COMMENTARY

# Feb 11, 2022 This Week in Cardiology Podcast

John M. Mandrola, MD

Disclosures

February 11, 2022

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 February 11, 2022, John Mandrola, MD comments on the following news and features stories.

#### LDL Equations

The Journal of the American College of Cardiology (JACC) has published a paper looking at the clinical implications of different LDL cholesterol (LDL-C) estimates.

I think we all know this, but nearly all LDL measurements are actually estimates using total cholesterol (TC), triglycerides (TG), and very low-density lipoprotein (VLDL). Three equations are used; the most common is the Friedewald equation, then the Sampson and Martin/Hopkins equations. Direct measurements of LDL are lengthy and expensive and not routinely done.

The other relevant fact here is that treatment guidelines often turn on driving the LDL-C below 70 mg/dL (1.81 mmol/L). So if you underestimate a higher LDL, you may undertreat some patients because you think the LDL is in range when it is not.

The study group, first author, Aparma Sajja, conducted an electronic health record review of patients who had heart disease and TG levels < 400 mg/dL (4.52 mmol/L). They estimated the LDL-C in 146,000 patients using the three equations.

• Patients were described as concordant if LDL was < 70 mg/dL (1.81 mmol/L) with each equation.

• Patients were described as discordant if LDL was < 70 mg/dL for the index equation and ≥ 70 mg/dL with another equation.

• The Martin/Hopkins equation consistently estimated higher LDL-C values than the Friedewald and Sampson equations.

• Discordance rates were 15% for the Friedewald vs Martin/Hopkins comparison, 9% for the Friedewald vs Sampson comparison, and 7% for the Sampson vs Martin/Hopkins comparison.

• Discordance increased at lower LDL-C cutpoints and in those with elevated TG levels.

• Among patients with TG levels of ≥150 mg/dL (1.69 mmol/L), a >10 mg/dL (0.26 mmol/L) difference in LDL-C was present in 67%, 27%, and 23% of patients when comparing the Friedewald vs Martin/Hopkins, Friedewald vs Sampson, and Sampson vs Martin/Hopkins equations, respectively.

I like this paper because it’s a nice use of observational data. The authors make a reasonable conclusion that “clinically meaningful differences in estimated LDL-C exist among equations, particularly at TG levels of ≥ 150 mg/dL (1.69 mmol/L) and/or lower LDL-C levels.”

Here is how I see the importance of this paper: if you have a person with known vascular disease, especially if the disease is substantial and the patient is otherwise in good shape, it makes sense to do as much as feasible to reduce risk. Note the qualifier—as much as feasible.

If you have a person with multivessel coronary artery disease (CAD) and the LDL by Freidewald comes back at 69 mg/dL (1.78 mmol/L) and they are on 20 mg of atorvastatin, it makes sense to bump it up. That is easy and low burden to most patients. If you have a patient with similar amount of CAD who has the same LDL but is on 80 mg of atorvastatin and ezetimibe and the next step is a PCSK9 inhibitor, and they are already on seven drugs and living on a fixed income, well, that is a different story. The accompanying editorial is excellent. You all should read it and save it, because if you are like me, the specifics of these equations will be hard to remember. The editorialists get into the nitty gritty of the equations and tell us which is best in certain situations.

My take homes are easier to remember:

• LDL-C is an estimate.

• There are three equations.

The most common equation – Friedewald — tends to give lower LDL readings when LDL is low or TGs are greater than 150 mg/dL (1.69 mmol/L). In these cases, we might consider use of other equations. For patients with TG ≥ 400 mg/dL (4.52 mmol/L), the Sampson method may be the most accurate. For those on intensive lipid-lowering therapies with LDL-C values < 100 mg/dL (2.59 mmol/L), the Martin/Hopkins method may be favored to reduce the risk for undertreatment, given that no safety signal has been identified for very low levels of LDL-C, and atherosclerotic cardiovascular disease risk reduction seems to be proportionate to LDL-C reduction.

Finally, these were patients with established heart disease and pertain to treating to goal LDL. For primary prevention statin decisions, I don’t know that these equations make a difference. For these patients, the decision is simple: you enter their numbers into the risk calculator and tell them their 10-year risk on and off statins, and then patients can decide if the spread is worth the disutility of taking a pill every day.

#### Acute Stroke Care

A quick note on an important trial in interventional stroke care. The New England Journal of Medicine (NEJM) published the results as a simultaneous publication from a presentation at the International Stroke Conference.

Endovascular (EV) therapy for stroke is generally not done for patients with big infarcts in the middle cerebral artery region. The worry is that patients will bleed after reperfusion and get worse. However, endovascular therapy and knowledge iterates, and Japanese investigators decided to study the role of endovascular therapy in these infarcts with a large ischemic core. And the only way to study such a decision is with a randomized controlled trial (RCT). They randomly assigned 100 to receive EV plus medical therapy, vs 100 to receive medical therapy alone.

The primary endpoint was the percent of people who had a modified Rankin Scale score (mRS) of 0-3 at three months.

• mRS 0 = No symptoms

• mRS 1 = No significant disability despite symptoms; able to carry out all usual duties and activities

• mRS 2 = Slight disability, unable to carry out all previous activities but able to look after own affairs without assistance

• mRS 3 = Moderate disability requiring some help but able to walk without assistance

The results were clear:

• In the EV group, 31% had mRS 0-3 vs 13% with medical care; P = 0.002.

• In the EV group, 58% had any intracranial hemorrhage (ICH) vs 31% in medical arm (or 85% higher); P < 0.001.

This sounds bad but recall that mRS is a functional score. So if more patients are functional at 90 days, then the higher ICH rate must not have been clinically significant. The number of “symptomatic” ICH were small and not significantly different.

I see these results as statistically robust and clinically meaningful. This is one trial. It had small numbers. It was carried out in Japan and had low use of recombinant tissue plasminogen activator (rtPA). The authors note that if rtPA had been used more often or at higher doses in the trial, the outcomes might have been improved in both groups, but there might have been an increased percentage of patients with ICH in both groups. You all know how I feel about rtPA in acute stroke. It’s dubious. So perhaps this trial was positive because of the low burden of lytic therapy.

I highlight this paper because interventional stroke care looks a lot like interventional cardiology looked years ago. I think we are going to be seeing more and more interventional stroke procedures and perhaps there will be a need for more stroke interventionalists. I also wonder why there is such separation in the fields of brain and heart interventions?

#### What to Do After PCI for ACS and the Network Meta-analysis

JACC Interventions has published a network meta-analysis looking at the reduction of anti-platelet therapies after percutaneous coronary interventions (PCI) and stenting for patients with acute myocardial infarction (MI). The paper is one of the most complex papers I have read in years. First some background. PCI for acute MI is one of the purest treatments in cardiology. It’s truly spectacular, and younger listeners simply don’t have the perspective that us older folks have. You think that PCI for acute MI in which a patient suffering from acute cardiac injury gets off the table with a band-aid on the wrist is normal.

In medical school, we treated acute MI patients with morphine, aspirin, and Swan-Ganz catheters. In internship and residency, we gave lytics and prayed the artery opened and the patient did not have a catastrophic brain bleed. But here’s the thing: Even though PCI for acute MI is beautiful, there is still the issue of how best to keep the metal cage from occluding while minimizing major bleeding, which is associated with higher death rates.

Current guidelines recommend 12 months of dual antiplatelet therapy (DAPT) for patients with acute coronary syndromes (ACS). But stents have iterated: newer stents have thinner struts and are less prone to clots. And this has driven the idea to reduce DAPT so as to reduce bleeding and still prevent ischemic events. The whole thing boils down the clot/bleeding tension inherent in any antithrombotic therapy. Same issue with stroke prevention in atrial fibrillation (AF) with oral anticoagulants.

There are two ways to reduce the burden of anti-platelet drugs and hence bleeding after a stent. You can de-escalate the DAPT or you can shorten the duration of DAPT. Deescalate means lowering the dose of the more potent P2Y12 blockers like prasugrel or ticagrelor or substituting the less potent P2y12 blocker clopidogrel. the DAPT duration is self-explanatory, but of course you can shorten it by changing DAPT to aspirin alone or to a P2Y12 inhibitor alone. Here’s the problem: trials have compared de-escalation to standard DAPT and shortening DAPT to standard DAPT but we don’t have massive three-arm trials looking at all three directly. So, we have an if A = B and B = C, then A = C issue, or the transitive property. That is where the network meta-analysis comes in.

Italian authors, first author Claudio Laudani, from Catania Italy, collected 29 studies including 50,000 patients for this meta-analysis. They analyzed the results using frequentist statistics—no prior assumptions and Bayesian statistics, which incorporate priors. The primary endpoint of interest was all cause mortality, secondary outcomes were net adverse cardiac events (NACE), and major adverse cardiac events (MACE). The idea was to study short DAPT vs de-escalation.

