DR. LAROSA: We are here to talk about high-density lipoprotein (HDL) and our knowledge about it and what, if anything, we should be doing about it. First, do we believe that HDL is a reliable biomarker of atherosclerosis risk? After all, many epidemiologic studies have demonstrated that HDL levels are low in people with low animal-fat intake, and so, a simple inverse relationship with HDL is not universally observed.
DR. WONG: It’s certainly well-established from studies dating back to the Framingham study1 that a low HDL level is associated with higher cardiovascular risk. We also have data from studies on patients with stable coronary artery disease (CAD), such as the Treating to New Targets (TNT) trial,2 showing that even in people with well-controlled low-density lipoprotein (LDL) levels (eg, approximately 70 mg/dL), there is a residual risk for subsequent coronary heart disease events in those with a low HDL cholesterol level. Therefore, I think most people agree that it’s a good biomarker of cardiovascular disease risk. Whether it is really a risk factor and whether we can really modify it and reduce the risk have been seriously contemplated on in recent studies, in particular, in the Atherothrombosis Intervention in Metabolic Syndrome With Low HDL/High Triglycerides: Impact on Global Health Outcomes (AIM-HIGH) trial,3 although that trial was conducted in patients with known CAD, and they were successfully treated to achieve an LDL level of approximately 70 mg/dL.
I think there’s a general consensus that healthy levels of HDL are good for one’s cardiovascular health and that we should strive to attain such levels (at least nonpharmacologically). In addition to controlling the LDL cholesterol level, which remains the main target for lipid treatment, the question remains about whether an increase in HDL cholesterol is going to provide added benefit. Therefore, we need to wait for further data from upcoming trials like Heart Protection Study 2-Treatment of HDL to Reduce the Incidence of Vascular Events (HPS2-Thrive)4 and some of the cholesteryl ester transfer protein (CETP) inhibitor trials7 to better understand the facts.
DR. GOTTO: I wish to iterate the same point as Dr. Wong: evidence about HDL having protective effects dates back to before the Framingham study,1 to Gofman’s analytical ultracentrifuge8 studies in the 1950s that showed an association between cardioprotection and higher levels of HDL fractions. A paper from 1951 in the American Journal of Medicine showed higher levels of HDL in women.9 In addition, there were a number of early studies dating back to the Lipid Research Clinics trial,10,11 in which the calculations appeared to show a benefit of high HDL. In this trial, there was a relative risk reduction of approximately 19%, which could be explained mainly by the LDL level reduction, and a small increase in HDL levels was found to be independently associated with a reduced cardiovascular risk in the treatment group. Similarly, in the Helsinki Heart Study,12 a reduced cardiovascular risk was noted primarily in individuals with the lipid triad, but again, this was explained by the combined reduction in LDL cholesterol levels and the increase in HDL levels. In the Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial (VA-HIT),13 the benefit was more closely related to the small increase in HDL than to the changes in triglyceride concentration, and the LDL levels did not change at all. So, while it hasn’t been conclusively proven that increasing the HDL level will reduce the risk and increasing the HDL level may depend on the mechanism involved in increasing the level and type of HDL that results from this process, our data over the years certainly support the concept that increasing HDL levels may be beneficial and protective.
DR. RADER: I certainly agree with that. I think HDL is actually an excellent biomarker of cardiovascular risk. Epidemiologically, there’s a very strong inverse association between HDL and CAD. There are certainly environmental factors such as a low-fat diet that lower the HDL level but don’t increase the risk, and very high levels of alcohol intake increase the HDL level but don’t necessarily reduce the risk. From a genetic standpoint, there are genes that cause a reduction in HDL levels that don’t seem to increase the cardiovascular risk, and there are other genes that cause an increase in HDL levels that don’t seem to reduce the cardiovascular risk. So, I think that for an individual patient, the HDL level is not necessarily a perfect predictor of risk, and HDL responds particularly to pharmacologic intervention but has not been proven to be a marker of the success of that pharmacologic intervention. Much more work is needed in that regard.
DR. LAROSA: Would you believe that HDL is a better predictor in populations in which the LDL levels are in the atherogenic range compared with populations in which the LDL levels are quite low?
DR. GOTTO: Yes. In Turkey, for example, there is a genetically determined low level of HDL. My impression from observing Turkish patients over many years is that outside the cities where the diet is different and the LDL and cholesterol levels are not as high, a low HDL level is not associated with such an increase in the cardiovascular risk. When people move into the cities and their diet changes, the LDL level increases and the low HDL level becomes a more significant predictor.
DR. RADER: Yes, I think Dr. Gotto made a good point, although LDL and HDL levels are clearly statistically independent predictors of risk when viewed separately in large populations. However, I think that as one particularly achieves lower levels of LDL, as Dr. Gotto suggested, the risk associated with a low HDL level seems to be attenuated.
DR. LAROSA: And that’s certainly in keeping with both the TNT trial2 and the Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin (JUPITER),14 isn’t it?
DR. WONG: Yes, and I think this is primarily why we’ve seen benefits in earlier trials, such as the Coronary Drug Project,15 the Familial Atherosclerosis Treatment Study (FATS),16 and the HDL Atherosclerosis Treatment Study (HATS),17 involving many more people who did not have well-controlled atherogenic lipoprotein levels. In the more recent studies, such as TNT2 and, of course, AIM-HIGH, you’ve had people who were very well treated, and so, the effect of a low HDL level and the impact of managing it were attenuated.
DR. LAROSA: I think we all agree that HDL is a good biomarker. In fact, it is probably one of the better ones, but it’s not perfect. A corollary question is “Should we be measuring the HDL cholesterol levels?” Is there some other fraction or component of HDL that would perform better as a biomarker?
DR. WONG: The recent Multi-Ethnic Study of Atherosclerosis (MESA) paper18 was quite interesting because it showed that looking at HDL particles was perhaps more predictive of future coronary event risk than HDL cholesterol concentration; so, these particles might be a more appropriate measure than HDL concentration. In that study, the authors found that HDL and HDL particle concentrations were associated with carotid intima-media thickness, and future coronary heart disease event risk was not affected when they adjusted for atherogenic lipoproteins as well as HDL and HDL particle size. However, the relation of HDL cholesterol concentration with coronary heart disease risk was attenuated if they adjusted for the other measures, which suggests that its relationship to risk might be best explained by HDL particles.
DR. RADER: I do think the HDL particle data are interesting, but I’m not sure if that particular study using surrogate measures of atherosclerosis is necessarily the definitive work. I think we need concrete data on HDL cholesterol versus HDL particles for hard cardiovascular endpoints. I’ll just add that apoA-I, the major HDL protein, has been of great interest as a potential—and perhaps improved—alternative to HDL cholesterol. Some available data suggest that apoA-I may be slightly better, but actually, it’s not much better when you look at the abundance of existing data. Further, there is the concept that some measure of HDL function, such as efflux capacity, might ultimately be a better predictor of risk. However, we have very little data on that particular topic at this point.
DR. LAROSA: Let’s segue into the issue of clinical trials that have been either completed or stopped and what they might reveal about increasing HDL levels as a beneficial therapeutic move. We could discuss the Investigation of Lipid Level Management to Understand its Impact in Atherosclerotic Events (ILLUMINATE) trial,19 the dalcetrapib study,20 and AIM-HIGH,3 all of which failed to answer the question of whether increasing the HDL level would be effective.
DR. GOTTO: In the ILLUMINATE trial, if you look at the changes in lipoproteins, you would think that there would have been some cardiovascular benefit with the level of HDL increase and LDL decrease, so it’s hard to blame one or the other lipoprotein fraction. I think it remains to be determined whether the phenomenon is completely explained on the basis of the blood pressure change and the increased aldosterone secretion. There may be some other toxicity related to the vessel wall if there was some off-target drug action. We don’t have a complete report of the dalcetrapib study yet,20 but we know that HDL levels increased and the LDL levels did not change, but there was an increase in the cardiovascular end points. Based on what we know at this point, I don’t think we can determine whether the mechanism is wrong or the drug wasn’t strong enough.
There’s a great deal of speculation about the AIM-HIGH study.3 The investigators still claim that they should have seen some benefit from the changes and the predictions, although there was only a 4 mg/dL difference in the HDL level between the 2 comparison groups. They also introduced a lower dose of nicotinic acid in the comparison group and used ezetimibe to equalize the LDL levels. So, the larger Heart Protection Study 2-Treatment of HDL to Reduce the Incidence of Vascular Events (HPS2-THRIVE study)4 should clarify the benefit of niacin therapy.
