Oct 22, 2021 This Week in Cardiology Podcast

John M. Mandrola, MD


October 22, 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 October 22, 2021, John Mandrola, MD comments on the following news and features stories.


COVID news continues to get better. US cases, hospitalizations, and deaths are trending down.

I’ve decided to stop Tweeting and discussing COVID-19. As someone who had enjoyed critical thinking, I had thought the pandemic was a great canvas for rational thinking, kind of like a mountain pass is great for bike training. But the polarization is just too much. When you can lose friends or a career over an interpretation of data-points, pragmatic John says it’s time to simply wait it out.

I do not know, but I hope that the combination of natural immunity and vaccination keeps cases down, and then normality returns. We can get back to arguing about PCI in stable coronary artery disease.

Speaking of normality, I traveled to Barcelona Spain to participate as faculty in the European Association of Cardiothoracic Surgery (EACTS) meeting. Thank you to the EACTS leadership for the invite. After nearly 2 years of Zoom, it was incredibly nice to meet colleagues in real life, and it was a true honor to be invited to give a Neutral Martian view of the recent trials. And thanks to Professor Whitlock for helping me set up my Garmin running watch.

Aspirin for Primary Prevention

Let’s first talk about aspirin (ASA) for primary prevention, which made news recently when the US Preventive Services Task Force (USPSTF) posted a draft statement that people aged 40 to 59 years should decide with their clinicians whether to start taking ASA and people older than 60 years should not start taking ASA.

The New York Time led coverage with the headline: “Aspirin Use to Prevent 1st Heart Attack or Stroke Should Be Curtailed, U.S. Panel Says.” Many print and TV news outlets covered it and a handful of patients asked me about the ASA news. What’s weird about this story is that most of the “new data” was published in 2018. 2018! In pandemic time, 3 years ago was a totally different era.

I am a huge fan of the USPSTF because they get as close as possible to a Neutral Martian view of evidence. And as part of the new recommendations, they also published the draft of a systematic review and meta-analysis of the ASA data. But first let’s revisit the 2018 big trials, which TWIC covered at the time.


  • Approximately 12,000 patients, mean age ≈ 64, 5-year follow-up, mean American College of Cardiology (ACC) 10-year risk, 17%, which is super high.

  • Primary endpoint (PEP; cardiovascular death [CVD], myocardial infarction [MI], unstable angina [UA], stroke, transient ischemic attack [TIA]): 4.29% in ASA vs 4.48% in placebo; hazard ratio (HR) 0.96, confidence interval (CI) 0.81-1.13.

  • GI bleeding: 0.97% in ASA vs 0.46% in placebo; HR 2.11 (CI 1.36-3.28) P = 0.007.

*Despite a 17% 10-year risk of events, the actual event rates were low at 4%, so the authors felt that despite trying to enroll a high-risk adult population, they actually enrolled a lower-risk group. My take: in this group, ASA had no significant effect on CV events but doubled the risk of GI bleeding.


  • Approximately 15,000 patients; mean age ≈ 63 years; 7-year follow-up

  • PEP [CVD, MI, stroke, TIA]:8.5% in ASA vs 9.6% in placebo; rate ratio (RR) 0.88 (CI 0.79-0.97), P = 0.01

  • Major bleeding: 4.1% in ASA vs 3.2% in placebo; RR 1.29 (CI 1.09-1.52), P = 0.003

*ASA prevented serious vascular events but it also caused major bleeding events. The absolute benefits were largely counter balanced by the bleeding hazard.

My take: the diabetics had a higher event rate, and while the CV event rate was lower, it was nonfatal events that drove the reduction, as vascular death was no different. The reduction in nonfatal events was countered by greater increase in major bleeding.

ASPREE – One trial, two papers, one on all-cause mortality and the other on CV events and bleeding with ASA.

