddenly was substantial.
But, we know from HF trialists, that the proportion of patients with HF, who have SCD has gradually decreased over the years. Read a paper in NEJM by Shen et al. The authors, HF trialists, analyzed data from 40k patients in HF/ICD trials from 1995-2014.
They observed a 44% decline in the rate of sudden death across the trials (P=0.03). The cumulative incidence of sudden death at 90 days after randomization was 2.4% in the earliest trial and 1.0% in the most recent trial.
The only way an ICD can help HF patients live longer lives is if the proportion of SCD is high enough. HF in 2023 is different from HF in 1999.
The other piece of evidence that primary prevention ICDs have lost their net benefit was DANISH. Published in 2016, in patients with nonischemic CM, the ICD showed no signal of benefit over background medical therapy.
DANISH has largely been ignored because it does not comport well with the other incentives to use ICDs.
I strongly suspect that PROFID-EHRA will be quite similar to DANISH. But of course I don’t know. So we wait.
But what’s really invigorating is that European leaders had the gumption to do this trial. To accomplish such a feat, they had to a) be neutral arbiters of the evidence, b) have the persuasion abilities to convince colleagues to randomize patients, and c) the energy to garner funding.
Peer Review:
JNO published a fun paper led by a group at Stanford looking at peer review.
The study set out to answer questions about the quality of peer review. Which is super-important, obviously.
Much has been written about the vital nature of peer review, but also its severe limitations—which include being too lax, and missing things that should have been found, but also, being too restrictive and insular, and not allowing findings that question the prevailing dogma.
First author of the current JNO paper, Pascal Geldsetzer, and his colleagues did a cross-sectional study of almost 12,000 peer reviews from 3 general medicine journals. This corresponded to about 4000 publications in BMC Medicine, The BMJ, and PLOS Medicine.
They made an interesting choice for measuring quality in peer review. Their surrogate measure: extremely short peer reviews and how often they inform editorial decisions. Their idea is that a peer review of less than 200 words is poor. (I will come back to that.)
The median word count of peer review was 425. The overall prevalence of very short (<200 words) peer reviews was 17.1%. Across the 3 journals, 20.9% were based on review sets containing 50% or more very short reviews.
The authors concluded:
In this study of 3 leading general medical journals, one-fifth of initial editorial decisions for published articles were likely based at least partially on reviews of such short length that they were unlikely to be of high quality.
My take:
I like the idea of more systematic empirical study of peer review. I’ve been on both sides of peer review: doing it, and having papers shaped by insular peer reviewers.
This paper is interesting because of its choice of surrogate endpoint—very short reviews. They say it occurs one-fifth of the time, and this is bad because short reviews are bad.
I am not sure about that. Many times, I read a paper, and want to simply write, it’s fine. The authors describe their study. It’s limited in scope. They say so and make cautious conclusions, and I don’t feel that adding minutia is helpful: let the readers decide; they aren’t stupid.
Most of the time, my main contribution to peer review is to stop the authors from making outsized conclusions. As a person who thinks short writing is quite worthy, I don’t agree that short reviews are necessarily bad.
Short writing is way harder than long writing.
The problem with most medical studies is not the limited methodology (small numbers, non-random comparisons) but the outsized conclusions. It does not take many words, to change those things.
In most cases, I think, medical writing would be improved with fewer not more words. When in doubt, cut words.
One final comment: this study was made possible because those three journals publish their peer reviews. This, I believe, would be a great advancement in cardiology.
If peer reviews were published, the reviewer comments would be far better. Even if you kept the reviews anonymous, it would still be beneficial to readers to see what the referees wrote.
When I peer review, I write as if it were public.
© 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: Jan 05, 2024 This Week in Cardiology Podcast - Medscape - Jan 05, 2024.
Comments