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 February 25, 2022, John Mandrola, MD comments on the following news and features stories.
Heart Failure Therapy
The normal way to take care of patients with heart Failure (HF) is to use guideline-directed medical therapy (GDMT) and manage diuretics to maintain fluid balance. Diuretic management typically turns on clinical signs and symptoms, like weight, edema etc.
Here is the proposal: a paperclip-sized device (placed in the pulmonary artery during a minimally invasive procedure) that is capable of monitoring pressure changes and remotely transmitting these daily pressure readings to a patient’s clinical team would be able to manage patients better than standard care.
The device is made by Abbott and is called CardioMEMS. It already has a specific indication based on the 2011 CHAMPION trial, which randomly assigned 550 patients with New York Heart Association (NYHA) functional class III heart failure—irrespective of left ventricular ejection fraction (LVEF)—and a heart failure hospitalization during the previous 12 months. CHAMPION found a significant 28% reduction in HF hospitalizations at 6 months, but the study was not powered for mortality.
After initially failing at the US Food and Drug Administration (FDA), the FDA finally approved CardioMEMS in 2014 for the specific indication outlined in CHAMPION. You had to have class III HF and a recent hospitalization for HF. That’s a pretty specific indication; if you are the makers of the device, you want to be able to use it in less ill patients with more general indications. (Also, if it really works as advertised, it would be great to expand indications.)
So, the makers of the paperclip-like device sponsored another trial – it’s called GUIDE-HF and I have discussed it before on the podcast. The Lancet published GUIDE-HF in August of 2021 simultaneously with its presentation at the European Society of Cardiology (ESC) Congress.
GUIDE randomly assigned roughly 1000 patients (double the size of CHAMPION) to an invasive arm using the device vs standard care. Randomization was after implant of the device. The treatment arm used data from the device; the control arm was standard care, even though those patients had the device. Patients were blinded to the treatment assignment, but clinicians were not.
Entry criteria in GUIDE HF was broader than in CHAMPION; patients could have class II to IV chronic HF, and either a recent HF hospitalization or elevated BNP. The primary endpoint was a composite of all-cause mortality and total HF events at 12 months assessed in all randomly assigned patients
Briefly, the patients were a mean age of 70 years, overweight, with a body mass index of 37; one-third had class II and two-thirds had class III HF, the mean LVEF was 38% to 40%
Here were the results (recall that there were about 500 patients in each group:
There were 253 PEP events in the treatment arm – 51% overall or 0.56 events per patient year.
There were 289 PEP events in the control arm – 57% overall or 0.64 events per patient year.
The relative risk or hazard ratio (HR) was 0.88, a 12% reduction, and the confidence interval (CI) ranged from 0.75, a 25% reduction, to 1.05, a 5% risk increase.
The P-value was 0.16.
40 patients died in the monitoring group vs 37 in the standard care group. So obviously, no signal of mortality benefit.
This was a solidly non-significant primary outcome.
Urgent HF events, HF hospitalizations, death, were also all non-significantly different.
You’d guess that this would be easy for the FDA: a pricy invasive device that did not reduce clinical outcomes in a multicenter randomized controlled study (RCT). You could also add that even if you want to ignore the P-value, the reduction of 12% was modest and it was driven by the softer endpoint of HF hospitalizations, not mortality. But no, it was not to be. GUIDE-HF actually underpinned the FDA approval of an expanded indication, an indication that is expected to expand device availability to 1.2 million more Americans. I know, you are puzzled.
The authors provided a “pre-covid impact analysis of the primary endpoint and components.” I am not joking. They say they did a prespecified COVID-19 sensitivity analysis to compare events before COVID and during COVID. (I am not sure how this could be prespecified because the trial started in March 2018).
The issue, the authors say, is that during the pandemic, event rates decreased in the control arm, and that narrowed the treatment effect. And the pandemic affected 28% of total follow-up time.
They also say this analysis “suggested” a treatment interaction; however, the P-value for the treatment interaction was 0.11—way above the 0.05 threshold for significance.
The authors tell us that this special analysis was endorsed by regulatory agencies like the European Medicines Agency (EMA), and FDA.
And voila... this special analysis made statistical significance. An HR indicated a 19% reduction, the CI ranged from 0.66 to 1.00, and the P-value was 0.049. The effect was driven by a 24% decline in HF events (P = .014), with no significant contribution from mortality. The makers of the device said in a press release: “Results of the GUIDE-HF trial supported the expanded indication.”
