Jun 18, 2021 This Week in Cardiology Podcast

John M. Mandrola, MD


June 18, 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.

In This Week’s Podcast

For the week ending June 18, 2021, John Mandrola, MD comments on the following news and features stories.

COVID Vaccine Myocarditis

Another week has brought more good news. COVID cases and hospitalizations are trending lower despite the relaxing of social distancing and masking. I read that the European Union will now allow US tourists. Even COVID Twitter seems less anxious—which is saying a lot.

But there is more news on the vaccine-induced myocarditis story. This week, Circulation published a series of 8 patients, mostly young men, mostly second shots, who had severe myocarditis. A number of these patients required ICU admission. All of them had high levels of troponin and abnormal scans. Along with that study, came an American Heart Association (AHA) press release in which they essentially say, no worries, the Centers for Disease Control and Prevention (CDC) is investigating, but in the meantime don’t delay your COVID shot.

In their press release they make a misleading statement: “According to the CDC, fewer than 1,000 cases of myocarditis-like illnesses were reported as of May 31, 2021, and nearly 312 million doses of COVID-19 vaccines have been administered in the U.S. to-date.” (As of May 31, less than 10% of the total doses of vaccine in the United States have been given to young adults.) What is misleading is the denominator. Just as it was with the clotting issues in young women with the adenovirus-based COVID vaccine, we mustn’t use the total number of shots given, but the total number of shots given in the at-risk cohort of young men.

The CDC was to hold a special meeting today to review the evidence of this issue. (It was postponed due to the newly declared Juneteenth holiday). It seems wise to wait for this meeting, look at the data, discuss the data, consider the extreme low-levels of circulating cases in most communities, and take a conservative—not aggressive—stance to vaccinating young people. This gives time to consider key questions, such as should we use only one dose in young people or those who have had SARS-COV2 infection.

This discussion should not be construed as vaccine hesitancy. The incredible efficacy of the mRNA vaccine in vulnerable people is a core factor in my caution for vaccinating young folks: to wit, literally every adult who wants a shot can get one, and that essentially makes that person bullet proof—so the unvaccinated pose essentially no threat to a vaccinated person. As I said last week, the risk of myocarditis from vaccine remains low, but for the young, risk of severe COVID is also very low. There is room here for discussion and collection of more data.

Christian Eriksen’s Cardiac Arrest

The worlds of cardiology and sports collided this past week when Danish footballer Christian Eriksen suffered cardiac arrest on the field. He promptly received cardiopulmonary resuscitation (CPR) and a shock from an automated external defibrillator (AED), and regained consciousness. The team doctor said it so well: “He was gone.” Resuscitation of sudden death in a young person is truly glorious.

I’ve reviewed many of the news reports. When a 29-year-old athlete has cardiac arrest, the differential diagnosis includes hypertrophic cardiomyopathy, idiopathic cardiomyopathy, coronary malformation, and myocarditis. A BBC news report quoted the famous London sports cardiologist Sanjay Sharma who said that he had put Eriksen through a battery of cardiac tests before 2019 when Eriksen was at another team. Sharma says he did not have structural abnormalities, at least before 2019. Sharma also said his Milan team had likely done the same. None of the news reports that I could find have identified a cause, but Thursday it was announced that he would have an implantable cardioverter defibrillator.

A couple general thoughts:

  • Though this was a dramatic case because so many millions watched, I regularly see people saved by bystander CPR and use of AEDs. I hope Eriksen’s save advances the cause for more citizens to learn basic CPR and for more institutions to invest in life-saving AEDs. This is an uncontroversial public health measure.

  • Here is a man who had perhaps the most extensive screening anyone could imagine and still had cardiac arrest on the pitch. This doesn’t mean screening is useless, but like Wilson and Jungner famously wrote: screening for disease is admirable, but in practice, there are snags.

  • A snag here is that the best screeners in the world could not protect Eriksen from having a cardiac arrest. This is the problem of false negative: the tests Eriksen had did not detect an underlying propensity for cardiac arrest, and this is not at all uncommon.

  • Another snag of screening athletes is that the better we get at treatment, the less valuable early detection through screening becomes. Imagine a world where every sports field has an AED, or all citizens are skilled in CPR. The corollary here in cardiology and oncology is that the better we become at disease treatment, the less important is early detection.