  • About 19k patients; mean age ≈ 74 years; 4.7-year follow-up

  • PEP (CV events): 10.7 vs 11.3 per 1000 patient-years in the ASA vs placebo groups, respectively, HR 0.95 (CI 0.83-1.08)

  • Major bleeding: 8.6 vs 6.2 per 1000 patient-years in the ASA vs placebo groups, respectively, HR 1.38 (CI 1.18-1.62) P < 0.001

  • Overall death: 5.9% vs 5.2% in the ASA vs placebo groups; 0.7% ARR translated to an HR of 1.14 (CI 1.01-1.29)

  • The higher death rate looked to be driven by cancer related deaths.

Authors: ASA as primary prevention in older adults resulted in a significantly higher rate of major bleeding and did not result in significantly lower CV events. And in the separate paper on death, “higher all-cause death was observed among healthy older adults who received ASA. was attributed primarily to cancer-related death. In the context of previous studies, this result was unexpected and should be interpreted with caution.”


For CV events, they included 13 clinical trials –

  • Major CVD events: the HR was 0.89 (0.85-0.94)

  • For CVD: the HR was 0.95 (0.87-1.04)

  • For Major bleeds: the HR was 1.45 (1.33-1.58)

  • All-cause death: the HR was 0.97 (0.93-1.02)

When I tweeted out that this all was old news, dating back to the 2018 trials, trial expert Sanjay Kaul responded that the evidence against ASA use in primary prevention was even older than 2018. Kaul sent me a link to an FDA report from 2014, that said:

“The FDA has reviewed the available data and does not believe the evidence supports the general use of aspirin for primary prevention of a heart attack or stroke. In fact, there are serious risks associated with the use of aspirin, including increased risk of bleeding in the stomach and brain, in situations where the benefit of aspirin for primary prevention has not been established.”

My colleague Melissa Walton-Shirley wrote an editorial on ASA that had two themes: one was that patients with high coronary artery calcium (CAC) scores may still benefit from ASA, and that the major network coverage may have led many patients to stop taking ASA without consulting with their clinicians.

My takes:

  • Walton-Shirley is correct, this news applies only to primary prevention. Patients taking ASA after a vascular stroke, or MI or PCI, should stay on antiplatelets.

  • It is clear that ASA reduces CV events by about 10% in relative terms, but this translates to very small absolute risk reductions. So small that it is not enough to reduce CV death or overall death.

  • It is also incredibly clear that ASA increases the risk of major bleeding, by a greater amount than it reduces nonfatal CV events. Yet, this risk increase is also small in absolute terms.

  • And the signal of a 14% increased risk of death in the healthy elderly, in ASPREE, is worrisome. Keep in mind that the upper bound of that HR was 1.29, indicating that ASA could have increased death by 29%.

  • The contemporary studies are very strong. I put more weight on them because they studied patients in this era with modern background therapy. Each of the trials accrued many events and gave similar signals.

  • The USPSTF meta-analysis reveals little heterogeneity.

As for saying having high CAC will confer a high-enough risk, that there is more benefit than harm from ASA, I think that is plausible, but highly speculative. Remember, the 10-year risk of a CV event in the ARRIVE trial was 17%--way above the statin threshold of 7.5%. And that study found no compelling net benefit. A mere 4% relative risk reduction. Neither did ASCEND or ASPREE.

I suspect that whether or not CAC produces a heterogenous treatment effect of ASA might be gleaned from the data. I may be wrong here, and trialists, please correct me, but let’s say the authors pooled their data and studied those who had high CAC scores, you might get a sense of a signal. Of course, this would be a post-hoc test, and it would susceptible to the play of chance. Never forget that one’s horoscope sign produced spurious signals from ASA in the ISIS-2 trial. But Dr Walton-Shirley is correct in saying that if a patient with high CAC and low bleeding score wanted to take ASA after a discussion with his or her doc, then that is fine.

As we always discuss on this podcast, there are specific and larger lessons. The specific lesson here is that the net benefits of ASA in those yet to have an event is dubious. The larger lesson is that it is super hard for preventive therapies to substantially improve health in those without disease. We clinicians are at our best when we treat those who seek our help because they are sick. Due to unintended consequences—here, major bleeding preventive therapies have a much harder time delivering a net good for people.

Cooling for Cardiac Arrest

Even though this topic doesn’t bring clicks and attention, I love it because it teaches larger lessons about the sclerosis of evidence translation and the power of RCTs. JAMA has published yet another RCT looking at induced hypothermia after return of spontaneous circulation (ROSC) in patients with out of hospital cardiac arrest. It’s called the CAPITAL CHILL trial.

Before I tell you the results of this comparison of target temps of 31 vs 33 degrees Celcius in survivors of cardiac arrest, let me tell you results of recent trials of induced hypothermia, then you might be able to guess the results.

  • In 2013, the TTM trial compared 33 degrees to 36 degrees in 975 post-cardiac arrest survivors. There were no differences in death or neurologic function. Plus, the cooling arm had more adverse events.

  • In 2021, the TTM2 trial compared 33 degrees to normothermia in 1850 patients, and the results were clear: no difference in death nor neurologic outcomes. There were more arrhythmias, greater use of paralytics, and longer time on the ventilator in the cooling arm.

The New England Journal of Medicine (NEJM) published both these trials. These clearly establish that 33 degrees is no better, probably worse, than normal temp.

Now to CAPTIAL CHILL, which compared 31 vs 34 degrees in almost 400 patients. Well, if 33 was no better than normal temp, what do you think 31 degrees will do? You guessed it: no difference in the primary endpoint of mortality or poor neurologic outcome. In fact, 48% had a PEP in the 31-degree arm vs 45% in the 34 degrees arm. My friends, I think we can now say there is no longer equipoise for induced hypothermia. Cooling simply does not produce better outcomes. The larger lesson is that even though cooling makes sense, the evidence shows it does not help. And since it requires more effort, more nursing attention, and TTM clearly had higher adverse events, we should de-adopt this strategy. You could argue it could have been de-adopted in 2013 after TTM1. Once again, the RCT shows its power to inform us how best to care for patients. We mustn’t ever be scared to test our favorite strategies in the RCT.

Half-dose DOACs

The group led by Dr. Andrea Natale have long provided us with important information regarding anticoagulation after electrophysiologic procedures. I remember when I started ablating atrial fibrillation (AF) nearly 20 years ago, it was the Natale group that first recommended uninterrupted warfarin. This was provocative, early, and correct.

This month, JACC Interventions has published a study looking at a novel strategy to prevent device-related clots after Watchman implantation from this same research group. This was a non-randomized study enrolling consecutive patients who had Watchman at three centers over a 5-year period early in the Watchman experience—from 2014 to 2019.

The two strategies were standard antithrombic therapy (SAT) with novel oral anticoagulants (NOAC) plus ASA for 45 days, then ASA plus clopidogrel for 180 days, then ASA vs half-dose direct oral anticoagulants (hdDOAC) plus ASA for 45 days, then half-dose DOAC alone. Crucially, the choice of these two strategies were at the discretion of the clinician. That is a huge point and one that I will come back to.

The primary efficacy endpoint was a composite of both safety and efficacy endpoints: 1) thromboembolic (TE) events (ischemic stroke, TIA, peripheral thromboembolism); 2) device-related thrombosis (DRT); and 3) major bleeding events (severe or life-threatening, and moderate bleeding).

The results strongly favored the half-dose group. After a median follow-up duration 13 months:

DRT occurred in 3.4% of the SAT group vs 0.0% in the hdDOAC arm; log-rank P = 0.009).

  • The risk of nonprocedural major bleeding was 0.5% vs. hdDOAC group vs SAT.

  • The rate of the primary composite endpoint of DRT and TE and major bleeding events was 9.5% in SAT patients and 1.0% in hdDOAC patients (HR: nearly 10).

The authors concluded that clearly half-dose DOAC is the better strategy but acknowledge that the main limitation of the study was the nonrandomized choice of strategy.

Although the baseline characteristics table did not reveal major differences, regular listeners of this podcast know that many factors determine how a clinician chooses a therapy. Some of these factors come out in a spreadsheet but many of the intangible factors do not. This is why randomization is so important—because it balances all important co-variates, those that go on a spread sheet and those that do not.

What’s more, and this was pointed out in the editorial, the time course of therapies varied—more of the SAT were done early on and more of hdDOAC group were recruited later, which can also bias the results, because there is a learning curve to this procedure.

These are major caveats. But I agree with my colleague Jon Piccini, whom Steve Stiles quoted in the news story, that these are interesting and plausible findings. Half-dose DOACs may indeed be a better DRT prevention strategy than dual antiplatelet therapy (DAPT). The best way to test that is with an RCT, which is planned.

However, I cannot leave this topic without noting that the obvious: if hdDOAC, prevented DRT, reduced stroke and bleeding, perhaps an arm of any future trial of high bleeding-risk patients should include a half-dose DOAC alone, because if this cocktail worked well with a foreign body in the left atrium, it’s possible it could work even better alone.

Never forget the history of Watchman: PROTECT AF did not pass FDA muster; PREVAIL did not meet noninferiority in its first primary endpoint; procedural complications in real world data is 5% to 10%; and roughly 50,000 Watchman implants have been done in the US, and hardly any of these patients have been randomized in proper trials.

This important study on hdDOAC in high-risk patients highlights the tragedy of tens of thousands of patients being exposed to a dubious procedure outside of clinical trials. If we had the courage or incentive to randomize even a fraction of these 50,000 patients, we would know whether or not this invasive preventive strategy works.

Social Media

JAMA Open published a research letter survey on gender differences in physician use of social media. The survey generated about 580 responses; 56% identified as women.

Both men and women agreed that social media helped build their professional networks. Both agreed that social media increased collaboration inside and outside their specialty. And men and women agreed that social media improved their job satisfaction.

  • Compared with men, women physicians were less likely to report that social media use expanded their research portfolio (123 men [48%] vs 117 women [36%]; P = .005) or resulted in a speaking engagement (101 men [39%] vs 95 women [30%]; P < .001) or scholarship opportunity (63 men [25%] vs 69 women [21%]; P = .02).

  • Compared with men, women physicians were more likely to report using social media to build a support network (141 men [55%] vs 233 women [73%]; P < .001), but were less likely to report using it to learn about research (212 men [83%] vs 221 women [68%]; P < .001) or clinical topics (219 men [86%] vs 243 women [76%]; P = .003).

The authors wrote, “The findings of this survey study suggest that social media use by women physicians may not improve gender equity. It may be the case that the same biases that lead to fewer opportunities for professional advancement for women persist in the online physician community, hindering women’s professional advancement.”

Medscape Medical News covered the paper, and I agree with comments from Dr. Vineet Arora. She said she has mixed feelings about social media. That it was a necessity for one’s career, but also that there is negativity and trolling. And, I will add, downright nastiness. I recommend reading the news article because the notion that not engaging at all seems like a bad idea. I mean, there is a ton of great information out there, and oodles of opportunities for networking.

But the pandemic has changed my view of social media. Before 2019, I had traveled all over the place giving lectures on the benefits of social media. Get online. Speak your mind. Be rational, went my argument. Well, the polarization of the pandemic has shattered the notion that rational free-thinking always works out well, especially on Twitter.

I still use Twitter, but the ability to persuade on that platform approaches nil. One of the granddads of social media, Dr Bryan Vartabedian was correct when he wrote that you need a home base. Twitter, Instagram, FB, are like outposts.

A home base could be a blog, a website, a YouTube channel, a podcast. I recently started a substack. Home bases are places where you can, in longform, tell the world about you and make logical arguments. Longform writing, video, and conversation are the future, I think, and they seem much safer places for clinicians and scientists.

As for Twitter, I agree with Drs Arora and Shapiro who said in the news article that simply not responding to provocations is best. Resist the urge. I favor mute over blocking.

I will close with this important message about social media and Twitter: it’s not real. It’s a small sample of vocal people who interact without personal involvement. Real life is in your clinic, at your hospital, at home with your spouse, at your kids soccer games, in the airport when you help an older person get their luggage in the overhead bin. Or on the Appalachian trial when you share a bag of trail mix or a few baby wipes. Never forget that the algorithm of Facebook and Twitter is to provoke and to anger, because that leads to more engagement. The pandemic has reminded me of that. And I try to not fall prey to the algorithm.


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.