Sep 3, 2021 This Week in Cardiology Podcast

John M. Mandrola, MD


September 03, 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 September 3, 2021, John Mandrola, MD comments on the following news and features stories.

I am not saying a word about COVID today. Let’s move to Cardiology. It will take 2, maybe 3 weeks to get through the big studies from the European Society of Cardiology (ESC) meeting.

EMPEROR-Preserved: Empa vs Placebo in HFpEF

SGLT2 inhibitors have proven beneficial for the reduction of cardiovascular (CV) events in patients with type 2 diabetes (T2D), the reduction in renal outcomes in chronic kidney disease (CKD), and reduction of hospitalizations for heart failure (HHF) and CV death in heart failure with reduced ejection fraction (HFrEF). The EMPEROR-Preserved trial was the first RCT testing an SGLT2 inhibitor (empagliflozin) in patients with HF with preserved EF (HFpEF).

The primary endpoint was the time to first hospitalization for HF or CV death. Recall that PARAGON-HF, the HFpEF trial of sacubitril/valsartan vs valsartan, studied total HHF or CV death.

EMPEROR-Preserved investigators screened more than 11,000 patients and randomized 6000 of them. That’s important because all trials carefully select patients, but EMPEROR-Preserved excluded almost as many as they randomized. Most were excluded because they did not meet the threshold for brain natriuretic peptide (BNP). The average age was 72 years, about half were women, and one-third had slightly decreased left ventricular EF of 40% to 50%. So it wasn’t all totally normal EFs.

  • At slightly more than 2 years of follow-up: 13.8% of patients in the empagliflozin group experienced CV death or an HF hospitalization compared with 17.1% of the placebo group.

  • The 3.3 percentage point absolute risk reduction translated to a 21% relative risk reduction. The result was statistically robust with a P value of .003.

  • HHF was reduced by 29% but CV death was only reduced by 9%. The HHF hospitalization rate was statistically significant but the CV death reduction was not.

Before I comment, let me tell you about a second presentation on empagliflozin. Dr. Milton Packer presented an “Emperor-pooled” analysis of renal outcomes. The pooled part is the combination of EMPEROR-Reduced, which was empagliflozin vs placebo in patients with HF, and EMPEROR-Preserved. Packer points out that the two trials were sort of sister trials, with similar protocols, sites, and infrastructure; the difference was simply and EF of 40%. Above it and you got into EMPEROR-Preserved, below 40%, and you got into EMPEROR-Reduced.

The pooled analysis was interesting. In EMPEROR-Reduced, the drug reduced a composite renal outcomes by 50% whereas in EMPEROR-Preserved, there was no reduction in renal outcomes.

Summary: EMPEROR-Preserved gets to be the first truly “positive” trial in HFpEF. And since many patients with HFpEF also have CKD and T2D, there are good reasons to use an SGLT2 inhibitor—specifically empagliflozin. (A placebo-controlled trial of dapagliflozin vs placebo in HFpEF is coming soon.) But I think the buzz over this drug is overdone. Here is why:

  • The only signal of benefit in this trial was in HHF, not CV death or all-cause death or renal outcomes. Preventing admissions for HF is important, but only if HHF represents a big chunk of all-cause admissions.

  • In this group of patients with tons of other medical problems, HHF represents less than 20% of all hospitalizations. The number of HHF in the empagliflozin arm is 407; the total admissions is 2566.

  • Pause there and think about this—we are about to accept an expensive drug based on an endpoint that represents a very small fraction of the reasons this group gets admitted to the hospital. Later in the podcast I will discuss another trial of HF in which HHF represents nearly 90% of all-cause HF. When that is the case, HHF is a good endpoint. But not in a condition like HFpEF.

  • The lack of improvement in renal outcomes, makes me think that empagliflozin acts simply as an expensive diuretic, not an important disease-modifying drug.

  • Finally, the mean difference in the Kansas City Cardiomyopathy Questionnaire (KCCQ), which goes from 0-100 and measures quality of life (QOL) was less than 2. Most people feel a meaningful change in KCCQ is 5. So empagliflozin underperforms on QOL measures as well.

My friends, I am big fan of SGLT2 inhibitor drugs, but EMPEROR-Preserved fails to convince me that we should accept such an expensive drug in the absence of reductions in CV death, all-cause death, renal outcomes, or QOL.

