Thomas Dayspring, MD, FACP, FNLA, NCMP, discusses the promises and pitfalls of the 2013 American College of Cardiology/American Heart Association Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults.
Internal Medicine World Report Editor-in-Chief Simon Douglas Murray, MD, consulted with Thomas Dayspring, MD, FACP, FNLA, NCMP, Director of Cardiovascular Education at the Foundation for Health Improvement and Technology and Clinical Assistant Professor of Medicine at the Rutgers New Jersey Medical School, on the 2013 American College of Cardiology (ACC)/American Heart Association (AHA) Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults.
When I looked at the new cholesterol guideline, I was somewhat confused. It's a long document and I came away thinking I didn't know what it really said. Is there a clear take-home message that internists can apply to their practices?
First of all, I agree with your assessment. One of the guideline’s missions was to simplify things, but it’s almost an unreadable document. It’s so entrenched in trying to link everything to various degrees of evidence that an average, everyday clinician will get lost in the statistical minutia.
But the takeaway is if you look at the previous National Cholesterol Education Program (NCEP) Adult Treatment Panel (ATP) III guidelines, the new guideline wanted to minimize the number of risk categories, treatment complexities, and even some diagnostic categories. Basically, they’ve come down to defining risk in certain people, and depending on which risk category a patient falls in, he or she is either statin eligible or not statin eligible, and that’s it.
Of course, there are also some lifestyle recommendations, but gone are all of the better ways to define risk.
Do you think there is bias in that?
Here’s the problem: the guideline only used what’s called Level 1 evidence, and that evidence has been garnered from large, empowered, randomized, double-blinded trials. So, basically, they eliminated about 90% of the knowledge we currently have on understanding atherothrombosis, its risk assessment, its diagnosis, and its aggressive treatment. Now, this would have been good if statin therapy had numerous trials showing 100% event reduction. But, as we all know, if you do a meta-analysis of the statin trials, you get somewhere between a 20%, 30%, or if you’re lucky, 40% event reduction, which spun the other way would be a failure to reduce 60-70% of the events.
The guideline should have encouraged a physician to use his experience and lesser degrees of evidence from clinical trials and say, “We’ve got you covered with your statin, but we still see some abnormal tests, and we want to try additional lifestyle, additional medication in your case to at least eradicate things that tell me you are at risk.” But the guideline just eliminated that.
What other biomarkers were not included in the guideline?
In the old days, they did want you to follow up with low-density lipoprotein cholesterol (LDL-C) and triglycerides, and then try to identify people who have residual risk. But now, they don’t really even care if you do a follow-up lipid profile because the patient is on the statin, and they will tell you that statins work no matter what they do to LDL-C.
If you want to make it even simpler than that, the old days would tell you that if a patient’s statin or lifestyle normalized their LDL-C, then make sure you look at triglycerides or high-density lipoprotein cholesterol (HDL-C), because if those are still abnormal, the patient is probably going to have residual events, despite their good-looking LDL-C.
That is where the guideline introduces what it calls emerging biomarkers, like C-reactive protein (CRP), inflammatory markers, lipoprotein testing, and coronary calcium scoring. Many people thought they were going to suggest calculating the non-HDL-C level, which is a better marker than LDL-C in correlating with apolipoprotein B (apoB) and LDL particle (LDL-P) concentrations. But because there’s never been a trial designed to enroll people using non-HDL and use that as the goal, there’s no Level 1 evidence, so the guideline ignored that incredibly useful biomarker.
Can you explain what apoB is?
Remember, you have to understand the disease we’re trying to eliminate is atherosclerosis, which is a build-up of cholesterol in deposits called plaques in your artery wall. That can ultimately lead to a narrowing or blockage of the artery wall. So, we know a patient cannot have atherosclerosis if they don’t have cholesterol in their artery wall. But the only way cholesterol, which is a lipid, gets into your artery wall is by coming out of cells that produce it. It’s packaged within a protein-wrapped vehicle called the lipoprotein; hence, it solubilizes the cholesterol in the plasma.
So, you’ve got to come to the understanding that atherosclerosis is a lipoprotein-mediated disease, in which case you then have to say, “How do I assay lipoproteins in the laboratory?” Well, the simplest test that’s been around for 50 years is cholesterol measurements like total cholesterol or the cholesterol within the LDL, because that correlates well with atherosclerosis. That’s why we’ve been measuring cholesterol for a long time, but we now know it would better if we just counted the lipoproteins that have the ability to crash the artery wall — and that’s where LDL-P count or apoB comes in.
The protein that enwraps LDL and your very low-density lipoproteins (VLDL) is apoB. There’s one molecule of apoB on every VLDL and every LDL, but because an LDL has a half-life of several days and a VLDL has a half-life of a few hours, almost all of your apoB particles are LDLs. So, if we measured 1 apoB per LDL-P, that is a way of counting how many LDL-Ps are in your blood. That is very different than assaying LDL cholesterol, which simply is the mass of all the cholesterol molecules in all LDL-Ps.