DR. WONG: An important point about AIM-HIGH3 is that this was a secondary prevention trial, and many subjects were on several other recommended therapies for coronary diseases. Additionally, the subjects were very well treated to achieve an LDL level of <70 mg/dL. So, I think the failure of AIM-HIGH3 suggests that, perhaps, niacin therapy may not work in such a well-controlled population, ie, those who are well treated on the basis of LDL level. We’ll have to see whether a larger study like HPS2-THRIVE4 is going to show a different result, but I think this does not disprove the speculation that an increase in the HDL level might work in other populations—perhaps, populations with metabolic syndrome or primary prevention—that do not necessarily have well-controlled LDL levels. There are still many open questions resulting from the AIM-HIGH trial.3
DR. GOTTO: There is another aspect about HDL trials worth mentioning. In the statin trials, individuals who have benefited the most from statin intervention—whether it’s on the basis of quantitative coronary angiography or clinical endpoints—are those who start off with the lowest HDL levels. So, there could be a difference based on what HDL level you begin with.
DR. RADER: The outcome of the HPS2-THRIVE trial,4 which has not yet been reported at the time of this recording, will shed some light on the niacin issue. With regard to HDL as a marker of therapeutic benefit, I think the biggest challenge to that hypothesis is the dal-OUTCOMES trial with dalcetrapib.20 Here, we have a drug that increased the HDL level up to approximately 30% and its administration was stopped for futility because there was no reduction in cardiovascular events. Thus, unless there was some sort of off-target adverse effect of the drug that counterbalanced the benefits of increased HDL levels, it’s very hard to continue to support the simplistic hypothesis that increasing the HDL cholesterol levels reduces the cardiovascular risk. This trial clearly suggests that the situation is more complicated than that.
DR. LAROSA: Perhaps, one of the problems we’re having with the CETP inhibitors is that we’re looking at an HDL molecule that’s already chock-full of cholesterol and may not be functional in the way we would like it to be.
DR. RADER: It’s very possible that an increase in the HDL level by CETP inhibition may not necessarily be very protective by nature because the mechanism by which it is being increased is not desirable; however, there could be other mechanisms that increased HDL cholesterol levels through different approaches that might still be beneficial. I’m not suggesting that an increase in HDL levels is never going to be beneficial under any circumstances, but it certainly poses a challenge to the generic hypothesis.
DR. WONG: Yes, I think we really need to see how CETP inhibitors affect the structure and function of various HDL sub-fractions, and not much data are available on that. There’s still hope for some of the other CETP inhibitors, namely, evacetrapib and anacetrapib, which increase the HDL levels above 120%—a much larger increase than you get with dalcetrapib. However, I think we’ll need to await the endpoint trials to determine whether this really has an advantage over statin therapy and therapy involving lowering of the LDL cholesterol levels, specifically to understand how HDL structure and function is affected by these drugs.
DR. GOTTO: I’d be interested in Dr. Rader’s comments on this because he studied HDL particles that resulted from CETP inhibition with torcetrapib and found that they were fully active in promoting cholesterol efflux in cells.
DR. RADER: Yes, it’s a complicated area because torcetrapib may have had off-target effects that counterbalance the beneficial effects of efflux. However, it’s also possible that there are different ways of measuring efflux, and certain pathways may be beneficial, while others may not. We just don’t have the data yet to know how to fully interpret the different types of efflux assays for different types of HDL particles. We also don’t know whether those pathways are themselves atheroprotective.
DR. LAROSA: That’s a good segue into the next topic: mechanistically, how does having a high HDL level benefit us? The mechanism that we have focused on the most is the ability of HDL to promote reverse cholesterol transport, but HDL has other properties that might be beneficial, including its antiinflammatory and antioxidative properties. In fact, some investigators believe that these properties rather than reverse cholesterol transport are responsible for HDL’s alleged benefits.
DR. GOTTO: Regarding the antiinflammatory effects, HDL is associated with paraoxonases, which have strong antioxidant properties, and it inhibits the induction and formation of vascular cell adhesion molecule (VCAM), intercellular adhesion molecule (ICAM), and E-selectin. This may be due to the inhibition of sphingosine kinase, which is required for the action of tumor necrosis factor-alpha on ICAM and VCAM. HDL also inhibits LDL oxidation and monocyte chemotactic protein-1. In fact, these are just a few of the numerous actions of HDL; it also binds endotoxin/lipopolysaccharides. Therefore, HDL has a number of antiinflammatory and antioxidative effects that can be demonstrated by in vitro experiments.
DR. RADER: There are several interesting properties of HDL that are potentially antiatherogenic. It certainly can be argued that they would be. What we don’t know yet is which of these, if any, are operative in humans, and what is the difference among different HDL particles and different individuals in these different assays. One of the problems is that these are difficult assays that are not standardized by any means and haven’t been examined in large epidemiologic studies with regard to their relationship to cardiovascular risk. This is clearly the direction for future research, but it’s a big challenge for the field to amass that degree of high-quality data in order to actually make sense of these different properties and their roles in HDL atheroprotection.
DR. LAROSA: Is there currently an assay available for reverse cholesterol transport or cholesterol efflux that might be clinically useful?
DR. RADER: If you mean one that can be ordered by a clinician in the office, then the answer is no. I think these are still research tools that aren’t amenable for use in the clinic at this point. Having said that, I believe that it is feasible to try to develop an assay for efflux capacity, or even one or more of these other functions, possibly with a cell-based assay. For example, Dr. Gotto mentioned that sphingosine-1-phosphateon HDL was a very interesting marker, which over time, if validated, could potentially be used in clinical practice. However, this validation is going to take a while, several years maybe.
DR. LAROSA: Are any of these assays being incorporated into ongoing trials now?
DR. RADER: I am aware of interest in measuring efflux capacity as a component of a clinical trial. I suspect that some of these other effects (eg, antiinflammatory effects) are possibly of interest as well.
DR. GOTTO: I think that some of these effects were measured in the dalcetrapib trial,20 but they were ancillary studies. I haven’t come across any reports about other trials.
DR. WONG: Do you know whether HPS2-THRIVE4 or any of the other CETP trials are measuring them?
DR. GOTTO: No, I don’t.
DR. RADER: Yes, but I don’t know what’s in the public domain at this point.
DR. WONG: We do have some emerging epidemiologic data indicating that the protective function of HDL may be attenuated in the presence of an inflammatory state such as that indicated by high C-reactive protein levels and potentially by other inflammatory factors as well.
DR. LAROSA: So, what we can say at this point is that we don’t know to what extent it could be protective. We have evidence of various properties of HDL that might be theoretically useful for describing its potential benefit, but we don’t have either a clinical trial or other prospective evidence indicating that any of these properties are explanatory. Is that a fair statement?
DR. RADER: I think it’s a fair statement.
DR. WONG: Yes, I agree.
DR. LAROSA: Okay, let’s talk a little bit about new studies that are currently in the field or new drugs that are currently on the horizon that might help us clarify some of these issues.
DR. RADER: We’ve already talked at length about the CETP inhibitors. I think this is a very interesting class, and I won’t add much more except that we are awaiting results from the ones that are still in clinical trials. I personally am moving more toward the concept of the HDL flux hypothesis (versus the HDL cholesterol hypothesis), ie, promoting flux might be a better mechanism for reducing risk compared with just increasing HDL cholesterol levels. In that regard, there are interesting interventions that are in the relatively early stages of development that would potentially promote flux, including infusing reconstituted apoA-I through a variety of different approaches, increasing apoA-I production with a small molecule, increasing macrophage transporters of efflux such as an LXR agonist, or an anti-microRNA-33 (miR-33) approach. These approaches need to be tested, of course, but good pre-clinical data are available that suggest they may be antiatherogenic.
DR. WONG: Some of the newer upcoming trials include HPS2-THRIVE4 as well as the CETP7 outcomes.
DR. LAROSA: I think it would be useful if we defined that study just briefly, because it’s been mentioned several times and some of our readers might not know what it is.
DR. WONG: HPS2-THRIVE4 is a large niacin trial examining more than 25 000 subjects. The trial has been essentially completed now, and they’re basically analyzing the data. Patients with known coronary disease were randomized between the newer niacin product developed by Merck and placebo on top of statin therapy. They actually did report interim data at the European Society of Cardiology Congress in 2012 that showed a 20% reduction in LDL levels and a 17% increase in HDL levels after 8 weeks, and 76% of the treated subjects remained compliant. So, this trial is going to have many more endpoints than the AIM-HIGH3 did.
Because AIM-HIGH3 had virtually no hazard ratio, the question is whether this larger sample size with HPS2-THRIVE4 is going to produce any different results. It’s going to be important to try to determine how to separate the effects of the increase in HDL levels from the additional reductions in LDL levels that you’re going to get from niacin or even from the other CETP inhibitors in those trials.
