High-Density Lipoprotein as a Therapeutic Target:  A Systematic Review FREE

Despite numerous therapeutic advances, cardiovascular disease (CVD) remains the leading cause of morbidity and mortality in developed countries. Although statins reduce coronary artery disease by approximately 30%,^1- 2 substantial residual cardiovascular risk remains, even with very aggressive reductions in levels of low-density lipoprotein cholesterol (LDL-C).^1,3 Accordingly, attention has shifted toward strategies for targeting high-density lipoprotein (HDL) composition as adjunctive therapy to prevent and treat CVD.

In this article, we review published randomized clinical trials that relate to the effect of high-density lipoprotein cholesterol (HDL-C) levels on clinical outcomes and atherosclerosis imaging. Additionally, we discuss epidemiologic observations, proof-of-concept and mechanistic basic science studies, and key phase 1 through 3 trials of novel agents. We also address current guidelines and promising HDL-based strategies and provide clinical suggestions.

Assisted by a medical librarian, we performed a systematic search of the English-language literature published from 1965 through May 2007 in MEDLINE and the Cochrane database using the key words HDL-C and apolipoprotein A-I and the subheadings reverse cholesterol transport, CVD prevention and control, drug therapy, and therapy. This search, plus the inclusion of 1 report published subsequent to the May 2007 cutoff,⁴ yielded 754 studies, of which 357 were randomized controlled trials (Figure 1). We reviewed the abstracts of all of these studies, and in many cases, we analyzed the full-text reports. We extended the search by review of bibliographies from pertinent original reports of data and review articles. Two authors (I.M.S. and M.H.S.) independently assessed each study’s eligibility. When this determination was not unanimous, both authors reviewed the full manuscript and reached consensus with the third author (B.J.A).

We prioritized our selection to identify randomized controlled trials over meta-analyses over mechanistic studies. The inclusion criteria were (1) controlled clinical trials of increasing HDL-C levels or otherwise altering the composition of HDL through pharmacological therapy, (2) randomized treatment allocation, and (3) clinical or surrogate (atherosclerotic burden) outcome measures. A total of 31 core studies met our criteria (Figure 1).

Our review also includes novel mechanistic or basic science studies providing proof of concept, as well as key preclinical trials of novel agents. Furthermore, we reviewed observational and meta-analysis reports from 1965 to 2007 and abstracts presented at the American Heart Association and American College of Cardiology annual sessions from 2003-2007. We also reviewed position statements and guidelines from the American College of Cardiology, American Heart Association, National Cholesterol Education Program, European Consensus Panel, and American Diabetes Association, as well as ongoing drug trials via ClinicalTrials.gov.

Studies indicate that low HDL-C levels are relatively common in the general population, with reported rates of HDL-C less than 35 mg/dL (to convert to mmol/L, multiply by 0.0259) of 16% to 18% in men and 3% to 6% in women.⁵ In addition, low level of HDL-C is a component of the metabolic syndrome, which has a prevalence of 24% in US individuals older than 20 years.^6- 7

Multiple epidemiologic studies have established a low level of HDL-C as an independent risk factor for CVD.^5,8 For example, in the Framingham Heart Study, 43% to 44% of coronary events occurred in persons with HDL-C levels less than 40 mg/dL (22% of the total study population).⁵ Individuals having HDL-C levels less than 35 mg/dL had an 8-fold higher incidence of CVD compared with those having HDL-C levels of more than 65 mg/dL.^5,8 The strength of the relationship between low HDL-C levels and increased CVD risk also is significant in elderly individuals and may be greater in women than in men.^5,8- 9

Angiographic and ultrasonographic data indicate that low levels of HDL-C are associated with risk and severity of coronary artery disease, carotid disease, and postangioplasty restenosis.^9- 10 Observational studies have shown that each 1-mg/dL decrease in plasma HDL-C concentration is associated with a 2% to 3% increased risk of CVD.^9- 10 In addition, each 1-mg/dL increase is associated with a 6% lower risk of coronary death, independent of LDL-C level.^9- 10

Although HDL is thought to protect against CVD, the precise means by which it exerts its antiatherogenic effects are still being characterized. It appears that HDL is likely protective through multiple pathways, including both reverse cholesterol transport and non–cholesterol-dependent mechanisms.¹¹

Reverse Cholesterol Transport. Reverse cholesterol transport involves the transfer of excess cholesterol from lipid-laden macrophages (foam cells) present in peripheral tissues to the liver via HDL, with subsequent catabolism of cholesterol or excretion into bile (Figure 2).¹² In the vessel wall, cholesterol ester stored in macrophages can be converted to free cholesterol by cholesterol ester hydrolase, whereas acyl-cholesterol acyltransferase can esterify cholesterol within macrophages to form atherogenic foam cells.

The liver and intestine synthesize lipid-poor apolipoprotein A-I (apo A-I), which can interact with the adenosine triphosphate–binding cassette transporter A1 (ABCA1) located on the arterial macrophages, transporting free cholesterol to the extracellular lipid-poor HDL. Lipidation of the HDL particles generates nascent (pre-β) HDL.¹³ Subsequently, lecithin-cholesterol acyltransferase esterifies free cholesterol within nascent HDL to produce mature α-HDL particles (ie, HDL₃ [smaller, more dense particles] and HDL₂ [larger, less dense particles]). These mature HDL particles can further take up free cholesterol via the macrophage adenosine triphosphate–binding cassette transporter G1.

Mature HDL has at least 2 metabolic fates. In the direct pathway, cholesteryl esters contained within HDL can undergo selective uptake by hepatocytes and steroid hormone–producing cells via the scavenger receptor type B1 and subsequent excretion into the bile.^12,14 In the indirect pathway, cholesteryl esters within HDL can be exchanged for triglycerides in apolipoprotein B–rich particles (LDL and very low-density lipoprotein [VLDL]) through the action of cholesteryl ester transfer protein (CETP). The subsequent uptake of apolipoprotein B rich in cholesteryl esters by hepatic LDL receptors may be responsible for up to 50% of reverse cholesterol transport.¹² The triglyceride-rich HDL can then undergo hydrolysis by hepatic lipase and endothelial lipase to form small HDL for further participation in transport.¹²

The kidneys appear to play a less understood role in HDL catabolism by controlling the processing rate of lipid-poor apo A-I.¹⁴

Non–Cholesterol-Dependent Mechanisms. In addition to its major role in reverse cholesterol transport, HDL has other biological activities that may contribute to its protective effects against atherosclerosis.^9,11,15 These include antioxidant (counteracting LDL oxidation) effects, anti-inflammatory effects, antithrombotic/profibrinolytic (reducing platelet aggregation and coagulation) effects, and vasoprotective (facilitating vascular relaxation and inhibiting leukocyte chemotaxis and adhesion) effects.^9,11,15- 16

Collectively, HDL and its components (including apo A-I, paraoxonase, platelet activating factor acetylhydrolase, and other antioxidant enzymes) exert an array of effects that may help prevent atherosclerosis, acute coronary syndromes, and restenosis after coronary angioplasty.^9,17

High-density lipoprotein particles can vary substantially in size, density, composition, and functional properties, potentially affecting their relationship to atherosclerosis.^{11,14- 15,18- 19} Furthermore, levels of plasma HDL-C do not predict its functionality.

Acute or chronic inflammation can adversely influence the normally protective characteristics of HDL. For example, modification of apo A-I within HDL by leukocyte myeloperoxidase renders HDL dysfunctional with proinflammatory and atherogenic properties.²⁰ Patients with CVD or CVD equivalents have elevated levels of proinflammatory HDL compared with healthy controls.²¹ Furthermore, the inflammatory and anti-inflammatory properties of HDL are better predictors of CVD prevalence than HDL-C alone.²¹ In the Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial (VA-HIT) population, the large, cholesterol-rich HDL particle, α-1, was the most significant risk factor for recurrent CVD events among HDL particles.²² Rader²³ showed that 75% of patients with coronary heart disease (CHD) or CHD equivalents had proinflammatory HDL despite statin therapy. Despite the availability of several techniques to separate HDL by density gradient ultracentrifugation, proton nuclear magnetic resonance, and 2-dimensional gel electrophoresis, studies have not compared the ability of these techniques to predict vascular events. Moreover, mechanistic studies differ as to whether particle size is an important predictor of HDL function; some analyses suggest that small particles are anti-inflammatory, while others indicate that small particles may not be protective in settings of increased oxidative stress.^16,19- 21

The apolipoproteins within HDL are significant determinants of its function.¹⁵ The major HDL apolipoprotein, apo A-I, helps to stimulate the activity of ABCA1 and lecithin-cholesterol acyltransferase and is a ligand for scavenger receptor type B1.¹⁴ On the other hand, apo A-II, another component of HDL, has been shown to be proatherogenic in animal models.¹⁴ Thus, therapeutic strategies that selectively increase apo A-I levels may be more atheroprotective than those that increase levels of both apo A-I and apo A-II.

Aerobic Exercise. Frequent aerobic exercise has been shown to increase HDL-C levels by approximately 5% as early as 2 months from start of regular exercise in sedentary but otherwise healthy individuals through multiple mechanisms (Table 1).^24- 25 To increase HDL-C levels optimally, individuals should perform five 30-minute sessions of brisk aerobic exercise per week, with total duration of exercise more than 120 minutes per week being the strongest determinant of increased HDL-C levels.^42- 43

Tobacco Cessation. Tobacco, both smoked and smokeless, reduces levels of HDL-C (Table 1).²⁶ One study showed that HDL-C levels increase by approximately 4 mg/dL following smoking cessation without significant changes in levels of LDL-C, total cholesterol, or triglycerides.²⁷ Tobacco smoke also is a source of oxidative stress that can lead to HDL dysfunction.¹⁹ Tobacco cessation should be aggressively promoted via a multidisciplinary approach of counseling and pharmacological agents, as appropriate.⁴²

Weight Loss. Weight loss generally increases HDL-C levels (Table 1) in overweight or obese patients.²⁸ During active weight loss, HDL-C levels decrease slightly; however, when a stable weight reduction is achieved, HDL-C levels increase by 0.35 mg/dL per kilogram of weight lost.²⁹ Weight loss, achieved by lifestyle modifications with or without pharmacological support, has been associated with improved cardiometabolic risk factors.^30- 32 Overweight and obese individuals should aim to achieve a body mass index (calculated as weight in kilograms divided by height in meters squared) of less than 25 (less than 24 if of Asian descent), at a rate of 2 kg of weight lost per month.⁴²

Alcohol Consumption. Intake of moderate amounts of alcohol (30-40 g [1-3 drinks] per day) increases HDL-C levels and is associated with decreased risk of CHD independent of other factors.^33- 34 The link between moderate alcohol consumption and decreased cardiovascular risk may reflect the epiphenomenon of a “sick quitter” effect, whereby former alcohol users practice abstinence either due to illness or to medical advice but are misclassified as being nondrinkers.³⁵ Ingestion of 30 to 40 g of alcohol per day for 3 weeks can increase HDL-C levels by as much as 12%, irrespective of the type of alcohol consumed (Table 1).³⁴ However, current guidelines advise no more than 2 drinks per day for men and no more than 1 drink per day for women.⁴² Persons who do not drink should not be encouraged to initiate regular alcohol consumption.

Dietary Factors. Diets rich in saturated fatty acids and trans-fatty acids can increase HDL-C levels but also increase LDL-C levels and HDL-induced expression of proinflammatory endothelial cell adhesion molecules.³⁶ In contrast, diets rich in polyunsaturated fats improve the anti-inflammatory capacity of HDL.³⁶ Substituting dietary saturated fatty acids and trans-fatty acids with monounsaturated fatty acids and polyunsaturated fatty acids reduces the LDL-C:HDL-C ratio.³⁷

Ingestion of n-3 polyunsaturated fatty acids has been associated with increased HDL-C levels and cardiovascular benefits³⁸; however, studies of the cardiovascular effects of n-6 polyunsaturated fatty acids are lacking.³⁹

Replacing saturated fatty acids and trans-fatty acids with low–glycemic index carbohydrates may improve HDL-C profile, but the data are limited.⁴⁰ Recent data suggest that a low-fat, high-fiber diet for 2 to 3 weeks, in combination with exercise, converts HDL from a proinflammatory to an anti-inflammatory state.²⁵ This improvement in the functional property of HDL occurred despite a mild reduction in levels of HDL-C, suggesting increased turnover of proinflammatory HDL.²⁵

Patients should be advised to replace dietary saturated fatty acids and trans-fatty acids with monounsaturated and polyunsaturated fatty acids sources such as plant oils (olive, canola, soy, mustard, flaxseed), nuts (almonds, peanuts, walnuts, pecans), and marine foods (salmon, tuna, mackerel, marine oils).⁴²

In general, studies using overall lifestyle modifications have shown CVD benefit; however, these beneficial effects cannot solely be attributed to a single strategy or mechanism.^30,41

Currently available HDL-based therapeutic approaches can be grouped into 2 major categories: agents that increase or modify levels of the HDL components (HDL-C, apo A-I, and phospholipids), and agents that up-regulate reverse cholesterol transport and macrophage cholesterol efflux.

Agents That Modify HDL Composition. Nicotinic Acid. First recognized as a treatment for dyslipidemia in 1955, nicotinic acid can increase HDL-C levels by 20% to 30% as well as reduce plasma triglyceride levels by 40% to 50% and LDL-C levels by up to 20% (Table 2).^44- 56 When tolerated, nicotinic acid represents the most effective HDL-C–increasing agent currently available.⁹ The most compelling data to support the use of niacin are from the Coronary Drug Project (CDP), which evaluated niacin monotherapy in 8341 men with prior myocardial infarction. In this prestatin study, niacin was associated with a 27% reduction in the incidence of nonfatal reinfarction at 6 years,⁴⁴ and all-cause mortality was reduced by 11% at 15 years (Table 3).⁵⁰

Other clinical outcomes studies of niacin have been conducted in the setting of combination therapy, and thus the results should be interpreted in this context. In general, treatment with niacin, both as monotherapy and in combination with other antidyslipidemic agents, has produced consistent clinical and atherosclerotic burden benefit (Table 3). Most of the CVD benefit of niacin can be attributed to its HDL-C–increasing properties (Table 2).⁹ The therapeutic potential of niacin has been limited by its adverse effects, which can be reduced by use of an extended-release formulation. Use of niacin by patients with diabetes has been cautioned against by the American Diabetes Association, although recent niacin studies have shown no significant increase in long-term glycemic levels, use of oral hyperglycemic regimens, or niacin withdrawal in these patients.^85- 86

Statins. Statins modestly increase HDL-C levels by 5% to 15%, and this effect is most evident with rosuvastatin.⁶⁸ The mechanism for statin-induced increase in HDL-C levels is presently unclear, but statins increase levels of apo A-I and lipid-poor HDL and also augment the activity of the antioxidant enzyme paraoxonase (Table 2).^69- 72 In a recent post hoc analysis of 1455 patients with low HDL-C levels, statins reduced coronary atheroma independent of LDL-C when HDL-C levels were increased by at least 7.5%.⁷³ Statins confer a greater reduction in CVD among patients with lower HDL-C levels, independent of LDL-C effects.⁷⁴ However, due to their modest effect on HDL-C levels, statins generally are not adequate as monotherapy for increasing HDL-C.⁷⁵

Agents That Target Reverse Cholesterol Transport and Macrophage Cholesterol Efflux. Fibric acid derivatives (fibrates), which are agonists of peroxisome proliferator–activated receptor (PPAR) α, can increase HDL-C levels (by 10%-20%), modestly lower LDL-C levels (by 10%-15%), and substantially lower levels of triglycerides (by 40%-50%) (Table 2).^57- 58 Several studies have elucidated the effect of fibrates on reverse cholesterol transport (Table 2).^87- 88

In the Helsinki Heart Study and the VA-HIT trial, treatment with gemfibrozil was associated with reduction in CVD events (Table 3).^57- 58 The 22% reduction in coronary events in the VA-HIT trial was attributed to a modest increase (6%) in HDL-C levels.⁵⁸ In contrast, the results of studies using other fibrates such as bezafibrate and fenofibrate have been negative, while clofibrate has been associated with potential hazard (Table 3).^{44,57- 67,80} Some evidence indicates that the benefits seen in the VA-HIT cohort may be partly due to additional effects of fibrates on apolipoprotein synthesis and lipoprotein metabolism (Table 2).^89- 90

The combination of fibrates, particularly gemfibrozil, with statins requires caution and monitoring of creatine kinase levels because of the risk for myotoxicity, including rhabdomyolysis.⁴² Because of this safety concern and variable-outcome study results, the precise role of fibrate treatment remains uncertain but likely includes some patients at high risk of coronary artery disease, low levels of HDL-C, and increased levels of serum triglycerides. The cardiovascular effect of adding fenofibrate to statin therapy is being assessed in a large-scale clinical trial in patients with type 2 diabetes.⁹¹

Several novel HDL-based therapeutic strategies have been or currently are being tested in phase 1 through 3 clinical trials. These include newer formulations of nicotinic acid/receptor agonists, CETP inhibitors, cannabinoid-1 receptor antagonists, PPAR agonists, liver X receptor/farnesoid X receptor agonists, and apo A-I and/or phospholipid-derived therapies (Table 4). As with current therapies, new HDL-based approaches presently evolving also can be categorized into agents that increase or modify levels of the HDL components and agents that up-regulate reverse cholesterol transport and macrophage cholesterol efflux.