The results of this indirect comparison of short DAPT and de-escalation, using standard DAPT as a common reference, can be summarized as follows:

• Both short DAPT and de-escalation carried a similar risk for death and cardiovascular death.

• Short DAPT significantly reduced the risk for major bleeding, ranking first in the Bayesian hierarchy of treatment strategies with respect to this endpoint, and increased the risk for NACE.

• De-escalation significantly reduced the risk for NACE, increased the risk for major bleeding, and ranked first in the Bayesian hierarchy of treatment strategies with respect to MACE, MI, stroke, and stent thrombosis.

Recall that this was not a direct comparison of short DAPT or de-escalation vs standard DAPT but an indirect comparison. At first I was wondering why de-escalation would increase major bleeding, but I think this was because de-escalation was compared to short DAPT. As the authors note then, short DAPT may be a better strategy than de-escalation in patients with high risk for bleeding.

However, de-escalation resulted in larger net benefit compared with single-antiplatelet therapy. Such benefit in the context of a higher risk for bleeding implies a probable reduction in ischemic events.

The significance of this analysis is that currently patients with ACS and PCI are recommended to receive 12 months of DAPT but tailoring the duration of DAPT on individual characteristics for ischemic or bleeding events is an option. Shortening DAPT carries a 2a recommendation while de-escalation is given a 2b. But these results challenge this ranking as de-escalation showed a similar risk of death and a reduced risk for NACE compared with short DAPT.

There are limitations here. These were indirect comparisons with assumptions. They say the transitivity assumption was met, but it’s hard for me to understand exactly how that is known and if it is totally reasonable to make these comparisons.

On Twitter I was discussing the complexities of using this language in the paper and David Cohen Tweeted a slide he had used during a talk he gave at ACC that showed that Harvard might be better than Duke in basketball by transitive logic. Harvard had beat Boston College, Boston College had beat Maryland, and Maryland had beat Duke, so by this logic, should we infer that Harvard is better than Duke? Or is it possible that the studies or in this case games were not entirely poolable?

Final comment: Having looked at the post PCI studies including this one, I am not sure we are ever going to be able to reduce this to an algorithm. I may be wrong but there are just too many variables:

• PCI factors: Skill in placing it, stent size, how many stents, their location, etc;

• Patient factors: Bleeding risk, MACE risk;

• Drug factors: Different potency, different duration, different combinations.

The average effect from trials get us into a ballpark, but I think docs are going to have to individualize. Always remember that here we are talking ACS and AMI, and stents save lives in this situation. For those with stable CAD, we avoid these complexities if we treat CAD medically—which is just as good for mortality, prevention of MI, and in many cases angina reduction.

#### Lipoprotein(a)

Topline results of a study to be presented at ACC as a late breaker were released this week in a press release. It concerns lipoprotein(a) [Lp(a)]. Lp(a) is a low-density lipoprotein with an added apolipoprotein—hence the “a.” Etiologically, it may have provided a survival advantage by aiding in wound healing and reducing bleeding, particularly in childbirth.

An individual's Lp(a) level is 80%-90% genetically determined in an autosomal codominant inheritance pattern with full expression by 1 to 2 years of age and adult-like levels achieved by approximately 5 years of age. Outside of acute inflammatory states, the Lp(a) level remains stable through an individual's lifetime regardless of lifestyle.

High Lp(a) affects about one in five people worldwide and is a genetic risk factor for cardiovascular disease. There are no approved medications that selectively lower Lp(a), and levels cannot be significantly modified through lifestyle changes or any approved medications.

A company called Silence Therapeutics has designed a drug – without a name yet — that can reduce Lp(a) in a dose-dependent manner. The drug is a short interfering RNA (siRNA) injection. It works to lower Lp(a) production by using the body's natural process of RNA interference to target and silence messenger RNA that is transcribed from the Lp(a) gene in liver cells.

• The APOLLO trial is a phase 1 study and it found that the siRNA drug maintained reductions in Lp(a) of up to 81% out to 150 days.

• There were no safety concerns noted and fu has been extended to 1 year.

• The inimitable Steve Nissen is the principal investigator.

Of course, this is super super early. Phase 1 is a long way from Phase 3 when a drug must reduce clinical outcomes with adequate safety. But I have to say I think we are living through a pretty incredible period of drug development in the lipid space. I don’t know if this will pan out, nor if it will ultimately be affordable, but it is scientifically very neat.