COMMENTARY

Jul 17, 2020 This Week in Cardiology Podcast

John M. Mandrola, MD

Disclosures

July 17, 2020

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 July 17, 2020, John Mandrola, MD comments on the following news and features stories.

COVID Update: Surge Continues; Vaccine News

The short story: America is in a hot mess. Multiple cities and states are seeing increases (70k new cases in one day this week—a record). As expected, hospitalizations are increasing and early data looks like deaths are tracking up.

The United States now has 3.6 million cases of COVID. I saw that in Arizona, one in 4 people tested are positive. US deaths 2 weeks ago were 135,000 and this week they are 141,000; the rate of rise is 1.04x, which is steady, but this flat line belies the focal nature of the pandemic. Florida, Arizona, California, and Texas all show clear signals of increasing rates of deaths. Another problem we have here is inconsistent reporting. We don’t seem to have a central repository of data.

The big question for the United States now is schools. Should they be opened or stay virtual? The school issue parallels medical decision making surrounding COVID. Namely, the risk tradeoffs: the risk of the virus vs the risks of avoiding the virus.

Just as it was with the shutdown, viral avoidance measures harm disadvantaged lower-income people more than the rich. White collar workers are much more able to keep their children up to speed in a virtual classroom.

The school question forces people to think about the concept of how safe does safe have to be. Coronavirus is dastardly: it spreads like crazy. The vast majority of people do fine with the infection, but enough people get super sick or die from it, that it induces terrible fear. Then, fear can cause people to make dubious decisions. Like ignoring chest pain or stroke symptoms, or not coming in for a pacemaker check, or delaying treatment of dying patients on the off-chance that he or she may have COVID. Off-target morbidity and mortality from this virus are beginning to worry me as much as COVID.

Last week I mentioned the news of vaccine trials. This week, The New England Journal of Medicine (NEJM) published a phase 1 study of mRNA vaccine in 45 healthy adults. It induced anti–SARS-CoV-2 immune responses in all participants, and no trial-limiting safety concerns were identified. More encouraging however is this lead sentence from a NYT column this week: “Of the 150-plus-coronavirus vaccines in development....” My friends, I am so happy there are 150 vaccines in development.

EXCEL

This week, the NEJM published multiple letters to the editor on the 5-year report of the EXCEL trial, and the authors replied, including a table of data. This table is significant. Recall that the EXCEL trial was the largest RCT comparing drug-eluting stents vs coronary artery bypass graft (CABG) in patients with left main disease. The authors published the five-year results in November 2019. They reported three-year results in 2016. About 950 patients were in each group. The composite outcome was death, stroke, or MI. Sounds simple, right?

Death is easy to count. So is stroke. The problem came with MI. In EXCEL, the authors chose to use the Society for Cardiovascular Angiography & Interventions (SCAI) definition of MI, which uses biomarkers and is procedurally agnostic. That is a problem because heart surgery will cause a greater release of cardiac enzymes than a percutaneous coronary intervention (PCI). Thus, CABG had higher rates of MI early and then fewer later. In the end, the no-difference in MI rates drove noninferiority of the composite outcome in the three-year results. Scrutiny of the actual data complicates this simplistic take on EXCEL.

  1. In the five-year results, the easy to count outcome of death favored CABG. The authors say the higher death rate could be due to chance; when they looked at causes of death, death from cardiovascular causes was not different. Of course, the deciding of causes of death is tricky without autopsy.

  2. The choice to use a biomarker-heavy definition of MI biases in favor of PCI.

  3. In the five-year report the authors conclude that the rate of composite outcomes—22% PCI vs 19.2% in CABG arm—were statistically not different. But EXCEL was designed as a noninferiority trial. The noninferiority margin set out at 3 years was 4.2%, meaning PCI could be 4.2% worse and still be called noninferior. Thus, had the five-year data been analyzed as the three-year data were, the upper bound of 6.5% would have exceeded noninferiority and PCI would not have met noninferiority vs CABG.

  4. After the 5-year data was published, the BBC acquired raw data from the trial. This included the rate of MI as defined by the Universal Definition of Myocardial Infarction (UDMI). This showed that at three years the rate of MI was 80+% higher in the PCI arm. Worse, this particular outcome was a secondary endpoint set out in the trial protocol, but there was no mention of this key secondary outcome when the three and five year results were published.