Evolving Agents That Modify HDL Composition. Nicotinic Acid–Based Formulations. MK-0524, a selective prostaglandin D₂ receptor 1 antagonist, recently demonstrated reduction in nicotinic acid–associated vasodilation and flushing, both in mice and in humans (Table 4).⁹² The effect of MK-0524 and extended-release niacin on CVD is being assessed in 1 trial of patients with CHD and low HDL-C levels who are receiving statin therapy.¹¹⁰

CETP Inhibitors. Inhibition of CETP activity results in increased HDL-C levels and decreased levels of VLDL, LDL-C, and triglycerides (Table 4).⁹³ Inhibition of CETP also promotes the increase in levels of large, less dense HDL₂, an action that may lead to reverse cholesterol transport independent of pre-β- HDL and the ABCA1 pathway.⁹⁴ Epidemiologic evidence regarding the effects of CETP deficiency has been contradictory.^95- 97 However, the marked increases in HDL-C levels observed with several developmental compounds with significant CETP inhibitory activity nonetheless prompted the commercial development of these agents.

Torcetrapib, a direct CETP inhibitor, has been shown to decrease aortic atherosclerosis in animals and significantly increase HDL-C levels in humans.^111- 113 However, development of torcetrapib was halted in response to the results of the Investigation of Lipid Level Management to Understand Its Impact in Atherosclerotic Events (ILLUMINATE) trial, which showed a 61% increase in all-cause mortality and other cardiovascular events in individuals at high risk for CHD who were treated with torcetrapib plus atorvastatin compared with those who received placebo plus atorvastatin. Analysis from trials of torcetrapib revealed a mean blood pressure increase of 3 to 4 mm Hg associated with its use, a phenomenon that appears to be molecule-specific and unrelated to CETP inhibition.¹¹⁴ Three other trials with torcetrapib, which evaluated the effect of torcetrapib on coronary atheroma by intracoronary ultrasound and carotid intima-media thickness, showed no change in atherosclerotic burden despite a marked increase in levels of HDL-C (Table 3).^4,81- 82

While the exact mechanism for increased adverse events in the ILLUMINATE trial is not known, potential explanations include (1) increase in blood pressure and other off-target hemodynamic effects. In ILLUSTRATE there was an average 4.6-mm Hg increase in blood pressure in the torcetrapib group compared with the placebo group. This increase in blood pressure is a torcetrapib-specific toxicity, since such an effect has not been reported with other classes of CETP inhibitors or in animals or patients lacking CETP.⁹⁸ (2) Production of “dysfunctional” HDL, since CETP inhibition leads to production of α- HDL particles, which are not the preferred ABCA1-mediated cholesterol efflux transporters. (3) Production of “proinflammatory” HDL, possibly through oxidation of phospholipids in “stagnant” HDL particles that result from CETP inhibition.^16,20,99 (4) Unforeseen interactions between torcetrapib and atorvastatin. Future studies with other agents in this class that do not affect blood pressure may help clarify the mechanism(s) responsible for failure of torcetrapib.¹¹⁴

The CETP inhibitor JTT-705 reduces CETP activity by 95% in rabbits, accompanied by a 90% increase HDL-C levels, a 40% decrease in non–HDL-C levels, and an 80% decrease in aortic atherosclerosis (Table 4).¹⁰⁰ However, data from animal models on the effect of JTT-705 on atherosclerosis have not been consistent. A study in hypercholesteremic rabbits showed that even with similar changes in lipid profiles, JTT-705 did not affect aortic atherosclerosis.¹⁰¹ Potential limitations of these studies were the presence of high levels of total cholesterol and triglycerides, which might counter the beneficial effects of increasing levels of HDL-C.

In 2 recent randomized, double-blind, placebo-controlled phase 2 trials, JTT-705 alone or in combination with pravastatin was well tolerated, with no associated increase in blood pressure.^98,102 However, to date no clinical outcome trials with JTT-705 have been completed.

