DR. RADER: Our topic today is high-density lipoprotein (HDL); HDL as a target for new therapies. This is a topic that’s been of great interest to a lot of people, and not without its controversy. What we’ll do is talk a little bit about current approaches to raising HDL, or to targeting HDL, and then talk about new treatments in development.
So let’s first talk briefly about the epidemiology. There is incredibly strong epidemiologic association of low HDLs with coronary disease. That’s why we’re interested. But there have been recent issues that suggest that certain types of genetic causes of low HDL don’t seem to be necessarily associated with increased risk; as well as of course, some well known drug trials that were not all positive. There has been some questioning about HDL and its importance.
Peter, I wonder if you could start and address the issue of the epidemiology of HDL and the issue of causality versus association. What’s your take on the strong epidemiologic association of HDL levels and coronary disease?
DR. TOTH: The epidemiologic relationship between HDL and cardiovascular disease is remarkably consistent when you look at both genders, and people irrespective of race or ethnicity, throughout the world. We know that when it comes to HDL, when levels are low, risk for cardiovascular events is high. And, with some exceptions, when HDL is high, risk tends to be lower.
So ultimately what is it about HDL particles that give rise to this apparent atheroprotectiveness? Is the relationship between HDL cholesterol (HDL-C) and risk for coronary artery disease (CAD) biologically plausible? We believe the most important atheroprotective function of HDL is to drive reverse cholesterol transport (RCT): the process by which excess cholesterol is mobilized from the interior of macrophage foam cells resident within the subendothelial space of arterial walls, bound, and transported back to the liver for disposal. RCT has been verified in a variety of animal models and in humans.
Recently, it was demonstrated in humans that the capacity of HDLs to induce cholesterol efflux from foam cells is related to risk for coronary heart disease. For each one standard deviation increase in flux capacity, there is a 25% reduction in risk for coronary heart disease. It’s very impressive and confirms the importance of HDL functionality. We also know that HDL has a very complex proteasome, up to 75 different proteins, including apoproteins, enzymes, globulins, complement components, that influence its functionality. The HDLs are also important vehicles for delivering micro RNAs and sphingolipids to systemic tissues.
At least in vitro we know that HDL can exert a broad variety of anti-atherogenic effects, including reducing low-density lipoprotein (LDL) oxidation, modulating thrombotic capacity, regulating inflammation and insulin sensitivity, and participating in immunity. So there is biological plausibility to the epidemiological findings. And I think there is a body of evidence that in fact it does play into what goes on in vivo in humans.
DR. RADER: Bryan, I wonder if you could just give a very brief overview of the key players in HDL metabolism and then maybe give your take on this issue of flux and function compared to simple plasma levels of HDL.
Dr. Brewer: I think there was a big breakthrough in our understanding of how HDL removes excess cellular cholesterol with the discovery of the adenosine triphosphate–binding cassette (ABC) A1 transporter as the genetic defect in Tangier disease. The discovery of the ABCA1 transporter provided us with a mechanism by which HDL is able to efflux or remove cholesterol from cholesterol loaded cells. For a long period of time it was not clear how HDL was able to remove cholesterol from the cell. We also recently discovered that the major ligand for the ABCA1 transporter was pre-β-HDL or the lipid poor newly synthesized A-I.
After the discovery of the ABCA1 transporter, a second-transporter, ABCG1, was identified and shown to facilitate cholesterol efflux from cholesterol loaded cells. The ligand for the ABCG1 transporter is the mature αHDL. An additional major component of the cholesterol flux from the peripheral cells to the liver involves the lecithin-cholesterol acyltransferase (LCAT) enzyme, which esterifies plasma cholesterol and converts the pre β-HDL to the spherical αHDL.
There has also been a major conceptual change in our understanding of the pathway of HDL cholesterol transport back to the liver. Approximately half of the HDL cholesterol is transported to the liver via the cholesteryl ester transfer protein (CETP) to the β containing lipoproteins and ultimately back by the LDL receptor. The remaining approximately 50% of the HDL-C goes back directly from αHDL to the liver following binding to the hepatic scavenger receptor class B type 1 receptor. In the transport of cholesterol from peripheral cells it is important to note that the half-life of the apolipoprotein (apo) A-I HDL protein is about four days, whereas the cholesterol has a half-live of only hours. Thus the HDL particles load and unload cholesterol several time during the half life of the HDL particles.
