Jan 15, 2021 This Week in Cardiology Podcast

John M. Mandrola, MD


January 15, 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 January 15, 2021, John Mandrola, MD comments on the following news and features stories.


The big news this week seems to be the emergence of a more infectious COVID variant. F. Perry Wilson has a nice explainer on why a more infectious virus is so bad, even worse than an emergence of more deadly strain of virus. It gets down to math and exponential growth. Even though the chance of dying from the virus is low, say 0.5%, a more infectious strain exponentially raises the number of people infected and thus, you are multiplying 0.5% by an exponentially larger number.

This raises the stakes for improving the logistics of vaccination. The stories coming out of how well Israel has rolled out the vaccine are inspiring.

Left Atrial Appendage Closure

The use of left atrial appendage closure (LAAC) is supposed to be a stroke prevention strategy that avoids the need for long term anticoagulation. A Danish study compared percutaneous LAAC using Abbott’s Amulet device with direct acting oral anticoagulation (DOAC).

This would be quite an easy thing to study if there was a will to know. All you need to do is randomize patients to either arm and follow them for stroke or systemic embolism. That is not what the authors did. Instead, they took ≈ 1000 patients who had “successful implant” and compared them with 1000 patients from a Danish registry who received DOAC. JACC Interventions published this paper. The primary composite outcome of ischemic stroke, major bleeding, or all-cause mortality strongly favored LAAC; 14% of patients in the LAAC arm had an event vs 26% in the DOAC arm.

Pause on that for a second: these authors and this journal want us to believe that LAAC resulted in a 12% absolute risk reduction in hard events. A review of the components of the primary endpoint exposes the selection bias. There were no differences in ischemic strokes; a 42% reduction in bleeding, and a whopping 47% reduction in all-cause death.

The authors concluded that compared with DOACs, LAA occlusion (LAAO) may have similar stroke prevention efficacy, but lower risk of major bleeding and mortality. The lead investigator told a journalist from MDEdge: “The results suggest LAAO to be superior to DOACs in AF patients who have a predicted high risk of stroke and bleeding and adds to the evidence that LAAO is a promising stroke prevention strategy in selected AF patients."

What is wrong with this study? Not only is it nonrandomized but the LAAC data comes from an industry-sponsored post-market surveillance study. Unlike, say a Society of Thoracic Surgeons valve registry, which is mandated by law, this is likely a voluntary registry. I found this in the protocol from “The principal investigator is responsible for screening potential subjects to determine subject eligibility for the study.” My friends, these industry sponsored registries that enroll patients at the discretion of the enthusiasts doing the procedure cannot be used for comparisons. Period.

Second problem: propensity-matching, especially across countries, cannot be used to replace randomization. How can I tell there is obvious selection bias? The mortality curves diverge significantly after about 3 months. Not even the most enthusiastic enthusiast for this procedure will claim that you can see an all-cause mortality advantage in 3 to 6 months. What’s more, without a reduction of stroke events, the only way this procedure reduces death is by reducing bleeding; the bleeding curves don’t separate for more than 12 months and reduction in death was greater than the reduction in bleeding. Moreover, the most serious and immediately deadly kind of bleeding, that due to intracranial hemorrhage, was higher in the LAAC arm, 10 vs 8 respectively. Patients in the LAAC arm were healthier patients and that is why they did better.

There is one positive about this publication: if you are a learner, you can save it and put in a file as an example of how nonrandomized comparisons can be fatally flawed. My concern is that this paper will sit in the literature, be cited by proponents, maybe even incorporated in systematic reviews and those who take only a cursory look at the manuscript will be misled. Some would say, we don’t have a lot a data on the matter of LAAC vs DOAC, so this is something at least. I would answer that by saying—in the strongest terms possible—if this is the kind of data you cite, then no data is better.

AF Suppression

JACC published a super-interesting RCT looking at post-cardiac surgery atrial fibrillation (AF). The thing that is intriguing here is a potential clue about AF mechanism. Remember, literally no one understands the underlying pathophysiology of AF.