  • A third snag with screening is over-diagnosis: most young athletes aren’t seen by professors like Sanjay Sharma: they are screened by local cardiologists who are not accustomed to the ECG and ECHO patterns of elite athletes. And I’ve seen plenty of young people exposed to unnecessary cascade testing.

My final thought would be whether Eriksen will get a transvenous ICD or subcutaneous ICD (S-ICD). Bogdan Enache and I have argued in JACC Clinical Electrophysiology that the recent Class 1 FDA S-ICD lead and generator recalls combined with the dubious results of the PRAETORIAN trial argues against continued use of the S-ICD. I have stopped recommending S-ICD due to these issues. I will link to our article in JACC EP. I’d appreciate any feedback on it.

A Huge Reversal in the ICU

Fortunately, prompt treatment of Christian Eriksen meant that he did not need ICU care. But lots of patients with out-of-hospital cardiac arrest suffer the effects of prolonged cardiac standstill. Right now, the current practice is to cool these patients to 32 to 36 degrees Celsius. This is a labor-intensive and quite invasive intervention. Recall that all our cells in the body are optimized to work at 37 degrees.

Cooling as an intervention was established on very low-quality evidence. Pulmonologist Josh Farkas has an excellent summary on the PulmCrit blog. It all began with small unblinded trials that reported that cooling was associated with massive absolute reductions in mortality. These findings were then combined with the notion that circulatory arrest survivors often have fever, which could be causal in the poor outcome. Of course, fever is a surrogate and may also be just a marker of severe illness. Hypothermia then becomes codified in guidelines. Boom—there goes equipoise. Then came the TTM1 trial: 33 degrees vs enforced normothermia at 36 degrees. No differences in outcomes. Farkas said this was enough for him to go to 36 degrees. He writes that in addition to the extra labor and costs, 33 degrees may delay waking up, prolong ventilator time, and cause bradycardia. But the use of 33 degrees persisted, largely he writes, because of status quo bias.

In 2019, another trial of hypothermia (33 degrees vs 37 degrees) for non-shockable rhythms, called HYPERION, found improved functional outcomes in the cooling arm, but the P value was 0.04 and only 584 patients were enrolled. What’s more there was no difference in mortality at 90 days.

This was the background for the Targeted Hypothermia versus Targeted Normothermia after Out-of-Hospital Cardiac Arrest (TTM2) trial, which tested the status quo.

  • 1900 adults who presented with coma after an out-of-hospital cardiac arrest were randomly assigned to hypothermia to 33 degrees or normothermia (less than 37.8).

  • The primary endpoint was the unbiased overall death at 6 months; secondary outcomes included functional outcome at 6 months.

  • In the active arm, the 33-degree arm, the aim was therapeutic hypothermia with cooling initiated early, then slow rewarming and attempts to avoid rebound fever.

  • At 6 months the death rate was not significantly different – 50% vs 48%.

  • There was no difference in functional outcomes at 6 months, and more patients in the cooling arm had arrythmias—24% vs 17%.

The strength of this trial was its size and methods: it had many times the combined enrollment of the earlier and “positive” trials. To limit bias, decisions regarding the withdrawal of life-sustaining therapy were performed after 96 hours by blinded clinicians in a standardized fashion (with this information provided to the primary team).  Families were also blinded.  The specific message of this trial is that hypothermia is a nonbeneficial, aggressive, and potentially harmful intervention for treating patients with coma after cardiac arrest.

Practice should change. Note this doesn’t mean total nihilism but attention to maintaining normal temp as was done in the trial. The larger messages are:

  • We must always be alert to the acceptance of aggressive strategies based on plausibility and methodologically weak studies; the core belief of the medical conservative is that the onus is on proponents of expensive/invasive interventions to prove efficacy.

  • Surrogate measures and plausibility (in this case, temperature and fever) can lead us astray.

  • We must always have the courage to test our beliefs in properly done randomized controlled trials.

Revascularization in Chronic CAD

At the EuroPCR meeting last month, investigators stirred debate with a meta-analysis of studies looking at revascularization vs medical therapy in patients with stable coronary artery disease (CAD). The European Heart Journal (EHJ) published the paper. The inimitable David Brown and William Boden wrote the editorial—a harsh rebuttal.

