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 on Apple Podcasts, Spotify, or your preferred podcast provider. This podcast is intended for healthcare professionals only.
In This Week’s Podcast
For the week ending June 17, 2022, John Mandrola, MD comments on the following news and features stories.
Two recent papers addressed the issue of statin intolerance. One is a scientific statement paper on statin intolerance from the National Lipid Association and the other is a newly published randomized N-of-1 trial.
The first thing to say about statin intolerance is that the issue transcends the simple but important matter of whether a patient takes a statin. Yes, of course, we should care about whether a patient takes a medicine that reduces cardiac events. But the underlying themes here—causality, misattribution of causality, and nocebo effect—lie at the core of doctoring. I am especially drawn to the nocebo effect because words spoken in clinic or the hospital can harm and words can heal. Nocebo is the opposite of placebo effect.
A super-quick recap of the background on statin adverse events (AE).
Statins are the most studied medical intervention. Hundreds of thousands of patients have been randomized to statins or placebos. The drugs should not be thought of as cholesterol lowering drugs. Better is to think of them as heart attack and stroke preventers. There is a consistent 20% to 25% reduction of events on statins.
And in all these blinded trials—where patients do not know if they are on statin or placebo, adverse effects are no different. In fact, a meta-analysis found that more patients in the placebo arm stop the drugs for side effects.
But in observational trials, where patients know they are on statins, the rate of AE is quite high, and the proportion of patients who qualify for statins but do not take them is high. And up to half of all patients taking a statin stop the drug after 2 years.
Anyone who cares for patients knows that statin AE reports are super common.
There are three possible reasons for this disconnect. One is that the trials are wrong. Run-in periods can screen out intolerant patients, and we all know trials enroll ‘ideal’ motivated patients. Another reason is that patients misattribute causality of the muscle pain. They think it is due to the statin but it is really due to their diffuse arthritis. The third possible explanation of the disconnect among trials and observational studies of statin AE, the most intriguing reason, is the nocebo effect. That is, they expect the drug to cause them to feel unwell, and they do feel unwell.
Circulation Outcomes has published a pragmatic N-of-1 trial in which Oxford researchers studied the use of a behavioral intervention plus two types of N-of-1 trials for helping patients stay on statins.
First a word on N-of-1 trials. This is where the patient acts as his or her own control and active arm. Simply stated: 1 month on treatment, 1 month off treatment, and so on. We do this in regular practice a lot. We tell patients to try taking a med, then try stopping it and then going back on it. I do it a lot with antiarrhythmic drugs for benign arrhythmias. The problem of course is to do a proper scientific N-of-1 trial, you need a proper placebo tablet so as to ensure blinding, and you need a way to systematically record symptoms. And by the way, placebo tablets are very expensive and available only for science. Of course, #TWICPodcast has previously covered the incredibly smart SAMSON trial, led by James Howard at Imperial College London, in which statin intolerant patients took the real statin, a placebo, and no tablet in alternate months. Patients felt best during the no-tablet months but had essentially the same side effects on placebo as they did on statins. The conclusions were clear: the side effects were totally utterly real but were due to taking a tablet, not the statin chemical. It was nearly all nocebo effect.
But it’s hard to do such a trial as it required an expensive placebo, and an app to record symptoms. What the Oxford researchers, first author, Dr. Kate Tudor, reported in Circulation Outcomes was something simpler and more pragmatic. Their randomized controlled trial (RCT) had three arms and included 93 patients. Recall that in N-of-1 trials you don’t need as many patients because each patient is his or her own control.
Arm 1 had 36 patients who underwent behavioral counseling plus unblinded N-of-1 trials: Take the statin 1 month, don’t take it the next.
Arm 2 had 37 patients who underwent behavioral counseling plus blinded N-of-1 trials with a placebo. One month statin, one month placebo.
Arm 3 had usual care.
The behavioral counseling here may be quite important: the doctor elicited participants’ concerns about statins and sought to address these by positively endorsing the cardiovascular benefits of statins, explaining their mechanism of action to create trust, and the prevalence of AE in trials vs practice.
Who were the patients? Primary care physicians searched electronic health records for patients with prior statin intolerance or who had previously declined statins and in whom statins were indicated. Patients were contacted by letter.
The investigators found that 71% of the patients who embarked upon the unblinded arm completed the full 6 months, versus 82% of the blinded arm.
Similar proportions of patients restarted statins after the trial in the blinded (46%) and unblinded (44%) N-of-1 trials, but fewer in the usual care arm (20%).
The authors concluded that an unblinded N-of-1 trial approach with behavioral counseling can improve medication compliance when compared with usual care.