Comments. I am dismayed by this analysis, the regulatory agencies allowance of it, and the expanded approval of the device.
First, the obvious issues with the “special COVID analysis.” I see it as a post-hoc analysis that at least gives the appearance of fishing for a positive result after the main analysis. While I feel for trialists conducting trials these days, the pandemic is nobody’s fault. I find this an extremely dubious analysis. I don’t find the authors’ explanation persuasive. Why would the control arm be more affected than the treatment arm? And if you are going to posit an HTE, why isn’t the interaction P-value significant.
Second point: you have a clearly nonsignificant result in the primary analysis. And even if it were significant, the effect size was small and driven by HF hospitalizations.
Third, even if you let the COVID post-hoc analysis pass, it barely meets significance, and again is driven by HF hospitalizations.
Fourth: I cannot find where the authors tell us total hospitalizations. This is crucial because these are older patients with co-morbidities. Reducing HF hospitalizations is only an important outcome if it substantially reduces total hospitalizations.
Fifth problem: this is a single-blinded trial. This means clinicians at enrolling centers knew the treatment assignment. Hospitalization is a clinician decision. The clinicians who participated in this trial are most likely proponents for the device. This is not nefarious. It is normal subconscious human bias.
When this trial came out, and Lancet allowed the spin, I said this sort of thing tempted me to be cynical. Now the FDA has also allowed the spin. It’s depressing that this happens in clinical science.
I hear some of you say, who cares if the FDA approves it? Doctors can look at the data and see that there was very little benefit. Doctors don’t have to use it. But here is what happens in real life: Abbott revs up the marketing machine. Hospitals and docs who decide against doing it, because it’s low value, will struggle against the fear of missing out (FOMO). There will be pressure to use it because it’s new and invasive, and all the cool kids are using it. The HF department across town will be doing ads about how they are using it.
This FOMO thing happens in electrophysiology all the time: new devices, catheters, procedures, get huge marketing support. Key opinion leaders use these devices, post their experience on social media, discuss the experiences at meetings, and then, if you aren’t using this new thing, you are less cutting edge.
Finally, always remember that overuse of low-value care exacerbates underuse of high value care and it surely exacerbates the disparities so dominant in US healthcare.
The US Preventive Services Task Force (USPSTF) has published a draft document on the use of statins to prevent disease that is yet to occur. The jargon term for this is primary prevention. The document updates its 2016 document. It’s a massive meta-analysis and systematic review including more than 85,000 patients.
I like the USPSTF because they are independent judges of evidence. They conclude with moderate certainty that a statin will clinically benefit adults aged 40 to 75 years without heart disease (HD) and one major risk factor and a 10-year risk score of 7.5%. They go on to say that the net benefit of a statin is “at least small” for those in whom the 10-year risk is 7.5 to 10%.
Steve Stiles has a nice news summary. He points out that the USPSTF recommendations vary some from the American Heart Association (AHA) and American College of cardiology (ACC) 2018 guideline. But I would say not that much. Some key points:
USPSTF found insufficient evidence to guide primary prevention statin use in elderly individuals. Again, as I often tell older people, the point of preventive therapies is to be old. Once you are old, there is clearly less benefit.
The USPSTF and AHA documents also diverge on use of coronary artery calcium (CAC) scoring in primary prevention. CAC is not part of the USPSTF recommendations, but it figures prominently in the 2018 guideline as a "tiebreaker" for physicians and patients on the fence about whether to start the drugs. Here is what the impartial judges write about CAC:
A 2018 USPSTF review found that the addition of CAC score for risk assessment can improve both discrimination (improvement in the C statistic and reclassification.
However, it noted that the CAC score could also result in reclassification in individuals who do not experience cardiovascular events into higher risk categories, potential harms related to low radiation exposure from use of computed tomography and additional testing, and the absence of studies on clinical effects of risk assessment with CAC.
This is exactly what Andrew Foy and I wrote in our case against CAC in the American Family Physician journal.
The proponents of CAC say it should be used for statin decisions. But I believe they need to be more aggressive about messaging because in the real world, CAC is a super-efficient fear-inducing train to more lucrative procedures for cardiologists and hospitals.
Electrophysiologists call ventricular tachycardia (VT) that is difficult to stop a VT storm. The analogy works.
JAMA Cardiology has published a sham controlled RCT from the UPenn group looking at autonomic modulation to stop VT storm. The specific intervention is transcutaneous magnetic stimulation (TcMS), which can noninvasively and nondestructively modulate nervous activity in many disease states. It likely works by reducing sympathetic innervation.
This trial was conducted at UPenn and enrolled patients with at least three episodes of sustained VT in the preceding 24 hours. Patients were randomly assigned to a single session of TcMS or sham stimulation. The patient and treatment team did not know the treatment. The primary endpoint was freedom from sustained VT in the day following randomization.
The TcMS device was placed close to the left stellate ganglion in the left neck. For patients assigned to receive TcMS, stimulation was delivered at 80% of the local motor threshold and frequency of 0.9 Hz for 60 minutes. For patients assigned to receive sham stimulation, sham therapy was delivered for 60 minutes.
26 patients were randomly assigned (remember, VT storm isn’t common).
During the 24-hour period following randomization, sustained VT recurred in four of 14 patients (29%) in the TcMS group and in seven of 12 patients (58%) in the sham stimulation group. The difference was not statistically significant
Secondary endpoints favored the treatment arm:
The mean number of episodes of sustained VT after 72 hours was 4.5 in patients in the TcMS group compared with 10.7 in patients in the sham stimulation group
They also looked at physiologic effects and found that among those not receiving vasopressors, four patients in the TcMS group and no patients in the sham stimulation group had evidence of peripheral blockade.
None of these four patients had sustained or nonsustained VT within 24 hours after randomization.
There were no safety issues.
Comments. Technically, the primary endpoint did not reach statistical threshold of freedom from VT in the first 24 hours. But there were signals in the secondary endpoints that this non-invasive way to modulate the autonomic nervous system may hold promise for these patients. Emphasis here on its non-invasive nature.
The authors write: “Transcutaneous magnetic stimulation is a well-developed technique that exploits the plasticity of a patient’s nervous system and can promote excitation or inhibition depending on stimulation parameters.”
I laud the UPenn team for systematically studying this. You all might think I am going soft because the primary endpoint was null, and the authors point to secondary endpoints, which favor the treatment arm. But I think this is different thing. VT storm is a terrible condition. Sometimes, often perhaps, it’s a terminal event, and palliative care would be the best option, but sometimes VT storm is survivable and even though patients are sicker than heck, they can survive for years afterward. If we could have a non-invasive way to treat this condition, it could be a huge advance. Plus, I think we are in our infancy of knowledge about using neural modulation for treatment of arrhythmias.
I’d say this technique warrants further investigation in larger trials. The other thing to say about VT storm is it’s a little like studying a viral infection: outcomes occur in days not years, as with most cardiovascular (CV) trials. So, trials seem doable to me.
Fish Oil and AF
The topic of fish oil and atrial fibrillation (AF) is coming up almost every day I am in clinic. Sometimes more than once. My unofficial poll is that fish oil has overtaken aspirin for the drug I most often deprescribe.
Circulation published a meta-analysis of fish oil trials in late 2021 that found a clear increase in AF risk. TheHeart.org | Medscape Cardiology has published the editorial from Michelle Samuel and Stanley Nattel. It is excellent.
I have likely covered this topic previously but due to the clarity of the data, and the commonness of fish oil, I think it’s worth a brief revisit.
The first thing to say is that if fish oil had clear CV benefits, it’s possible an increase in risk of AF would be worth it. But fish oil doesn’t have clear CV benefits. One trial, REDUCE-IT, was positive, but others have not replicated this finding. Yes, there is infinite discussion about dosing, components of the capsule, and choice of placebo controls, but the short story is there are no clearly discernible benefits. That is why the increased risk of AF is important. The authors of the meta-analysis found seven studies that included 80,000 patients. The follow-up was 5 years.
The use of marine omega-3 fatty acid supplements was associated with an increased risk of AF (n=2905; P=0.013).
In analyses stratified by dose, the HR was greater in the trials testing >1 gm/day (HR, 1.49; P=0.042) compared with those testing ≤1 gm/day (HR, 1.12; P for interaction <0.001).
In what’s called a meta-regression, the HR for AF increased per 1 gm higher dosage of omega-3 fatty acids dosage (HR, 1.11; P=0.001).
I can’t explain why this is. But it’s a clear signal. So in patients with AF or at risk of AF, without monster levels of triglycerides, I’ve recommended that they save the money. I say: eat fish not fish pills.
© 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: Feb 25, 2022 This Week in Cardiology Podcast - Medscape - Feb 25, 2022.