  5. These problems led to multiple surgical societies pulling their support of guidelines and a call for an independent review of EXCEL.

The letter-writers to the NEJM addressed these issues. The newsworthy issue came when the authors’ reply included a table with the results of MI as defined by UDMI. Again, this was a prespecified secondary outcome. Here there were 89 MI events (9.6%) in the PCI arm vs 43 MIs (4.7%) in the CABG arm. The authors give us the risk difference of 5% but the relative risk was about 2. Meaning, if you had PCI, you were twofold more likely to have an MI as defined by the UDMI.

The data were there. The secondary endpoint of MI by UDMI was specified in the trial protocol. Why wasn’t this data published in the three- or five-year manuscript? And, where were the editors and peer-reviewers of NEJM? Should it not be normal to look at what the scientists said they were going to report and see if they reported them?

Atrial Fibrillation

Last week, Jason Andrade and Canadian colleagues published a subanalysis of the CIRCA-DOSE trial that gets us thinking about the definitions of AF and also the huge knowledge gaps we have in AF ablation. CIRCA-DOSE, published in Circulation in November 2019, was a three-armed AF ablation RCT done in multiple Canadian centers: contact force–guided radiofrequency ablation vs 4 minute cryoballoon ablation (CB) vs 2 minute CB. About 350 patients. The primary endpoint was time to first AF or asymptomatic AF. Importantly, all pts got an implantable loop recorder (ILR).

The main finding was that there were no differences in 1-year efficacy in freedom from AF, which was only 53%. I know that sounds terrible; but the authors also showed that all three groups had a 90+% reductions in AF burden.

The present substudy, published in JAMA Open, was to evaluate the association of baseline AF episode duration with post-ablation outcomes. About three-fourths of the patients had AF lasting less than 24 hours, 7% had AF lasting 1-2 days, 12% had AF lasting 2-7 days, and 5% had AF more than 7 days.

The authors used the less than 24 hour group as the comparator and found that

documented recurrence of any atrial tachyarrhythmia following ablation was significantly lower in patients with a baseline AF episode duration of less than 24 continuous hours compared with those with longer AF episodes.

These findings add to the accumulating knowledge that patients with AF longer than 24 hours may be a different entity. Recall that cardioversion is more successful when AF is less than 24 hours; and stroke risk seems to increase substantially with more than 24 hours of AF. The authors propose that any AF lasting longer than 24 hours be considered persistent AF.

The two humbling caveats I would add to this data are that it’s possible that the reason patients with short-duration AF did better post ablation is natural history variation. In 2018, Steinberg and colleagues studied patients with AF detected by pacemaker and found that:

  • “Of patients with a first episode of 30 seconds to 2 minutes, 35.8% were free from subsequent episode >2 minutes at 180 days.”

  • Median AF burden was significantly less for patients with first episodes between 30 seconds and 3.8 hours versus >3.8 hours.

And a 2019 Danish study of the natural history of AF detected by ILR, the authors observed that AF episodes were present 2.7% (interquartile range: 1.0% to 15.7%) of monitoring days after debut. Progression to 24-hour episodes was seen in only 16%, whereas 22% had no AF episodes in the last 6 months of monitoring or longer.

My 4 take-home messages from this data are:

  1. We ought to have sham-controlled AF ablation trials.

  2. We ought to be mindful of the natural history of AF especially short-duration AF.

  3. We must keep reminding ourselves that the freedom from AF results in 2020 (with all our high-tech gear) is not much different that what Hassaguire reported in 1998.

  4. The inability to eliminate AF with a catheter is a reminder that in many cases the AF is the symptom not the disease. In many case, the real target is upstream from the atria.

Device Implantation Techniques

A quick note on a nice trial presented at the virtual European Heart Rhythm Association meeting regarding pacemaker implant techniques. This was a nifty RCT comparing self-taught ultrasound-guided axillary access for pacemakers vs cephalic vein dissection.

Brazilian authors from Porto Alegre studied the two techniques. About 44 pts in each arm. The primary endpoint was success; secondary endpoints were access time and procedure time and complications.

The axillary approach won: 98% success for the axillary approach vs 55% for the cephalic. The axillary approach was faster and associated with shorter procedure time. It’s a small trial. We ought to have a bigger one. But I like the axillary approach. It’s fast and elegant. That said, it’s good to know how to do both techniques.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
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.
Post as:

processing....