DR. LAROSA: The strategy of AIM-HIGH3 was to avoid those issues by essentially titrating LDL so that they wouldn’t interfere.
DR. GOTTO: One of the interesting aspects about this HPS2 study4 is that approximately a quarter of the participants are from China, and I think we have limited data about the effects of niacin in Chinese patients. Also, the preliminary data they released showed that during the screening leading up to the randomization, about one-third of the 25 000 participants dropped out mainly due to flushing. So, despite the antiflushing component of the drug, there’s still a problem with flushing.
DR. LAROSA: We shall see about that in the results soon, but I don’t think sample size is an issue here. I agree that the inclusion of a group who may be genetically different in their response to niacin is a potential issue.
DR. LAROSA: Nobody’s mentioned VA-HIT;13 is that for a good reason?
DR. GOTTO: I mentioned that the trial showed an association between increasing HDL levels and reduced cardiovascular risk. There was no association/correlation in this trial between triglyceride reduction and LDL concentration. LDL cholesterol levels remained the same, triglyceride levels decreased by 31%, and HDL levels increased by 6%, and according to the investigators, the 22% decrease in risk could be explained mainly in individuals who had metabolic syndrome or insulin resistance. So, that’s where there was most benefit. There was also a correlation between event reduction and HDL increase, and the individuals with clinical events in this trial had lower concentrations of the large HDL particles and higher concentrations of the small, pre-beta HDL particles.
DR. LAROSA: What do we think that doctors should be doing with HDL measurements at this point?
DR. RADER: We should still measure HDL levels in clinical practice, and such measurement should be part of a full lipid profile. It has some advantages and value in terms of predictive ability, as we’ve discussed, and it helps identify people who might be at high risk. I think the issue is about what we are supposed to do with a low HDL level when you identify it. My personal viewpoint is that we need to view that patient as presumably having a higher risk, and treat him/her slightly more aggressively than we might have otherwise, including a very aggressive reduction in the LDL level and, of course, active lifestyle management. I don’t think we’re currently in a position to actively recommend pharmacotherapy to increase HDL levels in patients with low HDL levels; however, the results of the HPS2-THRIVE4 trial may change that.
DR. WONG: I completely agree, and I think a low HDL level should certainly be an impetus to encourage the patient to note whatever lifestyle issues might be responsible for that. Cigarette smoking can reduce HDL levels by up to 20%, and appropriate weight control is not emphasized enough. So, this would be another reason to help emphasize weight loss, particularly in someone with visceral adiposity, which could certainly contribute to low HDL levels. Aerobic exercise, even brisk walking, can help. We need to encourage our patients to improve their lifestyle. This is, of course, an issue that is often not adequately dealt with, particularly in the primary care setting.
DR. LAROSA: Are there any other final comments?
DR. GOTTO: We know that patients with low HDL levels who are at high or average risk benefit remarkably from statin treatment. We carried out a study with lovastatin with individuals who had HDL levels below 50 mg/dL and who had no evidence of cardiovascular disease. At that time, they were considered to have average rather than elevated cholesterol and LDL levels, although we’d consider those levels too high now. They benefited remarkably from statin treatment within 5 years even though they had no signs or evidence of cardiovascular disease, and we put them on an intensive program of diet, exercise, and lifestyle modification.21
DR. LAROSA: I’d like to thank Dr. Wong, Dr. Gotto, and Dr. Rader for a very stimulating discussion, and I will now declare these proceedings closed.
*Since this interview was conducted, the full results of HPS2-Thrive were presented at the American College of Cardiology Scientific Sessions in March 2013. The trial was prematurely discontinued due to lack of efficacy as well as futility in reaching the primary endpoint of coronary death, nonfatal myocardial infarction, stroke, or coronary revascularization that occurred in 15.0% of patients in the control arm and 12.5% of patients in the niacin/laropiprant arm, and the difference between these 2 values was not significant. Moreover, there was an excess of adverse events noted in the niacin/laropiprant arm (approximately 30 adverse events per 1000 treated patients), including a 3.7% absolute excess risk of diabetic complications and 1.8% excess risk of new-onset diabetes.5 In addition, there was a greater risk of myopathy, particularly among the Chinese patients enrolled in the study.6 The investigators concluded that in light of these findings, the role of extended-release niacin for the prevention of cardiovascular disease needs to be reconsidered.
FoxP2 Media LLC is the publisher of The Medical Roundtable.