There are two difficulties in looking at HDL, and specifically HDL-C in terms of the efficiency of RCT. The first problem was the discovery that approximately 95% of the HDL-C is synthesized by the liver and intestine and <5% of the HDL-C is coming from peripheral cells including the cholesterol from the coronary arteries. The second difficulty as we have discussed is that the cholesterol flux through the HDL pathway from peripheral cells to the liver is not reflected in the plasma HDL-C level.
You can have low HDL with a very efficient system and effective HDL remodeling. A low plasma HDL-C level suggests that there is reduced RCT when in fact it could be normal or even increased. In addition, a high HDL-C level may mean that the system is not fluxing very effectively. Thus, the level of cholesterol in HDL is not very useful in evaluating the potential efficiency of the RCT process.
DR. RADER: The drug that we primarily use in clinical practice for raising HDL is niacin. This is certainly the most effective HDL raising drug we have. And many clinicians for years have prescribed niacin for people with low HDL including, I suspect the four of us around this table. So we recently had a bit of a shock when a trial designed to test the benefit of niacin added to a statin, in people with coronary disease and low HDL, the Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglycerides: Impact on Global Health Outcomes (AIM-HIGH) trial1 was stopped because, basically there was no evidence of benefit.2
Jean-Claude, I wonder if you could give us your take on that trial; what’s known, which at this particular point is not much. Based on what we know now, what is your take currently on the use of niacin in clinical practice based on that result?
DR. TARDIF: Well first, before going directly to the AIM-HIGH trial, there had been a lot of data generated with niacin over the last decades, so it’s not only looking at AIM-HIGH, but I think it’s putting a lot of data into perspective. I’m thinking of the Familial Atherosclerosis Treatment Study (FATS)3 and the HDL Atherosclerosis Treatment Study (HATS)4 that Greg Brown had led that had suggested that niacin would induce favorable effects in terms of the progression of disease with different imaging modalities. There was also a non-invasive carotid ultrasound study called Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol (ARBITER) 2 that also suggested that you could favorably alter carotid atherosclerosis using niacin.5
Then finally there was also the long-term follow-up of the Coronary Drug Project that had suggested significant benefits of niacin, both in terms of non-fatal myocardial infarctions and mortality in the long-term.6 Now these studies were conducted, especially in the Coronary Drug Project, in a different era, but there have been a lot of data suggesting that niacin would work.
Then there was the news that AIM-HIGH was stopped because of no benefit. AIM-HIGH was a study of 3,300 patients with vascular disease, low HDL, and high triglycerides that were randomized to either simvastatin 40 mg alone, or simvastatin plus niacin.1 The primary endpoint was a composite of hard outcomes; the study is going to be presented within the next two days and published in a major journal.
We know that there was no trend in the right direction for the primary endpoint, which was the composite of cardiovascular events, non-fatal myocardial infarction, and non-fatal stroke. And there was also a trend in the wrong direction for strokes. Several of these strokes actually occurred after patients had been off medication for a while.
There are a couple of issues in terms of the study design of AIM-HIGH that may have had an impact on these findings. The investigators were not blinded to the LDL cholesterol (LDL-C) values; that means that we’ll need to be looking carefully at imbalances in terms of statin use and dosage, as well as the use of ezetimibe. It would not be surprising, for example, that more patients in the placebo or simvastatin alone arm would have used higher doses or more ezetimibe. So what part this could have played into what was observed needs to be rethought.
Since your question was also what to make of these findings, there is also a major study called HPS2-THRIVE that’s ongoing, with more than 25,000 patients.7 These 25,000 patients are treated to goal first, and then are randomized to either niacin with laropiprant or placebo. There will not be further adjustments in terms of LDL lowering. I think this is probably a significant methodological issue. If implied in your question was, “Does niacin have an effect on other HDL raising drugs?” I think niacin has a number of other effects, not only in HDL but on triglycerides and LDL-C, perhaps on the vessel and inflammation, so I think it’s impossible to say that the HDL hypothesis was really tested in this study.
In terms of how to use niacin after AIM-HIGH, these methodological issues I think will have a serious impact, and we’ll probably have to wait until Treatment of High Density Lipoprotein to Reduce the Incidence of Vascular Events (HPS2-THRIVE)7 reports to really know exactly. I don’t want to avoid the question, but I would not personally stop niacin in patients. I may be slightly less inclined to start new patients on the medication.
DR. RADER: Peter, do you have a comment on that?