First of all, it’s unusual for a scientific meeting to label a meta-analysis of anything as “late-breaking.” A word on meta-analysis from a non-statistician: my understanding of a meta-analysis is that it is a pooling of results from multiple studies that address the same question. Ideally, the aggregation of information leads to greater strength of evidence on a question. Let’s say a number of RCTs found trends in favor of a treatment; each one taken alone was inconclusive statistically, either not enough patients or not enough events. A meta-analysis can combine these results and perhaps show the treatment is in fact effective.

But there are huge caveats:

  • By definition, a meta-analysis is a post-hoc retrospective review. I want to quote from a beautiful paper in the Lancet from Thomson and Pocock. They wrote: “The overall treatment effect estimated in a meta-analysis provides a useful guide to some average treatment effect in the trials, but it is a naïve oversimplification to regard it as applying directly to future patients.”

  • You have to be very careful about what studies to include. Studies need be sufficiently similar, lest you are comparing apples and oranges.

  • The other challenge comes in the interpretation of plots. You can’t just look at the summary effect. You have to assess for heterogeneity and weighting; too much heterogeneity adds doubt to the conclusions and if one trial is huge and the others are small, the huge trial drives the results.

The EHJ meta-analysis combined 25 trials of nearly 20,000 patients spanning nearly 50 years; it included the early CABG trials from the 1970-80s.The authors made the case for doing such a meta-analysis on the grounds that it included longer-term follow-up and they specifically look at cardiac death. They pre-specified cardiac mortality as their primary endpoint and spontaneous myocardial infarction (MI) as a secondary endpoint. Note that the included trials mostly measured all-cause mortality and (all types) of MI as a primary endpoint.

Compared with medical therapy alone, revascularization yielded a 21% reduction in cardiac death [RR 0.79 (0.67–0.93), P < 0.01] as well as spontaneous MI [RR 0.74 (0.64–0.86), P < 0.01]. They also did meta-regression, a technique that assesses the effect of a variable on effect size. In this case, they found that the cardiac death risk was lower in studies with longer duration. The authors did sensitivity analyses in which they repeated the analysis excluding groups of trials, say trials with patients following acute coronary syndrome (ACS) or those that included patients with chronic total occlusions, or those with CABG surgery. In each, the cardiac death signal remained significant.

Their conclusion was, “In stable CAD patients, randomisation to elective coronary revascularisation plus medical therapy led to reduced cardiac mortality compared with medical therapy alone.”

I think we could put this in a textbook of problematic meta-analyses. Here I will lean heavily on the brilliant editorial from Brown and Boden.

  • When the authors meta-analyzed the two primary endpoints that were most often measured in the RCTs—death and MI—they found no benefit from revascularization.

  • By definition, a meta-analysis is an analysis of data that is already known; that is, you know the results of the RCTs. It is therefore deeply problematic to simply pick and choose an endpoint. Here the authors chose cardiac (CV) death and spontaneous MI.

    • While choosing cardiac death may seem reasonable, the problems are that one-third of the included RCTs did not report CV death, and in unblinded trials, CV death is going to be more prone to bias than is overall death.

    • And the problem meta-analyzing just on spontaneous MI is that there is no agreement on reporting of MIs. Many people feel we should report all MIs, including those after procedures.

  • Perhaps the largest issue is that medical therapy has changed since the 1980s. Brown and Bowden include a Forest plot of studies from COURAGE (2007) to the present—a time when medical therapy included disease modifying drugs, such as statins. When you meta-analyze contemporary studies, there was no advantage to revascularization.

Of course, this should be of no surprise because, as Brown and Boden cite in their editorial, numerous meta-analyses have already been published that find no advantage of revascularization over optimal medical therapy. Brown and Boden conclude that the choices made in this meta-analysis are akin to putting a finger on the scale of scientific objectivity. They compare them to the statistical tricks that Thomson and Pocock warned against in the seminal Lancet paper. I agree; and this sort of stuff saddens me.

That this was presented as a late-breaker at a major scientific meeting, that it was published in a major journal, got by peer-reviewers who then allowed the authors to refute numerous properly done RCTs, tempts me to be cynical. Atherosclerosis is a systemic disease; skillfully done revascularization of focal disease has a clear role in helping selected people who fail medical therapy; why isn’t this enough?


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.