Key Points
- •
HDL-C is inversely correlated with coronary heart disease.
- •
In addition to its pivotal role in reverse cholesterol transport, HDL possesses anti-inflammatory, fibrinolytic, and antioxidant properties.
- •
Levels of HDL-C inadequately characterize HDL functionality.
- •
Therapies that raise HDL-C levels are associated with reduced atherosclerotic progression.
- •
Clinical outcome studies are evaluating whether raising of HDL-C levels independently reduces coronary heart disease risk.
Epidemiologic and Pathophysiologic Considerations
Epidemiology
During the past several decades, numerous observational studies have strongly portrayed high-density lipoprotein cholesterol (HDL-C) as an independent risk factor for coronary heart disease (CHD). In the United States, the Framingham Heart Study provided the strongest examples of this relationship. As demonstrated in Figure 13-1 , the risk of CHD was highest at the lowest levels of HDL-C in the Framingham Heart Study, even when low-density lipoprotein cholesterol (LDL-C) levels were not elevated (i.e., <100 mg/dL). Conversely, high levels of HDL-C conferred relative cardioprotection compared with lower HDL-C levels, even when they were accompanied by high LDL-C levels (i.e., >220 mg/dL).
Inverse relationships between HDL-C and CHD have also been demonstrated outside of the United States as exemplified in Tromsø, Norway, where a threefold greater risk of future CHD was conferred by low HDL-C independently of any other variables. In western Europe, the Prospective Cardiovascular Münster (PROCAM) study observed more than 25,000 men and women without symptomatic CHD at baseline (1979-1991). Among the most prominent findings was that low HDL-C, defined as <35 mg/dL, conferred a 2.5-fold increase of incident CHD in the absence of elevated total cholesterol (<200 mg/dL) and a 5-fold increase at higher total cholesterol levels.
Other prospective studies confirmed HDL-C to be inversely correlated with incident myocardial infarction even after adjustment for other risk factors, such as age, smoking, blood pressure, weight, and diabetes mellitus. If low HDL-C predicted CHD events, might there be a useful metric to gauge clinical effects related to raising of HDL-C levels? In fact, before clinical outcome studies addressing the effect of raising HDL-C on CHD events (see later), observational data from the Multiple Risk Factor Trial, the Lipid Research Clinics follow-up trial, the placebo arm of the Coronary Primary Prevention Trial, and the Physicians’ Health Study suggested that an HDL-C increment of 1 mg/dL would translate into an approximate 3% reduction in CHD risk.
Taken together, observational data from prior decades (1970s-1990s), with few exceptions, paint a vividly convincing picture in favor of HDL-C as an independent risk factor for CHD. Similarly, clinical trials assessing the effect of lipid-altering therapy on outcomes found baseline measurements of low HDL-C to be predictive of both initial and recurrent CHD events. Moreover, patients with low HDL-C assigned to placebo therapy demonstrated increased atherosclerotic progression and posed the highest CHD risk in clinical trials evaluating statin therapy.
Pathophysiology of High-Density Lipoprotein
In addition to reverse cholesterol transport, HDL possesses antioxidant, anti-inflammatory, and antifibrinolytic properties, all believed to contribute to its putative atheroprotective role. However, the extent to which these properties translate into clinical improvement vis-à-vis CHD outcomes awaits the results of ongoing clinical trials (see later).
The physiologic pathway underlying reverse cholesterol transport is shown in Figure 13-2 . Originating as an HDL precursor, lipid-depleted and hepatically or intestinally derived apolipoprotein (apo) A-I receives phospholipids and free cholesterol after hydrolysis of triglyceride-rich lipoproteins. The relatively lipid-poor pre-β (based on electrophoretic mobility) HDL interacts with the ATP transport protein ABCA1 to shuttle free cholesterol and phospholipids from peripheral cells (e.g., macrophages). Cholesterol is sequestered into the HDL core through esterification by lecithin-cholesterol acyltransferase (LCAT) to form small spherical HDL 3 (or α 2 , α 3 HDL) particles with approximately two molecules of apo A-I and a small percentage of particles containing hepatically derived apo A-II.
Additional contributions of cholesterol mediated by ABCG1 and ABCG4 (and, to a minor extent, the scavenger receptor B1 [SR-B1]) result in HDL maturation and larger HDL 2 particles (or α 1 ) that contain three or four molecules of apo A-I. Whereas apo A-I plays an important role in stabilizing HDL, activating LCAT and promoting reverse cholesterol transport, the role of apo A-II in this process is less clear. The fate of cholesteryl esters contained within HDL includes transfer to lower density lipoproteins (e.g., LDL, VLDL) in exchange for triglyceride, a process mediated by the cholesteryl ester transfer protein (CETP).
The cholesteryl ester transferred to very-low-density lipoprotein (VLDL) is converted to LDL and taken up by LDL receptors for hepatobiliary delivery and excretion. Alternatively, HDL may deliver cholesteryl ester to steroidogenic tissues (liver, adrenal, testis, ovaries) by binding to a SR-B1 protein, when it may serve as a precursor for hormone and gonadal steroid production. In cases of CETP inhibition (see later), apo E–enriched HDL may also deliver hepatic cholesterol by LDL receptor–related mechanisms. Other less established contributors to reverse cholesterol transport include apo A-I uptake by high-affinity hepatic beta-chain ATP synthase receptors and by the renal proximal tubule endocytic receptor cubilin.
In addition to its role in reverse cholesterol transport, HDL possesses other atheroprotective properties ( Fig. 13-3 ). Specifically, HDL downregulates expression of interstitial vascular cell adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) in vascular endothelial cells. This limits transendothelial migration of monocytes and macrophage conversion, thereby reducing a proinflammatory milieu. In addition, the HDL-associated enzymes paraoxonase 1 (PON1) and platelet-activating factor acetylhydrolase (PAF-AH) inhibit LDL oxidation. The inverse association between PON1 activity and CHD validates the potential clinical relevance of non–reverse cholesterol transport characteristics of HDL.
After the discovery that apo A-I is a prostacyclin-stabilizing factor, attention has been drawn to the HDL-associated phospholipid sphingosine 1-phosphate (S1P) because of an array of biologic effects on vascular endothelial and smooth muscle cells. For example, HDL-associated S1P improves arterial tone by upregulation of endothelial nitric oxide synthase. Moreover, in cellular studies and animal models, administration of HDL-associated S1P reduced infarct size and protected endothelial cells against apoptosis.
HDL has also been shown to reduce platelet activation by upregulating prostacyclin and nitric oxide production and decreasing thromboxane A 2 synthesis as well as by downregulating tissue factor expression. Taken together, reverse cholesterol transport is likely to account for only a proportion of the cardioprotective benefit attributable to HDL.
Beyond reverse cholesterol transport and other putative atheroprotective properties is the recent identification of serine protease inhibitors and complement-modulating proteins associated with HDL ( Fig. 13-4 ) that may protect against proteolysis and plaque rupture. Although in its early developmental stages, HDL proteomics is likely to advance our understanding of the diverse roles that HDL plays in vascular biology and atherothrombosis.
Genetics of High-Density Lipoprotein
High-Density Lipoprotein Deficiency States and Coronary Heart Disease Risk
To date, chromosomal aberrations in the APOA1/C3/A4/A5 gene complex or mutations in ABCA1, LCAT, and APOAI have been implicated in HDL deficiency states (e.g., HDL-C <10 mg/dL). Yet, in the absence of other CHD risk factors, premature CHD, especially before the age of 40 years, has not been reported with heritable HDL deficiency. This is in marked contrast to familial hypercholesterolemia, in which, irrespective of risk factor status, homozygotes commonly develop CHD in childhood or adolescence and heterozygotes may manifest CHD before the age of 30 years.
The lack of premature CHD in the absence of other risk factors (e.g., smoking, diabetes) suggests that there are other mechanisms enabling cholesterol efflux as the initial step in reverse cholesterol transport. For example, in cases of ABCA1 deficiency, cholesterol efflux occurs through SR-B1, passive diffusion, or upregulation of sterol 27-hydroxylase (CYP27A1) and caveolin 1 ( Fig. 13-5 ). Similarly, in the absence of apo A-I, other apolipoproteins such as macrophage-derived apo E may contribute to reverse cholesterol transport.
A third possibility is that from a teleologic standpoint, HDL was designed to remove only excess cholesterol to maintain cellular lipid homeostasis. If this is correct, “isolated” low HDL-C as defined by physiologic levels of LDL-C and triglyceride (e.g., less than 70 mg/dL and 100 mg/dL, respectively) without traditional CHD risk factors would not be associated with increased risk of premature CHD.
In fact, as noted before, HDL-C deficiency as a consequence of monogenic abnormalities has rarely if ever been associated with premature CHD in the absence of accompanying CHD risk factors, most notably cigarette smoking. Rather, the increased CHD risk associated with low levels of HDL-C may in large part be ascribed to associated metabolic perturbations (e.g., visceral adiposity, insulin resistance) that upregulate proinflammatory signaling pathways and raise overall atherothrombotic risk. This is in striking contrast to familial hypercholesterolemia, in which, in the pre–statin era, premature CHD occurred in most affected subjects, irrespective of risk factor status.
In contrast, elevated levels of HDL-C (i.e., >60 mg/dL) have been viewed as a negative CHD risk factor. Yet, the association of inheritable high HDL-C (>60 mg/dL) with reduced CHD risk remains as debatable as isolated low HDL-C and premature CHD. For example, variation in the HDL regulatory gene CETP is associated with high HDL-C and is especially prevalent in Japan, where approximately 50% of high HDL-C is a consequence of genetic CETP deficiency. Despite several case reports suggesting reduced CHD risk ( Table 13-1 ), data are inconclusive as to whether intrinsic cardioprotection is afforded as a consequence of CETP inhibition.
| Mutation | Location | HDL-C (mg/dL) | Atherogenicity | Reference |
|---|---|---|---|---|
| Int14A | Japan | >70 | ↑ | |
| D442G | United States | ≤60 | ↑ | |
| >60 | ↓ | |||
| R37X | Sweden | 208 | ↓ | |
| R268X | Canada | 70 | ↓ | |
| IVS7+1 | Netherlands | 80 | ↑↓ | |
| Q87X/Q165X | Greece | 194 | ↓ |
* Defined by history of coronary heart disease or noninvasive evaluation (carotid intima-media thickness, computed tomographic angiography).
To address the potential benefit of CETP inhibition, a multicenter trial is currently evaluating whether the CETP inhibitor dalcetrapib reduces CHD outcomes (see later). In addition to CETP, variation in hepatic lipase ( LIPC ) and endothelial lipase ( LIPG ) has also been associated with high levels of HDL-C. However, LIPC variants have not been associated with cardioprotection, and it remains unclear whether the loss-of-function LIPG variant Asn396Ser affects atherogenicity.
In addition to loss-of-function variants that despite being rare in the general population (<1%) produce a significant phenotypic effect (e.g., very low or high HDL-C), genome-wide association studies investigate informative loci that may also contribute to HDL-C levels. In addition to known genes that regulate HDL metabolism and function, novel HDL candidates have recently been uncovered through genome-wide association studies, a sample of which is represented in Table 13-2 .
| Gene | Chromosome | ΔHDL | Reference |
|---|---|---|---|
| GALNT2 | 1 | ↑ | |
| PCYT1 | 3 | ↑ | |
| ELOV2 | 6 | ↓ | |
| VNN1 | 6 | ↑ | |
| BMP1 | 8 | ↓ | |
| MMAB/MVK | 12 | ↓ |
The potential mechanism of action of these encoded proteins on HDL metabolism is outlined in Figure 13-6 . Although the impact on CHD risk has yet to be defined, the addition of genome-wide association studies to rare variant identification is likely to provide new insights related to HDL metabolism and overall atherothrombotic risk.
High-Density Lipoprotein Functionality and Classification
One of the quagmires relating HDL to CHD risk assessment is how best to assess and to represent HDL functionality. Although the level of HDL-C is most commonly used to gauge CHD risk, it inadequately characterizes the HDL proteome in terms of inherent metabolic complexities and effectiveness of reverse cholesterol transport. This in part reflects the inherent difficulty in accurately quantifying the contribution of macrophage-derived fecal cholesterol, a small measure of reverse cholesterol transport. Nonetheless, the recent development of macrophage-specific reverse cholesterol transport assays in murine models may hold promise as a future diagnostic tool if results are reproducible in humans.
In addition to cholesterol efflux and reverse cholesterol transport, other putative measures of HDL-mediated atheroprotection are currently under exploration. They include evaluation of reconstituted HDL, apo A-I Milano and mimetic compounds on indices of inflammation, hemostasis, thrombosis, and endothelial function.
Although HDL-C levels have traditionally been used as a surrogate for reverse cholesterol transport, experimental evidence favors additional measurements that may contribute to CHD risk stratification. For example, in the setting of insulin resistance or in hypertriglyceridemic states ( Fig. 13-7 ), free fatty acids are mobilized from adipocytes to drive hepatic VLDL production. The enhanced synthesis of triglyceride-rich lipoproteins leads to upregulation of CETP, resulting in greater exchange of triglyceride and cholesteryl ester between VLDL and HDL. Hypertriglyceridemic HDL particles exhibit reduced efficiency of cholesterol efflux and are subsequently hydrolyzed by hepatic lipase to produce small, dense cholesterol-depleted HDL particles.