Apo A-I–Directed Therapies. Current apo A-I–based strategies include infusion of recombinant phospholipids/apo A-I complexes (rHDLs), recombinant proapoliprotein A-I, apo A-I, or apo A-I mimetic peptides, thereby increasing circulating levels of lipid-poor acceptors of cholesterol (Table 4).

Weekly rHDL infusions with ETC-216, a recombinant apo A-I Milano (mutant)/phospholipids complex, were shown to promote regression of coronary atheroma plaque after 5 weeks, as measured by intravascular ultrasound, in patients with acute coronary syndromes (Table 3).⁸³ Additionally, the Effect of rHDL on Atherosclerosis–Safety and Efficacy (ERASE) trial evaluated the effect of rHDL in 183 patients with acute coronary syndromes using intravascular ultrasound and angiographic outcomes.⁸⁴ Compared with placebo, the primary end point of percentage change of atheroma volume with rHDL did not reach significance (Table 3).⁸⁴ Cholesterol efflux from foam cells via the ABCA1 pathway is one possible antiatherogenic mechanism of these infusion strategies.^83- 84

Mimetics of the 243–amino acid apo A-I are smaller peptides that resemble the lipid-binding domain of apo A-I and have been shown to possess atheroprotective effects in animal models.¹⁰⁵ One such mimetic, L-4F, was shown to improve in vivo vasodilatation and prevent monocyte adhesion, whereas D-4F, an orally active derivative compound that is protected from enzymatic digestion, significantly decreased atherosclerosis in animal models (Table 4).¹⁰⁵ The reported mechanism for these actions is the rapid production of β-migrating particles with increased paraoxonase activity, anti-inflammatory capacity, and ability to promote cholesterol efflux from macrophages.^{105- 106,115} Oral administration of D-4F promoted reverse cholesterol transport in both in vitro and in vivo models by accelerating the early steps of reverse cholesterol transport.¹⁰⁶

In a phase 1 human trial, a single oral dose of D-4F was associated with improved HDL anti-inflammatory function.²³ Results from a parenterally administered peptide, ETC-642, are expected soon. However, these apo A-I–directed therapies do not increase HDL-C levels but rather generate HDL-like particles, modify existing HDL, or both.

Phospholipid-Directed Therapies. Phospholipids are an integral component of HDL that also have become a therapeutic target (Table 4).¹⁰⁷ In a murine apoE null model, an oral synthetic phospholipid, 1,2-dimyristoyl-sn-glycero-3-phosphocholine, showed an increase in HDL-C and apo A-I levels, along with improvement in HDL function and reduction in aortic lesion size.¹⁰⁷ In a small (n = 16) human study, an oral derivative of soy lecithin, phosphatidylinositol, for 2 weeks showed an increase in levels of HDL-C as well as apo A-I (Table 4).¹⁰⁸ In addition, infusions of phospholipid particles (eg, phosphatidylcholine or large unilamellar liposomal vesicles) have been proposed to improve reverse cholesterol transport and atherosclerosis regression.¹⁰⁹

Cannabinoid-1 Receptor Blockers. Rimonabant, the first selective cannabinoid-1 receptor blocker with anorexant properties, has been shown to increase HDL-C levels by 10%, a phenomenon that is dose-dependent and weight loss–independent (Table 2).^31- 32 Compared with placebo, patients who received 20 mg of rimonabant had HDL-C levels twice that expected from weight loss alone, suggesting a direct pharmacological effect of rimonabant on lipid metabolism.^31- 32 Furthermore, HDL-C levels increased continuously throughout the 2-year study, even though body weight stabilized.^31- 32 Several large-scale trials to assess the effect of rimonabant on clinical outcomes are in progress; however, the drug was denied approval in the United States because of concerns regarding possible central nervous system toxicity.