Does that mean risk can be determined by LDL-P size?
LPL-P size often correlates with LDL-P count, but in many cases, it doesn’t. So, apoB or LDL-P counts are better tests.
To finish this story, apolipoprotein A1 (apoA1) is the protein that enwraps HDL. The one thing the new guideline got right is it told you not to follow HDL-C, because whatever you do to HDL-C with any therapy will have zero correlation with the patient’s outcome.
If you measure apoA1, it’s sort of a way of counting how many HDL particles (HDL-Ps) you have, which can be very different than the mass of cholesterol they are carrying, which is assayed by HDL-C tests. So, apolipoprotein tests are basically ways of counting particle numbers, and cholesterol is just how much cholesterol is in those particles. They often correlate, but in many cases, they don’t, so most people who want to do serious risk assessment prefer to count particles in addition to measuring cholesterol content.
If a patient has very few apoBs and a lot of apoA1s, then that correlates pretty well with low risk. And, of course, if the ratio is high — meaning way too many LDLs and too few HDLs — then that’s associated with high risk.
What is a good goal for apoB?
Every goal of therapy depends on the risk of the patient, so you’re going to have a stricter apoB goal if you’re treating a patient who’s already survived a heart attack, as opposed to a young person whose risk lies in the future. If we take the average person, we like to get lipid and lipoprotein concentrations below what’s called the 20th-percentile cut point, which for apoB would be 80 mg/dL, for LDL-P count would be 1,000 nmol/L, for LDL-C would be 100 mg/dL, and for non-HDL-C would be 115 mg/dL. If the patient was indeed a nightmare — had coronary events and was super high risk — many believe you should go for 5th-percentile cut points, and that would be 60 mg/dL for apoB, around 800 nmol/L for LDL-P counts, 70 mg/dL for LDL-C, and probably around 85 mg/dL for non-HDL-C.
In our population now, which tends to be very insulin resistant and have triglyceride abnormalities, those are the people who have what we call discordant levels of LDL-C and LDL-P counts or apoB. Their LDL-C looks really good, but if you count the number of atherogenic particles, they’re still high risk. And so, what I believe is you have to measure both, because in some cases, normal LDL-C is accompanied by the high particle counts, and that’s residual risk.
So, a patient with relatively low LDL can still be at risk?
In general, it takes more small than big LDLs to carry a given mass of cholesterol, but if for some reason the LDL is enriched with triglycerides, that means it’s not carrying very many cholesterol molecules. It can’t carry both, so people who have big LDLs that are triglyceride-rich have horrific LDL-P counts. Anything that depletes LDL-C — be it small particle size or triglyceride engorgement — will result in a high apoB or LDL-P count in the face of a normal LDL-C.
To explain how triglycerides get into LDLs, there’s a protein that traffics them in, which can happen in people with triglyceride levels as low as 120. It’s another reason to measure apoB or LDL-P count and not even worry about the size.
When I receive a lab analysis, one of the things that comes back besides LDL-C is remnants. What is the value of knowing what remnant particles are?
A remnant is sort of a shadow of its former self. The liver secretes VLDLs, which are triglyceride-rich particles that bring the fatty acids to the muscles or fat cells.
Think about this: a VLDL carries 80-90% triglycerides and only 10-20% cholesterol, but as it shrinks and loses the triglyceride, you still have a rather large apoB containing a particle full of cholesterol, like a very big LDL-P. Now, you would hope the liver would clear that remnant very quickly as is supposed to happen. But if it didn’t — and there are reasons why it might not — then that cholesterol-enriched VLDL, now called the remnant, could crash your artery wall, and particle for particle, it carries more cholesterol than an LDL.
Remnants contribute to atherogenesis, so we don’t like to see them floating around. Most believe if we had therapeutic measures — and we do with lifestyle changes and various drugs — we could eliminate remnants in addition to LDL-Ps, which would likely be beneficial.
If you could write your own guideline for the measurement of lipids and reduction of cardiovascular (CV) risk, what would it be? Would you lump stroke, coronary artery disease (CAD), and peripheral vascular disease together and say lowering cholesterol would reduce all of those risks?
First of all, I believe history is very important. Premature heart disease in a first-degree relative of either gender would be important to know, as well as smoking status. I would do a physical examination quickly to look for xanthomas or something bizarre that would immediately identify a genetic lipid disorder. Of course, I would check blood pressure and pulses, and if I don’t feel pulses in the feet, then the patient has peripheral vascular disease, which also tells me they are at very high risk for CAD and extracranial arterial disease.
