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 November 20, 2020, John Mandrola, MD comments on the following news and features stories.
COVID and COVID Vaccines
The COVID situation in the United States is bad. Even our hospital, which, for the entire pandemic, has stayed steady at about 5% of beds being for COVID admissions, is now seeing a near tripling of cases.
I see from the graphs that much of Europe is now coming down from the peak, and I suspect we will too soon. People self-adjust, because, generally, people are good. On the vaccine front, this week brought more good news in the form of the Moderna vaccine, which was also announced with a press release. Moderna vaccine results were similar to the Pfizer vaccine results—95 infections in a trial of about 30,000; 90 of the 95 were in the placebo group. There were no significant adverse events and the vaccine does not have to be stored at super-cold temps.
Also this week, Pfizer announced it had hit its safety milestone and will be applying for an emergency use authorization (EUA) soon. The 90+% efficacy was upheld.
Still more trials of more vaccines are to come. This is good news. But one question that will come up is how hard will it be to keep doing trials of vaccines when there is an EUA? Many voices online have rightly voiced concerns over safety. Perhaps we should wait for more data? As a medical conservative, I disagree. There are times in medicine when waiting for better evidence is wise. But there are also times when waiting for perfect information is foolish. If a patient is crashing, you have to assess and decide, often on imperfect information. The pandemic is akin to a patient crashing; the sensible thing is to move forward, vaccinate the most vulnerable, but crucially, we ought to randomize as we go so we can continue to gain information on safety and efficacy.
What do I mean? Ok, consider that, initially, there will be limited vaccine supply and some people will want it and others may not. The economist Alex Taborrak had a great idea.
There are many ways to conduct clinical trials while releasing a vaccine—indeed, we can make the clinical trials better by randomizing a phased release. Suppose we decide health care and transit workers should be vaccinated first. No problem–offer the workers the vaccine, put the social security numbers of those who want the vaccine into a hat like draft numbers, vaccine a randomly chosen sub-sample, monitor everyone.
Virtual AHA Meeting
I saw three big stories from the virtual meeting.
In 2015, the Lancet published a massive meta-analysis of trials of statin drugs. At that time, 174,000 people had been enrolled in statin trials. The drugs have been found to show a 20%-25% relative risk reduction in future cardiac events. But the drug can’t help unless a person swallows it. Many people stop taking the drugs. As with any drug, statins can cause rare and severe side effects, like frank myositis or liver damage. We are not talking about these super rare events. I am talking about myalgia, brain fog, fatigue, etc. The common ones.
In the blinded clinical trials that measured these side effects, there is no difference between statins and placebo for the mild side effects. But in the observational studies, in which patients know they are on statins, the reports of side effects are much higher with statins.
There are two ways to explain this discrepancy: one is that the trials have run-in periods and people who complain about side effects don’t get randomized. The other way is the nocebo effect—which is when negative expectations produces a negative outcome. The nocebo effect is the opposite of the placebo effect. A good example is if I prescribe a drug and tell you it might cause difficulty getting an erection, erectile dysfunction becomes more likely to occur.
Researchers at Imperial College London set out to study the possibility of nocebo effects in a super novel way. They took 60 patients who had previously stopped statins because of side effects and gave them 12 blank pill bottles, one for each month, which they were to take in a random, predefined order. Four bottles contained a daily statin (atorvastatin), four contained a daily placebo, and four were empty. Patients and researchers were blinded. Patients recorded their sensations every day. The primary endpoint was the nocebo proportion.
Think of a bar graph with three bars. Symptoms are on the y-axis. The months when there are no pills patient symptom scores are low, but not zero. The months they take placebo, there are more symptoms. And the months they take statins, if there is a true side effect from the biochemical effect of the drug, the symptom score bar should be higher than the placebo score. I have a picture in my column.
Here is what the researchers found: The mean symptom score for all patients was 8.0 during no-pill months, 15.4 during placebo months, and 16.3 during the statin months. The difference in symptom scores between the placebo and statin months vs the no-pill months was highly significant (P < .0001). But the difference in symptom scores between the placebo and statin months was not significant (P = .39). The nocebo proportion was 0.9 – with wide error bars. This means that, yes, patients who take statins experience significant symptoms, but it is not from the statin, it is from the act of taking the statin tablet.
At the end of the trial, the research team showed each patient their results and 30 of the 60 resumed taking the statin. The specific message is that if we teach patients about the nocebo effect, many would or could resume a beneficial drug.
Fish Oil Controversy
Right before the SAMSON trial presentation, the investigators of STRENGTH presented their trial’s results. STRENGTH was an RCT comparing an omega-3 carboxylic acid containing EPA and DHA, 4g daily vs a corn-oil placebo for the reduction of major adverse cardiovascular events (MACE). Results were published online in JAMA.
Corn oil was selected as the placebo intentionally because olive oil may be a beneficial placebo and mineral oil could be a harmful placebo. The trial enrolled statin-treated patients at high-risk for cardiovascular disease with average triglycerides at baseline of 239 mg/dL. The trial was stopped for “futility” earlier this year. There were absolutely no differences. Kaplan-Meier curves totally overlapped.
Key in this trial was that EPA levels were measured and found to be nearly tripled. They also looked at patients with the highest EPA levels, and still no signal of benefit. They did find, as did the super-positive REDUCE-IT trial, that fish pills increased the risk of atrial fibrillation (AF). This trend is under-appreciated, I believe.
STRENGTH authors speculated on why STRENGTH was null and REDUCE-IT so positive. Recall that REDUCE-IT studied pure EPA in the form of icosapent ethyl in a somewhat similar patient population. REDUCE-IT found a massive 25% relative risk reduction in events. This caught people by surprise because all previous fish oil trials were null. The potential problem with REDUCE-IT was that the mineral oil placebo caused significant increases in low density lipoprotein, C-reactive protein, and Apo-B levels.
The STRENGTH investigators offered possible reasons for the different outcomes:
DHA is toxic.
REDUCE-IT enrolled sicker patients (established coronary artery disease in 71% vs 56% in STRENGTH).
Unfavorable effects of mineral oil. In REDUCE-IT; mineral oil increased LDL by 10%, Apo-B by 8%, and CRP by 32%.
In the Q and A the principle investigator (PI) of STRENGTH, A. Michael Lincoff from the Cleveland Clinic, questioned the results of REDUCE-IT. Dheepak Bhatt, the PI of REDUCE-IT countered, calling these comments absurd. Bhatt cited the fact that JELIS, another RCT of pure EPA found benefit, as well as concerns that DHA could be toxic.
My take is that the mineral oil placebo in REDUCE-IT clearly was associated with significant increases in atherogenic compounds. I concede that the pure EPA used in REDUCE-IT is a different product than EPA/DHA, and so it is possible that EPA alone is beneficial, but the mineral oil effects create lots of uncertainty.
AHA featured two trials of ablation of AF vs antiarrhythmic drugs (AAD) in patients who had yet to try rhythm control. Both the US STOP AF trial and Canadian EARLY AF trial were simultaneously published in the NEJM. Medtronic, the makers of the cryoballoon (CB) ablation system used in the trials, was involved in funding for both trials, though its role was much larger in the US STOP AF trial.
Current practice holds that AAD should be used in most patients first, and ablation performed if AF persists or drugs are intolerable. STOP AF and EARLY AF tested CB pulmonary vein isolation as an initial strategy. STOP AF chose as an endpoint a composite of treatment success—freedom from AF, freedom from crossover to ablation, cardioversion etc. EARLY AF simply used recurrence of any atrial tachycardia. Both trials enrolled relatively young patients (about 60) who had normal hearts, small left atria, and low-burden AF.
The results were similar to previous trials comparing ablation to drugs: both STOP AF and EARLY AF found that ablation reduced the endpoint of AF recurrences. However, there were many caveats: In STOP AF, most treatment failures in the drug arm occurred while patients were taking ridiculously low doses of drugs and in 13% of treatment failures the patients were taking no AAD.
What’s more, in STOP AF, crossover from drugs to ablation was part of their primary endpoint (PEP). In total, there were 51 PEP in the AAD arm. There were 99 patients randomized to the AAD arm, but 15 patients or nearly 40% of all PEP events met the endpoint not by having AF but by crossing-over into the ablation arm. Most of these crossovers were for drug side effects. Only two of these 15 had documented arrythmia recurrences.
Remember that these trials are done at ablation centers; the trialists are ablation proponents, and no one is blinded, so there could be a hair trigger to cross-over. Hence, a serious bias.
The Canadian-led EARLY AF was much more rigorous. They had a protocol to escalate drug doses so as to give them a fair chance; they used implantable loop recorders to capture all AF episodes; and they measured quality of life pre- and post, with multiple questionnaires. EARLY AF found that quality of life improved a lot in both groups but significantly more so in the ablation arm.
As for safety, STOP AF obfuscated the results by listing many mundane events such as heart rate increases, fluid overload, influenza, rotator cuff syndrome etc. This had a dilution effect. You had to go to the supplement to learn that 6% of patients in the ablation arm had procedural complications. EARLY AF was clearer about their complications, about 3% in the ablation arm and 4% with drugs.
These trials will now be cited by proponents as saying early ablation should be considered to arrest the progression of AF and that AF ablation is safe and superior to medicines. This scares me. The first thing to say about translating this evidence to the bedside is this: In many US locales, ablation of AF is already way overused. AF ablation is well compensated. The procedure makes lots of money for industry, for the hospital and for us doctors. Doctors who make money for the hospital earn “status” as high-producers. If you are lousy at AF ablation and your patients require redo procedures, you actually earn more money.
Another issue was the lack of proper control arms, not just in these trials, but in all ablation trials. Dr. Rod Gimbel from Indiana made the point that patients in these two trials were young and healthy with early AF and low-burden AF. He cited two studies of such patients that show extremely low rates of progression over time, and some patients even regress. Gimbel’s point is that we ought to have a conservative non-rhythm control arm. Imagine an arm with a reassuring AF-team approach where a patient learns not to fear AF, and they learn how to reduce risk factors.
I would add to Gimbel’s point and say we also ought to have a placebo-procedure arm or sham arm. Why? Consider the median AF burden in the EARLY trial. Median AF burden was 0% in ablation arm and 0.13% in the drug arm. Why then, do patients in the ablation arm report feeling so much better? Both groups are at least 98% free of AF. There has to be a huge placebo effect from the procedure.
Finally, these are one-year trials, we also know that over time, AF tends to recur.
© 2020 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: Nov 20, 2020 This Week in Cardiology Podcast - Medscape - Nov 20, 2020.