Evolving Agents That Target Reverse Cholesterol Transport and Macrophage Cholesterol Efflux. PPAR agonists currently studied to increase HDL-C levels belong to 3 classes: PPAR-α, PPAR-γ, and PPAR-δ (formerly PPAR-β), all members of the family of nuclear receptor transcription factors involved in fatty acid metabolism.¹⁷ Unlike the first-generation PPAR-α agonists (fibrates), newer PPAR-α agonists, eg, NS-220, are more selective and potent (Table 4).¹⁰³ Another PPAR-α agonist, LY518674, raised safety concerns including increases in levels of creatinine and LDL-C.¹¹⁶ PPAR-γ activators used in type 2 diabetes mellitus, such as thiazolidinediones, increase HDL-C levels by 5% to 15% (Table 2)^76- 77; however, they have been associated with increased CVD events.⁷⁸ A PPAR-δ agonist, GW501516, increased HDL-C levels by 80% in a simian model (Table 4).¹⁰⁴ The results of a phase 2 human study are pending.¹¹⁷

PPAR agonism has been postulated to increase macrophage cholesterol efflux and the anti-inflammatory properties of HDL.^14,17 Dual PPAR- α and PPAR-γ agonists, such as muraglitazar, have been shown to increase HDL-C levels by as much as 16%^118- 119; however, an analysis of phase 2 and 3 trials of muraglitazar showed a significant increase in the composite end point of all-cause mortality, nonfatal myocardial infarction, stroke, transient ischemic attack, and congestive heart failure, prompting termination of its development.¹²⁰

Basic science studies and clinical investigations both support the inverse relationship between HDL-C levels and atherosclerosis. HDL augments reverse cholesterol transport and has antioxidative, anti-inflammatory, antithrombotic, and vasoprotective effects.^11,15- 16 Despite the preponderance of evidence linking low HDL-C levels to cardiovascular morbidity and mortality, there is no definitive evidence proving that increasing HDL-C levels reduces the incidence of major cardiovascular events. Indeed, major clinical guidelines have avoided providing an HDL-C target, despite some consensus statements that increasing HDL-C levels has merit.^9,42

Current guidelines from the Adult Treatment Panel III emphasize targeting primarily LDL-C, secondarily non–HDL-C, and then HDL-C.⁴² The recent American Heart Association/National Heart, Lung, and Blood Institute scientific statement proposes that HDL-C be a “tertiary target” (after LDL-C and triglycerides), with goals of HDL-C levels more than 40 mg/dL in men and more than 50 mg/dL in women.⁷ The American Diabetes Association proposes that HDL-C be a “secondary target” along with triglycerides, with a goal of HDL-C levels more than 40 mg/dL.¹²¹ Both advisories recommend consideration of niacin and fibrates to increase HDL-C levels if lifestyle modifications alone are inadequate.^7,121

Evolving HDL-based approaches can be categorized into those that modulate HDL composition (HDL-C, apo A-I, and phospholipids) and those that enhance reverse cholesterol transport. A third category, modulating antioxidant and anti-inflammatory functions of HDL, is common to both of the above approaches and may become an independent strategy of its own. The failure of recently developed agents that substantially increase HDL-C levels suggests that functionality of HDL may be a more appropriate target than HDL-C levels themselves.^{18- 19,99,122} In addition, the relationship between systemic inflammation and HDL function may be particularly relevant.¹⁰²

The functionality of different HDL subfractions appears to vary substantially. Of the known forms of HDL-C (pre-β-HDL, HDL₂, HDL₃) pre-β-HDL appears to be the most antiatherogenic form.¹²³ Therefore, therapies that increase the most atheroprotective subfraction(s) of functioning HDL may be most promising. Additionally, functional testing of HDL may provide insight as to the therapeutic promise of investigational compounds.

While LDL-C–lowering strategies have consistently reduced CHD risk, HDL-based approaches are much more complex and sometimes disappointing. In the last year, the development of pactimibe,¹²⁴ an acyl-cholesterol acyltransferase inhibitor, and now torcetrapib, have been abandoned. These discontinuations followed prior adverse experiences with some members of the PPAR class.^78,116,120

The negative results with these compounds do not refute the concept of increasing HDL-C levels, targeting HDL function, or both to treat atherosclerosis. However, the simple goal of increasing levels of “good” cholesterol can no longer be applied to all forms of HDL without consideration of therapeutic effect on HDL function and ultimately cardiovascular risk. Correlation of HDL functional changes with long-term outcome studies may allow for prospective validation of assays that measure HDL function.

In the meantime, appropriate strategies to increase HDL-C levels include aggressive overall lifestyle modification (exercise, diet, weight loss, and smoking cessation). Patients at increased risk for CHD with low HDL-C levels benefit from niacin and statin therapy, even though the effects of statins on HDL-C levels are relatively modest.^73- 75 Niacin and fibrate therapy also merit consideration as monotherapies or in closely supervised combination with statin therapy in patients at high risk for CHD.