Yes, there will be patients with HFpEF who also have CKD, or T2D, or with an EF hovering around 40%, whom I might consider giving an SGLT2 inhibitor, but a blanket recommendation to treat all HFpEF patients with empagliflozin would be unwise and wasteful of healthcare dollars.

The LOOP Study

To date, nearly all atrial fibrillation (AF) screening studies have found that if you screen older patients with risk factors for stroke (with any technique), you can show more AF episodes and more anticoagulant (AC) prescriptions in the screened group. But finding more AF and initiating more AC does is not the endpoint that matters. The endpoint that matters. The core reason we use AC in AF is to reduce stroke without increasing major bleeding. Indeed, one of the reasons I was drawn to the LOOP AF screening study was that it measured stroke and systemic embolism as its outcome.

The LOOP study enrolled about 6000 patients who were about 74 years old, with a mean CHADSVAS score of 4; 1500 got an implantable loop recorder (ILR), 4500 got standard care via primary care. Any AF lasting 6 minutes or more in either group prompted a recommendation for AC.

  • AF was detected in 32% of the ILR group vs 12% of the control group. That’s more than 3 times different.

  • This led to nearly 3 times more use of AC in the ILR group.

Follow-up was 5 years.

  • Stroke was reduced by 20%; 4.5% of the ILR group had a stroke or systemic embolism vs 5.6% in the control arm.

  • The hazard ratio (HR) was 0.80 and the 95% confidence interval (CI) ranged from 0.61, a 39% reduction in stroke, to 1.05, a 5% increased risk of stroke. The P-value was 0.11. Thus, the reduction in stroke did not reach our threshold of significance of.05.

  • If there was truly no difference in the two groups, aka the null hypothesis, there is an about a one in 10 chance of observing this data or something more extreme.

So, it seems like a wash.

An implantable cardiac monitor (ICM) is the ultimate AF monitor. Unlike an Apple Watch or spot checks, an ICM is always on. In LOOP, the authors were careful to adjudicate each episode, so it’s likely the false positive rate was also quite low. Because stroke reduction was nonsignificant and bleeding rates were increased, this study clearly closes the chapter on using super-expensive ILR monitors to screen for AF.

But there’s some nuance here that is worth mentioning because the issue of how much AF warrants AC is like the # 1 question in all of cardiology.

The authors used 6 minutes of AF to trigger AC. That’s probably too low. Remember, in the ASSERT observational study of cardiac device patients, 6 minutes of AF did associate with 2.5-fold higher risk of stroke, but nearly all that increased risk stemmed from longer-duration episodes of more than 24 hours.

One wonders whether a higher threshold for AF may have had a bigger effect on stroke reduction with less major bleeding. I say that because the patients in LOOP may have had too many vascular risk factors to benefit from AC. A super-important point, one that is often forgotten by AF doctors, especially those that favor percutaneous closure of the left atrial appendage, is that AF is not the most common cause of ischemic stroke.

It’s like the HHF vs total hospitalization argument I just made. Reducing AF-related stroke may not have a major impact on patients with oodles of vascular risk factors, because for these patients AF-related stroke represents a small proportion of the reason for strokes.

I am not saying I know the sweet spot of who should get oral AC, but recall that in the seminal trials of AC, only patients with clinical AF diagnosed on a 12-lead ECG--done for a reason—were randomized. The lack of the ability to screen for AF back in the old days may have enriched the trials then with patients most likely to benefit from AC.

The good news is that there are ongoing trials looking at AC for subclinical AF. We will learn more about that threshold. And as wearables expand, we will also learn how much AF occurs as the normal part of the aging process. For now, it will have to be judgement in the use of AC.

Finally, I think the LOOP study beautifully confirms how hard it is to show that preventive screening works in real life.

Ablate and Pace

The APAF-CRT trial looked at cardiac resynchronization therapy (CRT) pacing and atrioventricular (AV) node ablation vs rate control in patients with symptomatic permanent AF that was deemed unsuitable for primary AF ablation or that ablation had failed; narrow QRS (≤110 msec); and at least one hospitalization for heart failure. The trial delivered stunning results.