This is provocative data. It’s different from SAMSON in which the investigators showed that statin intolerance stems from nocebo effect. SAMSON also found that half the patients in their trial restarted statins. On the surface, this trial shows that perhaps you don’t need the more elaborate science of SAMSON—the blinding from the matched placebo, the month of no tablets, and the app to record daily symptoms. That would be great. Just tell patients to try off and on months. It doesn’t matter why they restart statins, ultimately the goal is to lower their risk with the drugs.
The problem here is generalizability or external validity. Always look at the trial flow diagram, usually Figure 1. Here they screened 700 to enroll 93, or about 1 in 8 patients. That is a highly selective sample of motivated patients. Then all these motivated patients got what sounds like an extremely supportive and time-consuming behavioral intervention. Then, in this scenario, unblinded N-of-1 trials seemed as good as placebo-controlled trials for the endpoint of starting statins.
I love this type of work, but I concur with the editorialists that there are limitations in saying this technique will work in most patients with statin intolerance. That said, I see three take-home messages:
Clearly the AE profile of statins are complex, and the nocebo effect is strong. In other words, our words and actions matter.
Many patients with statin intolerance can restart the statin.
N-of-1 trials have great promise—and ...imagine a world, as the editorialists do, where we had an atorvastatin tablet with 0 mg in which we could let patients do their own experiments.
I would love such a 0 mg beta-blocker tablet, because I strongly suspect that there is a big nocebo effect here as well.
I know this is awful jargon, but the concept is core to using evidence at the bedside. External validity speaks to the generalizability of a trial’s result to the patient who is front of you. The classic example is that of guideline directed medical therapy (GDMT) for heart failure (HF). GDMT stems from trials done in mostly male outpatients with HF, those patients who are well enough to walk into an HF clinic and participate in the demands of an RCT. The problem now is that profoundly silly quality measures urge clinicians to ignore external validity and begin these therapies in acutely ill hospitalized patients who often have multimorbidity. One of the main goals of proper evidence-based medicine is wisdom in applying trial results.
But here is the tension: from a scientific standpoint, trialists are often selective in their enrollment. This is fine — the goal is to define *a* population in which a therapy works. Having selected patients, selected protocols, proper controls, and robust follow-up works to make a trial cleaner. (we call this internal validity). We can then say for this population under these circumstances, this tablet or surgery reduces this endpoint.
That makes things tough on us, the consumers of medicine, because it’s hard to tell if the effect size seen in the confines of a trial apply to our older patient who has tennis ball cutouts on her walker and advanced chronic kidney disease (CKD). This is external validity. Pragmatic trials, such as the one I just discussed for statin intolerance, study interventions more as they occur in the real world. They are designed to maximize external validity.
This week the story of oral nirmatrelvir plus ritonavir for the acute treatment of COVID-19 beautifully illustrates the concept of external validity. I won’t say brand names but you know what I am talking about.
Cardiology and especially electrophysiology has become acutely aware of the drug because hardly a day goes by that I am not messaged about whether a cardiac patient can take it. This is b/c nirmatrelvir plus ritonavir [Note: the name ritonavir is mis-stated in the audio] is a CYP3A inhibitor and there are oodles of cardiac drugs that are metabolized via this pathway.
In April of this year the New England Journal of Medicine published the EPIC-HR trial, an RCT of more than 2000 patients who had pre-Omicron COVID-19 and risk factors for severe disease. As the authors of the paper wrote: “Of importance, this trial was restricted to unvaccinated persons.”
In these patients, in this setting, the drug crushed it.
The incidence of COVID-19-related hospitalization or death by day 28 was lower in the active arm than in the placebo group by 6.3 percentage points
This was an 89% relative risk reduction and a massively small P-value. Huge effect and statistically robust.
All 13 deaths occurred in the placebo group.
This trial was completed in late 2021.
But now, in late Spring of 2022, the question became who should get this drug. Would the drug have the same efficacy in vaccinated patients infected with a milder strain of virus?
This week, Pfizer announced via a press release that the EPIC-SR trial of nirmatrelvir plus ritonavir did not meet its primary endpoint of symptom alleviation. These were lower risk patients, including those who had been vaccinated. Here were the key results:
0.86% in active arm vs 1.7% placebo arm had hospitalization or death.
This 50% relative risk reduction did not meet statistical significance.
Given that the absolute risk reduction was less than 1 in 100, the effect size was likely not clinically significant or at least a heck of a lot less clinically significant than it was in the higher-risk trial.
This is the same drug with the same pharmacology but because the second trial enrolled patients who were lower risk the clinical effect dropped substantially. The absolute risk reduction (ARR) was > 6% in a higher risk group and now was less than 1%.
The analogy I see is to implanted cardioverter-defibrillators. An ICD converts ventricular fibrillation (VF). Low-risk patients can have VF, but we don’t implant them in people with low risk because the benefits (the ARR) would be so small that it is neither clinically significant nor cost effective. Harms would outweigh benefits.
I am not trying to be an infectious diseases doc here. The specifics are not the point. The point is that we always have to think about how similar trial circumstances are in our patients, and it is not easy.
I’ve been reading a lot about the IV contrast shortage, but this week it came to Kentucky. The email from the cardiology leaders induced both concern but also some topics of wide interest. Concern, of course, because contrast is incredibly important for many things we do. You can’t easily fix ST-elevation myocardial infarctions (STEMI’s) without contrast.
But, in the United States, we massively overuse procedures and diagnostics that use contrast. So, some of the suggestions made me think.
Suspension of cardiac computed tomographic angiography (CTA) for coronary disease. We can do physiologic testing, or the volume of contrast used for a cath is half of the volume for a CTA. It turns out there remains great debate about the merits of functional vs coronary CTA. Even the coronary CTA proponents would concede that if there is a delta it is small.
For pulmonary embolism (PE) evaluation, start with screening D-dimer. One of our PE experts here recommends that for these low positive D-dimer (0.5 to < 1.0) we should move forward with a lower extremity ultrasound to look for a deep vein thromboaia and then an echo for right ventricle enlargement, strain, and elevated pulmonary pressure. If this is positive and with a higher possibility of PE, then move on to CTA.
The third suggestion is one I will paraphrase. In so many words, the note encouraged all clinicians to revisit the delayed vs invasive trial of patients who had positive stress tests. Specifically, the supplement figure S6a, which plotted the rates of cardiac angiography in the two groups. The delayed strategy of medical therapy first then angiography only for recurrent symptoms, resulted in cath rates of about 25%. The invasive arm had angiography rates of 90%.
The email-message to clinicians was that, in ISCHEMIA-like patients with quite positive stress tests, you get the same rates of major adverse cardiac events over 4 years with 65% less coronary angiography if you use medical therapy and reserve catheterization for recurrent symptoms.
I want now to propose a thought. Let’s say this shortage really hits hard and it persists. The sad thing is that there will be individual patients who will be harmed. The provocative question I want to raise is that on average, will the changing norm of being more selective about use of PE protocols, caths after stress tests, and coronary CTA over functional testing actually lead to better healthcare outcomes? Or at least the same outcomes at less cost?
New Statement on Treatment of Diabetes and CKD
During the recent American Diabetes Association (ADA) meeting, a joint nephrology and diabetes scientific statement strongly recommends elevating SGLT2 inhibitors class to first line in people with diabetes and worsening CKD. I totally support this notion. We in cardiology, whether it is general cardiology or electrophysiology, see bunches of patients who also have diabetes or CKD and are not on SGLT2 inhibitors. I think we should take the lead and switch them. I’ve begun to do that myself, at least in patients with type 2 diabetes.
A side note here is that there was also an observational analysis presented at ADA that found an association of SGLT2 inhibitor use and less atrial fibrillation. The news report from journalist and PhD Michel Zoler also notes other observational studies that find the same. The main barrier to use of these drugs is not side effects, but costs. I don’t have an answer to that problem because it is a policy issue, and I am a doctor not policy maker.
Another suggestion from that same scientific document bothered me a bit. And I want to raise it here more as a question than a proclamation. The document urges rapid addition of the new nonsteroidal mineralocorticoid receptor antagonist (MRA) finerenone for further renal protection. Let’s pause there for a moment because I am sure this new agent is far costlier than the two established MRAs—spironolactone and eplerenone.
Finerenone was studied in FIDELIO-DKD a placebo-controlled RCT in 5700 patients with diabetes and CKD looking at primary composite renal endpoint. Finerenone definitely did better with a statistically significant 18% reduction in the primary endpoint.
Finerenone also reduced cardiovascular (CV) outcomes in patients with CKD and diabetes in the placebo-controlled FIGARO-DKD trial. This was a statistically significant 13% reduction in composite CV outcome driven mostly by fewer HF hospitalizations.
The problem I see for elevating this drug is the control arm of these trials—placebo. The trials are fine for regulatory approval, but the question I propose is, would we get the same effect from less a costly MRA? Is it a class effect?
And before using a super-costly MRA, should we not have proof that is superior to other MRAs. Yes, I know some say it may be better tolerated or have less hyperkalemia, but we must always remember that money spent on one thing can mean less money available for other things.
If there are nephrologists who accidently happen on to this podcast, please let me know in the comments section.
© 2022 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: Jun 17, 2022 This Week in Cardiology Podcast - Medscape - Jun 17, 2022.