Apo A-I is catabolized by cubilin receptors in the proximal renal tubule, and higher apo A-I fractional catabolic rates account for the reduced HDL-C levels found in postprandial states as well as in obese, insulin resistance, and diabetic states. This in part reflects upregulation of CETP in triglyceride-enriched apo B–100 containing lipoproteins (most notably, VLDL), thereby permitting greater exchange of triglyceride–cholesteryl ester with HDL. The triglyceride-enriched HDL serves as an excellent substrate for hepatic lipase, resulting in small, dense, apo A-I–depleted HDL particles.
Whereas greater apo A-I stability might suggest a more resounding cardioprotective effect for larger HDL particles, observational studies have been inconsistent. For example, several studies have found HDL 2 to be associated with lower CHD risk, whereas others, including the Physicians’ Health Study and EPIC-Norfolk prospective population study, failed to find significant differences between HDL particle size and risk of incident CHD after adjustment for traditional risk factors and other covariates.
A final lingering concern is the potential conversion of HDL to a proinflammatory form ( Fig. 13-8 ). For example, in the setting of an acute coronary syndrome (ACS), the enzyme myeloperoxidase released from leukocytes may bind to and oxidatively modify apo A-I. In addition to reducing the effectiveness of apo A-I in reverse cholesterol transport, the associated nitration and chlorination of apo A-I convert HDL to a proinflammatory particle that may be selectively incorporated in human atheroma.
Moreover, displacement of apo A-I by the acute-phase reactant serum amyloid A and associated reductions in LCAT and the HDL antioxidants PON1 and PAF-AH eliminate many of the cardioprotective properties of HDL. Statins may offset some of the adverse effects associated with proinflammatory HDL particles after ACS.
Therapies that Affect High-Density Lipoprotein
Lifestyle Changes Leading to Increased Levels of HDL-C ( Table 13-3 )
Diet and Weight Loss
The International Diabetes Federation and National Cholesterol Education Program consider abdominal adiposity and low HDL-C part of the metabolic syndrome. Weight loss, especially when it is accompanied by aerobic activity (see also later), may have a significant impact on raising HDL-C levels. In the absence of weight loss, however, increasing the proportion of carbohydrates at the expense of fat reduces both HDL-C and LDL-C levels. Moreover, the active process of weight loss is commonly associated with transient reductions in HDL-C, especially when a low-fat diet is prescribed, because of reductions in apo A-I production.
| Lifestyle Component | Percentage Increase in HDL-C |
|---|---|
| Diet and lifestyle | 0-5 |
| Weight loss | 5-15 |
| Omega-3 fatty acids | <5 |
| Exercise | 5-20 |
| Cigarette cessation | 5-10 |
| Alcohol consumption | 5-15 |
However, once weight loss has been attained and body weight stabilized, a meta-analysis of 70 diet studies found minimal increases approximating an increase of 1.6 mg/dL in HDL-C for every 10 pounds of weight lost. Finally, high intake of marine-based omega-3 fatty acids has been associated with higher HDL-C in some populations, although the overall net increase is very modest. Specifically, fish oil consumption by diet or capsule form has been associated with a 3% increase in HDL-C in subjects without hypertriglyceridemia at baseline (i.e., <177 mg/dL) compared with no increase in HDL-C among subjects with higher triglyceride levels.
Exercise
Aerobic exercise has been shown to increase HDL-C levels on average by 5% to 10%, with the increase mostly related to the frequency and intensity of the exercise. Increased HDL levels with exercise are associated with upregulation of lipoprotein lipase activity. Overall, 10% to 20% increases in HDL-C are observed if at least 1200 kcal are expended weekly.
Conditioned athletes and endurance runners often have HDL-C levels that are 30% to 50% higher than those of sedentary subjects, probably reflecting the combination of enhanced lipoprotein lipase activity and increased production of pre-β HDL particles, which may facilitate reverse cholesterol transport. Smaller increases (5% to 10%) are observed in subjects who have baseline low levels of HDL-C accompanied by elevated triglycerides and visceral adiposity. On the other hand, isolated low HDL-C is difficult to effectively raise even after extended endurance training.