We all know these patients: older, often female, small stiff ventricles, difficult to control ventricular rate, tons of meds, soft blood pressure (BP), and severe dyspnea. AF ablation would be foolish because a) it wouldn’t work, and b) the procedural risk is too high. Whereas most patients with AF benefit from reassurance and treatment of risk factors, patients like those in APAF-CRT need action. In essence they are dying from AF. Emphasis on from.

APAF-CRT was a two-phase trial. In 2018, investigators reported morbidity outcomes: CRT/ablation was superior to medical therapy in reducing heart failure hospitalizations and quality of life. At ESC, they presented (and published in the European Heart Journal) the second phase, the hard outcomes.

  • APAF-CRT found a 74% reduction in death with CRT and atrioventricular node ablation.

  • After 29 months, seven of 63 patients in the CRT/ablation arm died vs 20 of 70 patients in the drug arm. Keep in mind that mortality is the most bias-free endpoint.

  • This is even more impressive given that 18 of 70 patients in the medical arm crossed over to receive CRT and node ablation but by intention to treat, they were still counted in the med arm.

  • LVEF did not modify the results; the benefits of ablate and pace were equally good in those with normal and abnormal EF.

I am not sure why this trial didn’t get huge attention: a 74% reduction in death is amazing. These patients are dying from AF. Many have small stiff ventricles, and they do terrible with rapid irregular rates. Mechanical rate control regularizes the rate, obviates the need for high doses of meds, and the CRT prevents the LV dysfunction that can accompany right ventricle-only pacing. While I don’t think we should be creating AV block and pacer dependence willy-nilly, this procedure offers patients huge relief, and now we can say a mortality benefit.

The next step is to study nodal ablation with biventricular pacing vs left bundle branch area pacing. The latter, if successful, can achieve CRT with one lead instead of 2.

SSaSS – Salt Substitute and Stroke Study

No matter where you fall on the salt debate, three things are clear: elevated salt intake and low potassium intake associate with high BP, salt reduction can reduce BP, and lower BP associates with lower risk of stroke.

Salt substitutes that replace part of 100% sodium content with potassium chloride content, work in two ways—they simultaneously reduce sodium intake and boost potassium intake. Studies have shown that this reduces BP.

The question SSaSS trialists asked is whether this translates to a reduction in hard outcomes, such as stroke and major adverse cardiac events (MACE). And, yes, it did. SSaSS was conducted in rural counties in China. This is important because in these areas, most of the sodium intake comes from added salt in cooking. There are no significant amounts of processed food—which is high in sodium. This fact allows for a study of salt substitute. It would be much harder to study in North America because here, the majority of salt intake comes from processed food, not cooking or seasoning.

Sufficient salt substitute was provided to cover all household cooking and food preservation requirements at an average of approximately 20 grams per person per day. The investigators did not randomize individuals, however, they randomized entire villages. In all, 300 villages were given the salt substitute (75% sodium/25% potassium) and 300 villages stayed with the normal routine.

They then chose 35 individuals per village, about 10,000 patients in each group. About three in four patients had a history of stroke and 88% were hypertensive at baseline. These were high risk patients. Follow-up was about 5 years.

  • Mean BP reduction was 3.3 mmHg

  • Urinary sodium excretion was lower and potassium excretion was higher in the active arm.

  • The rate of fatal or nonfatal stroke events was significantly lower in the salt-substitute group than in the regular-salt group (29.1 events vs. 33.6 events per 1000 person-years) (Table 1 and Figure 3)

  • There were statistically significant reductions in MACE (13%), and all-cause death (12%).

  • Safety was assessed by hyperkalemia; there were no sig differences.

I spoke with primary author Bruce Neal from the University of New South Wales and we will have a video coming on the | Medscape Cardiology. He assured me that despite this being in rural China, follow-up was complete in more than 99%. Interestingly, he pointed to the national healthcare system in China as one reason you could get such strong follow-up.

I also asked him about the mean BP reduction of just 3 mm Hg. How did this translate into big gains? He reminded me that this degree of BP reduction in high-risk patients can make a big difference. It wasn’t far off what we get with some drugs. I’ve thought about this trial more since our chat, and I guess it doesn’t really matter so much about the BP if the intervention works. Recall that BP isn’t a disease, it’s only a surrogate marker. While the absolute risk reduction in outcomes were modest, applied to populations, this trial is a big deal. The authors remind us that the prevented strokes, cardiac events, and deaths with this intervention could number in the hundreds of thousands.


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: