HIV Articles  
Lipids in 2012: HDL cholesterol studies-more of the same?
  Download the PDF here
Nature Reviews Cardiology (February 2013)

Jean-Pierre Despres

Jean-Pierre Despres, PhD, FAHA, FIAS, received his PhD degree in exercise physiology from Universite Laval, Quebec City, QC, Canada in 1984. After completing a postdoctoral fellowship at the Department of Medicine of the University of Toronto, ON, he returned to Universite Laval in 1986 to establish an independent research program. Dr Despres has published more than 590 papers in the areas of human adipose tissue metabolism, body fat distribution, lipid and lipoprotein metabolism, type 2 diabetes mellitus, exercise physiology, and nutrition using approaches ranging from basic research to clinical and epidemiological studies.


Studies published in 2012 in the field of HDL research have provided further evidence suggesting that a low HDL-cholesterol level, in the absence of related lipid or nonlipid risk factors, is not associated with increased risk of coronary heart disease.

Analyses of data from prospective observational cohorts published over the past 30 years have shown that increased levels of circulating HDL cholesterol are associated with a reduced risk of coronary heart disease events.1 Furthermore, numerous preclinical studies have revealed that HDL particles have properties that could provide protection against the development of atherosclerosis.2 On the basis of this evidence, raising the level of HDL cholesterol has generally been considered a legitimate strategy to reduce cardiovascular risk through reduced progression, or even regression, of atherosclerotic plaque. Considerable efforts have been devoted to the development of HDL-cholesterol raising therapies in the hope that they will add to the well-documented benefits of statin-mediated lowering of LDL cholesterol. In this article, four important papers published in 2012 that have reinforced our understanding about the role of HDL particles and HDL cholesterol in cardiovascular risk will be highlighted.

One pharmacological approach that produces considerable increases in the circulating level of HDL cholesterol is inhibition of cholesteryl ester transfer protein (CETP), an enzyme that promotes net mass transfer of cholesteryl esters from HDL particles to apolipoprotein B-containing lipoproteins.2, 3 Despite promising preclinical results, torcetrapib-the first CETP inhibitor tested in clinical trials-was not found to reduce the size of atherosclerotic plaques in imaging trials.4 Furthermore, torcetrapib was reported to increase mortality and cardiovascular events in the large trial ILLUMINATE.5 This phenomenon could be related, among other factors, to an increase in blood pressure and aldosterone levels. Nevertheless, post-hoc analyses of results from torcetrapib trials combined with preclinical data suggested that the off-target effects of torcetrapib were probably unrelated to CETP inhibition.3 Therefore, the hypothesis that CETP inhibition could reduce, or even reverse, atherosclerotic vascular disease remained to be tested.

Another CETP inhibitor, dalcetrapib, was assessed in the DAL-OUTCOME study and the results published in November 2012.6 Importantly, this compound was found to have little effect on the level of LDL cholesterol while increasing HDL-cholesterol concentration.3 These findings provided the appropriate experimental conditions to test the clinical benefit on cardiovascular outcomes of specifically raising the HDL-cholesterol level with dalcetrapib. In the DAL-OUTCOME study,6 15,871 patients who had experienced a recent (median 61 days) acute coronary syndrome were randomly assigned to receive dalcetrapib (600 mg per day) or placebo on top of the best evidence-based clinical management. During the trial, HDL-cholesterol levels increased from baseline by 4-11% in the placebo group and by 31-40% in the dalcetrapib group. As expected, the CETP inhibitor had little effect on LDL-cholesterol levels. At the interim analysis (median follow-up 31 months), the data and safety monitoring board recommended termination of the trial for futility. The cumulative event rate was essentially the same in the placebo and dalcetrapib groups (8.0% vs 8.3%, respectively), and the drug had no effect on any component of the primary end point (a composite of major adverse cardiovascular events) or on mortality. Furthermore, modest but significant (P <0.001) increases in systolic blood pressure (0.6 mmHg) and C-reactive protein level (0.2 mg/l) were noted with dalcetrapib when compared with placebo.6 Thus, this new trial did not provide evidence that selectively increasing HDL-cholesterol levels (at least through CETP inhibition) reduces the risk of recurrent cardiovascular events.