But let’s presume there’s no family history, all pulses are great, and I don’t see any xanthomas. Then, my only prayer of doing a risk assessment is looking at various biomarkers. There’s emerging evidence that total HDL-P count more than apoA1 level is incredibly desirable, and that low total HDL-P counts are quite a bad risk factor and a better predictor than any HDL-C measurement. So, those are 2 things I want right off the bat.
The most atherogenic lipoprotein known to afflict humans is called lipoprotein(a) [Lp(a)], which is an LDL-P that’s carrying a potentially pathogenic protein on its surface. This is way more common than familial hypercholesterolemia (FH), as it’s estimated a third of the American public have a high-risk Lp(a). It’s a genetic test that’s only needed once in a patient’s lifetime, but patients either have it or they don’t. I need that on the chart somewhere, because if they have that, then I’m worried about them even if everything else comes back clean.
What level of Lp(a) would alarm you?
If you’re measuring Lp(a) mass, considerable risk is typically above 30 mg/dL. If you’re doing it in molar concentration, it’s 75. There’s a new test called Lp(a) particle counts, and if it’s above 75, that’s considerable risk. So, there are different ways of assaying Lp(a) and you have to become familiar with the various ones out there, as each has different levels.
Is there a treatment option for high Lp(a)?
The treatment right now is to blow away LDL-P count and apoB level, and maintain blood pressure control, smoking control, and healthy diets. No drug has data showing if it lowers Lp(a), patients will have less heart attacks. In fact, we were stunned last year when a big trial with niacin — which does lower Lp(a) — reduced Lp(a) but didn’t reduce one heart attack. However, the Justification for the Use of Statins in Primary Prevention (JUPITER) trial said even though statins do not lower Lp(a), they significantly reduce clinical events in primary prevention settings in patients with high Lp(a).
Are there any other significant inflammatory biomarkers?
There are many biomarkers a lipidologist like myself uses to fully ascertain risk, but what you’ve got to do is count particles, perform some rudimentary inflammatory tests, and measure Lp(a) in addition to the historical and physical exam findings.
For the inflammatory markers I previously mentioned, there are plenty of trials showing no matter what type of lipoprotein abnormality you have or don’t have, they give you more CV risk information in a given patient. If 2 people came to see me with identical apoB or LDL-P counts, I would worry a little bit more about the one who has concomitant inflammatory elevations, as opposed to somebody who doesn’t. In my belief, the better a risk assessment I do, the better I know what intensity of treatment is necessary. And depending on what’s abnormal, I can customize the treatment.
Say a patient is at great risk for an impending event and maybe we better check out if he or she has soft plaque. Is it possible to decipher that?
No, but an elevation of inflammatory biomarkers would certainly suggest there is unstable plaque present. So, again, you’d look at particle counts and other treatable risk factors like blood pressure, smoking, and insulin resistance, in which case a low-carb diet or even metformin would help. You’d also assess platelet function and tell the patient to take an aspirin, or if they’re aspirin-resistant, perhaps a more potent platelet inhibitor. You could run a thyroid function on everybody who’s got a lipid abnormality. You might also look at other types of factors that might contribute to atherogenesis or unstable plaque, such as very high homocysteine levels, low vitamin D, and Lp(a).
If you are already practicing that way, did the guideline add anything?
They took away half of the tools we were using and limited us to only one treatment, even though they know statins, as good as they have been, leave a horrific residual risk. If you look at it, the only thing different from the old Framingham score is they stuck the patient’s race in there. I’d almost say flip a coin: heads you treat, tails you don’t. And I think you would do just as well as you would by following the new guideline.
We’re in a world now where you have to do far better risk assessments using imaging and biomarkers. Many people do both, and without that — unless you have somebody with FH — you’re going to miss so many people, and you’re probably going to over-treat some people with a statin who don’t need one.
Are there other available treatment options besides statins?
Even though we have all sorts of big trials, there are secondary analyses that show if a patient is on a statin and has high triglycerides and low HDL-C, then drugs like fibrates, fish oils, and even niacin bring additional benefit to the table. We have studies where they’ve added ezetimibe, a cholesterol absorption blocker, to a statin and drastically reduced the ischemic events in patients with kidney failure and aortic stenosis. So, there’s lots of evidence that these other add-on drugs — which they’re not telling you to use in place of a statin — will help the statin do better in high-risk people.
This is not Moses from the mountains. This is not the Ten Commandments. You’re not a felon if you go beyond what they’re telling us now, and if we don’t, then I don’t think we’re going to make any progress in coronary heart disease.
Thomas Dayspring, MD, FACP, FNLA, NCMP, is Director of Cardiovascular Education at the Foundation for Health Improvement and Technology in Richmond, VA, and Clinical Assistant Professor of Medicine at the Rutgers New Jersey Medical School in Newark, NJ. He is the author of several textbook chapters and peer-reviewed articles relating to lipids and lipoproteins, including their pharmacologic modulation and relationship to gender, estrogen, and raloxifene.