The investigators randomly assigned 200 patients who were having coronary artery bypass grafting (CABG) to receive an injection of calcium chloride (CaCl2) to each of the 4 ganglionated plexi (GPs) adjacent to the pulmonary vein ostia in the left atrium, or sham injection. CaCl2 basically shuts down autonomic activity. The primary endpoint was postoperative AF (POAF) and POAF incidence was reduced from 36% to 15%.

Heart rate variability data showed a reduction in both high-frequency and low-frequency power in the CaCl2 group with a preserved low-frequency/high-frequency ratio, suggesting that the sympathetic/parasympathetic balance was not perturbed by CaCl2 injection. In other words, the CaCl2 injection created neural modulation of the heart. This makes you wonder about the role of neural pathophysiology as an underlying mechanism and possible therapeutic target. Recall that both sympathetic and parasympathetic neural trafficking in the atrial autonomic neural network is processed and integrated in the GPs.

Previous data is mixed and confusing. Cardioneural ablation is undergoing a resurgence of interest. If the result of this study is a real finding, it should be easy to replicate in multicenter studies. And if it reduces POAF by that much it will be a big gain in societal costs.

SGLT2 Inhibitors

As a class, SGLT2 inhibitors have pretty amazing cardiovascular and renal outcomes data. But they cost a lot. High cost raises the question of patient selection. A group of authors reported on the application of a biomarker-based risk score to identify individuals with dysglycemia who are at high risk for incident heart failure (HF). Their goal was to gather information on better allocation of the drug.

They used three longitudinal cohorts:  the Atherosclerosis Risk in Communities (ARIC), Dallas Heart Study (DHS), and Multi-Ethnic Study of Atherosclerosis (MESA) epidemiologic studies. The risk score was integer-based, like the CHADSVASC, and included high-sensitivity-assay cardiac troponin-T (hs-cTnT) and C-reactive protein (hs-CRP); N-terminal proB-type natriuretic peptide (NT-proBNP); and left ventricular hypertrophy by ECG. There was 1 point per category.

The final cohort comprised about 7000 patients, one-third with diabetes and two-thirds with pre-diabetes. The 5-year risk of HF among subjects with a biomarker score of ≤1 was low and comparable to participants with euglycemia (0.78%). HF increased in a graded fashion with an increasing biomarker score. The same pattern was seen in patients with pre-diabetes.

If you use the drugs in all patients, you prevent HF events–roughly 11 out of 1000 over 5 years. If you use the drug in patients with higher risk, you prevent even more events—roughly 44 out of 1000. An important distinction in this data is that none of these patients had any heart disease.

Obviously, this is a statistically complicated study using point measures of biomarkers that could easily change over time. What’s more, you’d want to validate this approach in trials. But the fact remains that SGLT2 inhibitor uptake in patients with diabetes is low. probably due to costs.

In an ideal world, we would have no need for risk scores of beneficial drugs. But the fact is that in cost constrained systems, risk stratification remains important.

ACE/ARBs in Chronic Kidney Disease

One thing I see all too often is the reflex to stop angiotensin receptor blockers (ARB) or angiotensin-converting enzyme (ACE) inhibitors in patients with chronic kidney disease (CKD). The creatine goes up a bit and boom, that’s it for the beneficial class of drugs. Guidelines recommend making this decision on a case-by-case basis.

Swedish, Dutch, and Canadian authors studied the association of outcomes in patients who did or did not stop RAS inhibition in patients with advanced CKD. They used the Swedish Renal Registry of patients over 10 years. These were patients who developed CKD, with an estimated glomerular filtration rate (eGFR) less than 30 mL/min/1.73m2 while on renin-angiotensin system (RAS) blockade. They had about 10,000 patients in this category; 15% stopped RAS.

Compared with continuing RAS inhibition, stopping therapy was associated with a 14% higher 5-year risk of death and 12% higher risk of major adverse cardiovascular events (MACE), but an 8% lower risk of kidney replacement. The problem of sorting through this nonrandomized comparison is confounders, of course. Were the patients who stopped ACE or ARB sicker? The clue is that mortality reduction is larger than reduction in MACE.

I won’t make causal inferences. I will just say that the routine stopping of RAS inhibition in CKD is unwise; close follow-up is needed. There is an ongoing RCT called the STOP-ACE trial. That’s great, because randomization is the only way to answer this important question.


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: