Dec 17, 2021 This Week in Cardiology Podcast

John M. Mandrola, MD


December 17, 2021

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.

In This Week’s Podcast

For the week ending December 17, 2021, John Mandrola, MD comments on the following news and features stories.

First, an Announcement

This Week in Cardiology will take a 2-week break for the holidays and return on January 7, 2022.

I want to thank all of you for your support and comments. It means a lot to me that you listen. I speak into this microphone every week and never know who it reaches. Remember, it really helps to give TWIC a rating and review on whichever app you use. Also, when you disagree with me, let me know in the comments section on THO.

Patent Foramen Ovale Closure

JAMA has published a nice meta-analysis of the patent foramen ovale (PFO) closure vs medical therapy trials in patients with cryptogenic stroke. First, some background:

Cryptogenic stroke is more common than you would think. And it’s so weird. Why would this person who is relatively healthy, devoid of vascular risk factors, and in sinus rhythm have an embolic stroke? The goal of finding a cause is obviously to prevent another stroke. Since these patients are usually young, the therapy has to continue for a long time—which is an important factor.

Your thoughts immediately go to the possibility of a PFO and paradoxical embolism of clot from the venous system. Here’s the problem: up to 25% of humans have a PFO, so it is hard to tell if the PFO is the cause or an innocent bystander.

Devices have made PFO closure possible via a percutaneous approach. PFO closure has been studied in six randomized trials. You can break these into the three before 2016 and three after 2016. They have different features and results.

In the pre-2016 trials — CLOSURE 1, RESPECT (early follow-up), and the PC Trial – the short story was that PFO closure trended better but when taken together results were basically null, and in 2016 the American Academy of Neurology recommended against routine PFO closure outside the research setting.

The possible reasons these trials failed to show superiority of PFO closure included a) low event rates leading to low power; b) the inclusion of patients with TIA, and older patients with competing causes of stroke that would be unaffected by PFO closure; c) some trials also had mixed medical arms that included oral anticoagulation (AC) instead of just aspirin.

Three newer trials were more selective and kept medical therapy as aspirin. Gore REDUCE, CLOSE, RESPECT (longer term follow-up), and DEFENSE-PFO all found PFO closure to be superior to aspirin. Some caveats: Event rates are low and absolute risk reductions (ARR) are small. Two trials had significant amounts of missing data, so you wonder about the play of chance. Further, the most positive trials (the ones with the larger shunts) also had the least use of AC. Thus, one is tempted to wonder whether the later trials were positive because of the better selection of patients more likely to benefit or were they positive because aspirin is inferior to oral AC.

Now to the JAMA study. The idea was to first take individual patient-level data from the trials and put them together to get an overall treatment effect. That’s not novel. It’s been done before. The novel part was then to test two risk scores for determining the likelihood of PFO being the cause of stroke (the Risk of Paradoxical embolism [RoPE] and PFO-Associated Stroke Causal Likelihood [PASCAL] scores) and test whether these scores produced heterogenous treatment effects.

I realize heterogeneous treatment effect is jargon, but the idea is simple. A trial includes patients with tons of individual characteristics—men, women, younger, older, large PFO, smaller PFOs, diabetes, no diabetes, etc, etc. Then the trial produces an average effect size.

  • In the case of the PFO vs meds, the meta-analysis of all the trials found that the yearly incidence of stroke with medical therapy was 1.09% and with device closure was 0.47%; P < .001. A 59% reduction.

  • So, you could say PFO closure reduces the risk of stroke by 59% in relative terms and about 0.5% per year in absolute terms.

But now you want to know if the patient in front of you is likely to get that average benefit or perhaps more or perhaps less.

This is where the authors tested two prediction scores:

RoPE is a 10-point scale — the higher the score, the more likely the PFO was to cause the stroke. For instance, you get five points if you are 18 to 29 years old, vs only one point if you are 60 to 69 years. And you get one point for being nonsmoker or having a cortical infarct and one point for not having a history of hypertension of diabetes. The main limitation of the RoPE score is that it did not include any features of the PFO itself.

The PASCAL system takes the RoPE score and adds echo-derived shunt size and presence of an atrial septal aneurysm. PASCAL reports three categories of PFO causality in the stroke: unlikely, possible, and probable. Let’s look at how these two scores affect results.

  • For patients with a RoPE score less than seven, the hazard ratio (HR) of PFO vs medical therapy in all the trials was 0.61 (0.37-1.00).

  • For RoPE score of seven or above, the HR was much greater at 0.21 (0.11-0.42).

  • In other words, the relative risk reduction was greater for those with clinical features suggesting PFO was causal. The P-value for interaction was significant at 0.02.

However, that’s relative risk reduction. You also have to look at ARR.

  • The ARR in RoPE scores less than seven was 1.1% with confidence intervals (CI) ranging from -0.9 to 3.1, so not significant.

  • For higher likelihood RoPE scores seven or higher, the ARR was 2.1%. Both favored PFO closure, but the CI for lower ROPE scores crossed 0.

Now let’s look at the results using PASCAL— an arguably better score because it takes into account both clinical and anatomic and functional characteristics of the PFO.

  • In patients graded as unlikely there was no reduction of PFO vs med. HR right at the 1.0, no effect line.

  • For patients graded as possible , the risk reduction of PFO closure was significant with an HR of 0.38 (0.22-0.65), a 62% reduction.

  • For those graded as probable , the HR was 0.10 (.03-0.35) or a 90% reduction.

Now let’s look at how the unlikely-possible-probable PASCAL score affects ARR.

  • For the unlikely patients, the ARR with PFO was null.

  • For possible and probable patients, the risk reduction was a statistically significant 2.1%.

Before I discuss this there is another important aspect of the study: Tthe risk score also predicted a harm of PFO closure, namely atrial fibrillation (AF). It is well known that PFO closure increases the probability of AF, most but not all of which, occurs in the short-term.

The absolute risk increase of post periprocedural (occurring more than 45 days after randomization) AF with a device was 4.41%, 1.53%, and 0.65% in the unlikely, possible, and probable PASCAL categories, respectively. In other words, those least likely to benefit based on unlikely scores were also more likely to get AF after 45 days.

Comments: I love this study and the idea of “personalized” evidence-based medicine. Trials are important because they show us average treatment effects in a large group of patients. But within that trial, there are going to be patients who benefit more or less than that average.

In reviewing the methods of this trial, I came across a superb review from primary author David Kent from Tufts. In that review, Kent and co-authors explain the limits of the usual way of looking for heterogeneous treatment effect, which is subgroup analyses based on a single feature — say gender, age, diabetes or no diabetes. They explain that these are fraught with problems and highly likely to deliver false positive results, mainly because trials are barely powered for treatment effects of a whole group.

But in this paper in JAMA they have individual patient-level data from six trials, and they explored heterogeneous treatment effect in not one characteristic but in biologically plausible and previously validated risk scores. Basically, they looked at results of PFO vs medical therapy based on what the attributable fraction of the PFO was to the original stroke. Patients with a high attributable fraction of PFO did better with closure.

Now there are limitations – stroke after cryptogenic stroke is a very low-incidence event. ARR are small, in the order of 1% to 2% and there is substantial missing data regarding the prevention of disabling stroke.

On Twitter, I heard from some medical statisticians who were worried about the dichotomization of, say, the RoPE score at seven. Their’s was a complicated criticism that centered on how to combine within trial and across trial variation. A variant of the Lord’s paradox.

The other limitation has nothing to do with the analysis but has to do with the best choice of medical therapy. Since the likely mechanism of stroke is paradoxical venous clot, and the clear best treatment for venous clot is oral AC, an argument can be made for oral AC as the proper comparator. You wonder how a trial of PFO closure would do against direct oral AC.

There is a great editorial from Robert Yeh and Doreen Defaria Yeh where they outline the ideal scenario when a patient presents with cryptogenic stroke.

  • A thorough evaluation for other sources of ischemic stroke—especially exclusion of AF.

  • Next is echo imaging of the atrial septum and consideration of sinus venosus or atrial septal defect. Make sure it is a PFO.

  • Once a PFO is found, the next step is to characterize the proclivity for paradoxical embolism.

  • Then, and I think this would key, put all this data together during a multi-disciplinary meeting with neurologists, hematologists, cardiologists, echocardiographers, and interventionalists.

  • The summary of the findings could then be brought to the family for shared decision making.

It's a tough call but the addition of these risk scores to better define the likelihood of benefit of device over aspirin.

Transcatheter Mitral Procedures

Remember the saying: Science tells us what we can do, trials tell us what we should do, and registries tell us what we are doing.