DR. TOTH: Yes, I think we have to be very careful in our interpretation of AIM-HIGH because the temptation is now to discount niacin as an agent that is of therapeutic value and efficacy. That’s the wrong thing to do because AIM-HIGH was light years beyond the other studies that we’ve seen.
You’re talking about patients who, at the time of randomization, have an LDL of 71, non-HDL 106, and an apoB of 81, and intensive background therapy with aspirin and thienopyridenes, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta blockers, and intensive regulation of blood sugars among diabetic patients (mean a1c 6.6%), it is going to be challenging to observe incremental benefit from yet another drug. Moreover, a lot of these patients had already been chronically treated with statin plus or minus ezetimibe. HDL-C was only mildly reduced at 35 mg/dL. This is not the same population as what you encountered in HATS, FATS, the Armed Forces Regression Study (AFREGS), or the Coronary Drug Project. I think when you look at patients who approximate what you saw in other trials, the patients who have atherogenic dyslipidemia, niacin clearly is therapeutically beneficial.
I think another temptation we have to avoid is that AIM-HIGH was a definitive test of whether or not the results of HATS were real. That’s wrong because HATS was a placebo-controlled study with a mean starting LDL-C of 124 and HDL-C of 31. Clearly there was benefit with an 89% reduction in risk for the primary composite endpoint (when comparing simvastatin/niacin therapy to placebo) and evidence for coronary atherosclerotic plaque regression as assessed by quantitative coronary angiography. The study was small. But they’re very different studies.
I think if we ask the question, if I saw a patient with the same level of background therapy as we see in AIM-HIGH, would I give that patient niacin? No. But that probably represents about 12% or 15% of the total patient population in the US and elsewhere with coronary disease.
I guarantee you that the average patient you encounter in community practice is not this intensively treated. We know that the percentage of patients who hit a very high risk target of LDL less than 70 and a non-HDL approximating 100 is around 13 or 14%. So I think we have to be very careful about drawing conclusions.
DR. RADER: So I get the feeling you haven’t given up on niacin yet?
DR. BREWER: I haven’t given up on niacin as yet. I think two aspects of the AIM-HIGH trial that we need to look at very carefully are the difference between the placebo group and the treated group in terms of changes in their HDL lipoproteins, and the fact that the study was designed with a projected goal of a significant reduction in clinical events as great as 25%.
There was only a final difference of 4 mg in HDL-cholesterol between the niacin treated group and the control group. This difference in HDL-C levels would not be able to achieve a 25% reduction in clinical events. Some of these points are going to be very important parameters to review in order to decide whether this trial effectively tested the HDL hypothesis, or the role of niacin per se in the treatment of high risk cardiovascular patients.
DR. TOTH: I reject the proposition that this study rigorously tested “the HDL hypothesis.” And I think the other thing we have to be careful about is that during the trial they talked about a numerical excess risk of ischemic stroke, but on therapy this was only 19 compared to 12 in the niacin and placebo groups, respectively. That’s seven excess events, and I’m afraid that’s not going to convince anyone that there’s an excess hazard attributable to niacin for ischemic stroke. Moreover, no other trial has ever shown an excess risk of ischemic stroke attributable to niacin.
DR. RADER: Bryan, could I ask you to address the theoretical issue of maybe the HDL raising with niacin doesn’t promote flux and doesn’t promote a good form of HDL raising? It’s purely theory, but could you just comment on that?
DR. BREWER: I think that’s an interesting question, because you need to look at not only the HDL-C level, but also the HDL particle number, which is another parameter that we’re beginning to use to evaluate the changes in the lipoprotein profile with the use of a given drug. Obviously what we’d like to do is increase the number of HDL particles as well as increase the cholesterol flux through the HDL pathway. Based on the currently available data with niacin there is a minimal increase in the number of HDL particles but what you have with niacin treatment is very large cholesterol filled HDL particles with an increase in lipid content per particle. Whether those large HDL particles will be able to function in a number of different ways and reduce atherosclerosis is not clear. It would be interesting to test these particles in the in vitro cholesterol efflux system that you and George Rothblat have used to see how effective these particles are in removing cholesterol from cholesterol loaded cells. We also don’t know if the cholesterol flux through the HDL pathway to the liver is increased with niacin treatment.
DR. RADER: Jean-Claude, could we talk briefly about fibrates, which also raise HDL modestly? There have also been some disappointments with fibrate trials recently. Do fibrates have any role in the management of patients with low HDL?