Another approach fuelling the HDL-cholesterol target debate relies on Mendelian randomization analyses. If a direct relationship between HDL cholesterol and cardiovascular outcomes exists, genetic variants associated with altered HDL-cholesterol levels, but that have no influence on other lipid or nonlipid risk factors, should be related to cardiovascular disease. Voight et al. examined a single nucleotide polymorphism in the endothelial lipase gene (LIPG Asn396Ser) in 20 studies that included a total of 20,913 patients with myocardial infarction (MI) and 95,407 healthy control individuals.7 Their findings were published in May 2012. Carriers of the LIPG 396Ser allele (2.6% frequency) had higher HDL-cholesterol levels (+0.14 mmol/l) than noncarriers, but showed no difference in levels of other lipid or nonlipid risk factors for MI. This finding enabled the investigators to test the hypothesis that lifetime exposure to a high HDL-cholesterol level might be cardioprotective. However, the increased HDL-cholesterol level linked with the 396Ser allele was not associated with a reduced risk of MI. This finding differed markedly from the 13% risk reduction (OR 0.87, 95% CI 0.84-0.91) that the investigators predicted on the basis of observational epidemiology studies.7 In addition, Voight and colleagues calculated a genetic risk score, which included 14 single nucleotide polymorphisms exclusively associated with increased HDL-cholesterol level and, again, found no relationship between these alleles and cardiovascular events.7 These results indicate that genetic variation associated with lifetime increased HDL-cholesterol level, but not with other cardiovascular risk factors, does not seem to be associated with a reduced risk of MI.

The results of this interesting analysis do not, however, provide the final answer as to whether or not HDL should be targeted to optimize protection against cardiovascular events. At this stage, one point can be made with certainty-not all approaches that increase the HDL-cholesterol level will translate into clinical benefits. For example, CETP inhibitors that do not also reduce LDL-cholesterol levels are probably useless. Whether more-potent CETP inhibitors, such as anacetrapib and evacetrapib that markedly increase HDL-cholesterol levels and decrease levels of non-HDL cholesterol, reduce atherosclerosis and related cardiovascular events is currently being tested in large trials.

In this context, we need to keep in mind that HDL cholesterol is only one feature of HDL. These lipoprotein particles have numerous properties that contribute to slow the development of atherosclerosis, but are not captured by the measurement of HDL-cholesterol concentration. HDL cholesterol only reflects the cholesterol content of the HDL fraction that is most often isolated by precipitation techniques. Numerous techniques are now available to assess the various other properties of HDL, including composition, size, migration on 2D gels, metabolomics, and the capacity to promote cholesterol efflux and act as antioxidants (Figure 1a). Thus, an expanded panel of HDL 'metrics' should be studied to further our understanding of the features of HDL that could be targeted for cardioprotection.

Analyses of MESA8 and the MRC/BHF Heart Protection Study,9 both published in 2012, suggested that the association between HDL-cholesterol level and cardiovascular outcomes was largely attenuated after control for other features of the lipoprotein profile. On the other hand, HDL particle number assessed by NMR was independently associated with cardiovascular outcomes. These findings suggest that some properties of HDL might provide information beyond HDL cholesterol itself, which seems to be largely a marker of other lipid and nonlipid risk factors. We have previously suggested that low HDL-cholesterol level is usually accompanied by abdominal obesity, lack of physical activity and sedentary lifestyle, insulin resistance, elevated triglyceride and apolipoprotein B levels, and by a state of low chronic inflammation (Figure 1b), key correlates that are predictive of increased cardiovascular disease risk.10 Therefore, low HDL-cholesterol level in isolation is a rare disorder. Such key correlates of low HDL-cholesterol level might explain why findings from observational epidemiology studies are discordant with results from clinical trials and Mendelian randomization studies. From population health and clinical stand points, although clearly associated with a low risk of coronary heart disease, a high HDL-cholesterol level is largely the consequence of healthy lifestyle habits, a low level of visceral adipose tissue and ectopic fat, and high insulin sensitivity.

Although studies on HDL published in 2012 have generated 'more of the same', such literature adds to the debate. Additional preclinical and clinical data will be required to answer the key question-is HDL cholesterol only an HDL-related marker of cardiovascular risk, or could other features of HDL be targeted as well? Meanwhile, to generate high physiological levels of functional HDL particles, we should watch our waistlines and stay physically active.

Key Advances

· Selectively increasing the HDL-cholesterol level with a cholesteryl ester transfer protein inhibitor (dalcetrapib) that did not decrease the LDL-cholesterol level did not reduce the risk of recurrent cardiovascular events6

· Genetic variants that are associated with altered HDL-cholesterol levels, but that have no additional influence on other lipid or nonlipid risk factors, were found to be unrelated to cardiovascular events7

· HDL particle number, assessed by NMR, was consistently found in two studies to be one feature of HDL that predicts cardiovascular events8, 9

  iconpaperstack view older Articles   Back to Top