JACC has published the latest update from the Society of Thoracic Surgeons–American College of Cardiology Transcatheter Valve Therapy (STS-ACC-TVT) Registry. The registry included data from 2014 to 2020.

  • It comprised about 34,000 transcatheter edge-to-edge repairs (TEER) and 3600 transcatheter mitral valve repairs (TMVR).

  • TEER procedures have gone up 10-fold since 2014.

  • Most of the procedures are for primary mitral regurgitation (MR), however, there has also been an increase in use for secondary MR; from 7.6% in 2016 to 21% in 2020.

  • For TEER, in 2014, 30-day mortality was 5.6%; from 2015 to 2019, the 30-day mortality ranged from 4.1 to 4.8%. Not much changed.

  • Over the study period, there were no significant changes in rates of stroke (1.3% overall), mitral valve re-intervention (1.1%), new onset postprocedural atrial fibrillation (3.0%), new dialysis requirement (0.9%), acute kidney injury (1.8%, inclusive of all stages), vascular complications (0.5%), or major/disabling bleed (4.0%)—again, surprising that these things haven’t improved over time.

  • One limitation of this registry is that there was a fair amount of missing data. At 1 year, 25% of the patients’ mortality status were unknown.

Just a couple of comments. 36,000 procedures is a lot. I went back and looked up the evidence base: for TEER in primary MR, the EVEREST trial comparing TEER to MV repair had 279 patients. For secondary MR, the positive COAPT trial had 614 patients, and the null MITRA-FR had 300.

Is that enough? I am asking. I would have thought there’d be more. I looked at the guidelines and the only citation for TEER for primary MR was EVEREST. TEER was actually the inferior procedure compared with surgery, but it had fewer major adverse events.

I am sure percutaneous procedures help selected people, but as an outside observer, it seems like a pretty weak supporting evidence base. I am happy to be corrected, any structural listeners out there, let me have it in the comments. The medical conservative view would be to have strong evidence for benefit.

Yoga and Vasovagal Syncope

JACC-Cinical Electrophysiology has published a small but really cool study looking at yoga as therapy for vasovagal syncope (VVS). The authors are from New Dehli, India; first author Gautam Sharma.

While I am terrible at yoga, I think it is a fantastic and uber-healthy pursuit. Did you know, for instance, there are two RCTs of yoga for the treatment of AF? Both were positive. The mind-body connection is hugely under-rated in medical practice.

The study was fairly simple and elegant. Patients with documented VVS were randomly assigned to yoga or standard care. The primary endpoint was the number of episodes of syncope and presyncope at 12 months.

  • The mean number of syncopal or presyncopal events at 12 months was 0.7 ± 0.7 in the intervention arm compared to 2.52 ± 1.93 in the control arm (P<0.01).

  • Quality of life at 12 months showed significant improvement of all Syncope Functional Status Questionnaire scores.

Regarding the yoga intervention, here is one of my favorite sentences from a medical journal: “The (yoga) module was further validated by 5 yoga experts of national repute.”

There were eight supervised yoga sessions in the first 2 weeks, after which patients had to continue practicing yoga at home for at least five daily sessions in a week. There were two supervised follow-up sessions in the second month and then one guided session per month till the sixth month.

The authors include a beautiful description of why yoga might benefit patients with vagal-mediated syncope. It could work through central or peripheral mechanisms. Here’s some possibilities:

  • Yoga-enhanced vascular and muscular tone, especially in lower limbs, not only blunts the venodilation phase of a syncope episode but may also accelerate the venous return.

  • Yoga breathing and relaxation techniques have been shown to increase vagal tone and improve autonomic balance, which could potentially curtail the sympathetic overdrive phase and interrupt the activation of the c-mechanoreceptors, which is a critical step in the syncope cascade.

Here's the thing; it doesn’t matter how it works. What matters is that it appeared to work, and right now, there are no great options for young people saddled with this condition

Of course, this study had oodles of limitations: it was open-label, compliance was self-reported. Yoga in India may also be of higher quality than in middle America.

Year-end Review | Medscape Cardiology posted my picks for the top 10 stories of 2021. Take a look and see what you think. Use the comment section to tell me what you liked and what I missed. I’ve received some positive reviews and some negative ones — see what you think. Surely there were more than 10 important stories.

Also check out the year-end summary from real academics Bob Harrington and Mike Gibson. They are smart and have a wealth of trial experience.

See you all next year. Happy Holidays.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.