Smoking Cessation
Overall, cigarette smoking impairs LCAT activity and reduces levels of HDL-C. In women, 1 pack per day smokers evidenced HDL-C levels that were 10 mg/dL lower compared with those of nonsmokers, whereas an approximate differential of 3 mg/dL in HDL-C favoring nonsmokers was observed in men. In contrast, a meta-analysis of 27 studies found significant increases in HDL-C levels after smoking cessation.
In a study performed by Moffatt, participants who stopped smoking for 60 days saw an average increase of 12.5 mg/dL in HDL-C; reinitiation of cigarettes resulted in reversion to precessation levels.
Alcohol
Alcohol inhibits hepatic lipase, thereby raising both HDL-C and its HDL 2 subfraction. A dose-response relationship exists; 1 ounce of alcohol consumed daily is associated with up to a 15% increase in HDL-C. It has been suggested that raising of HDL levels represents approximately half of the CHD benefit attributable to alcohol use.
The type of alcoholic drink does not appear to be as important as the quantity. HDL-C increases due to alcohol intake may be more noteworthy in sedentary subjects compared with those who exercise on a regular basis. Caution should be exercised in patients with low HDL-C who have elevated triglycerides (i.e., >200 mg/dL) as alcohol may substantially raise these levels as well.
Pharmacologic Therapy
In addition to lifestyle recommendations, medications have been employed to raise levels of HDL-C. It remains to be established whether and to what extent raising of HDL-C in and of itself reduces CHD risk. We review current and investigation agents affecting HDL metabolism as well as ongoing clinical trials that it is hoped will provide insights into the virtue of raising HDL-C or improving its functionality.
Niacin
Nicotinic acid, or niacin (vitamin B 3 ), is currently the most potent HDL-C–raising medication in the United States, with increases ranging between 20% and 35%. Of the three formulations, immediate-release (IR; three times daily), slow-release (SR; twice daily), and extended-release (ER; once daily), the IR form raises HDL-C to the greatest extent, followed by ER and SR formulations. For example, at doses of 1000 mg, the IR formulation raises HDL-C 25%, compared with 15% to 20% with ER and 10% with SR formulations. At doses of 1500 to 2000 mg daily, observed increases range from 25% to 35% (IR) versus 20% to 30% (ER) and 10% to 20% (SR).
Niacin reduces HDL catabolism by inhibiting hepatic apo A-I removal. Niacin also inhibits adipose hormone–sensitive lipase, reducing free fatty acid flux to hepatocytes, thereby decreasing VLDL production and triglyceride output. Based on the inverse association between triglycerides and HDL-C, for every reduction of 50 mg/dL in triglycerides, HDL-C levels rise by approximately 1.6 mg/dL in addition to apo A-I–mediated effects. Preliminary data for a combination of niacin and statin therapy found increases in HDL 3 levels of apo J and phospholipid transfer protein, raising the specter of improved reverse cholesterol transport, although the direct contribution by niacin (and statin) treatment remains to be determined.
In the Coronary Drug Project (CDP), 3 g of IR niacin daily was associated with reduction in nonfatal myocardial infarction after 5 years and a total mortality benefit of 11% compared with placebo. Coronary arteriography and studies evaluating carotid intima-media thickness have demonstrated that the addition of niacin to statin-based therapy is associated with reduced progression and in some cases regression of atherosclerotic disease.
For example, in the Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol 3 (ARBITER 3), the addition of niacin to statin therapy resulted in decreased progression and mild regression of carotid intima-media thickness at 12 and 24 months, respectively. Moreover, the HDL Atherosclerosis Treatment Study (HATS) found that the combination of niacin and statin decreased atherosclerosis by angiography as well as decreased clinical events compared with placebo. In this study, despite a low sample size, those randomized to the combination (compared with placebo) had a significant 90% lower rate of recurrent cardiac events, noteworthy given that most statin studies have shown maximum event reductions on the order of 30% to 40%. HATS set the stage for the clinical outcome trial Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglycerides and Impact on Global Health Outcomes (AIM-HIGH) designed to evaluate whether the combination of niacin and statin (with or without ezetimibe) is clinically superior to a predominant LDL-C–lowering regimen alone.
The most common side effect of niacin therapy is the prostaglandin D 2 –mediated cutaneous reaction that includes flushing and, to a lesser extent, urticaria. Pretreatment with 325 mg of chewable or nonenteric aspirin may limit or abort this reaction. Whereas the intensity of this side effect is reduced with food and ER or SR formulations and often eases with continued administration, instances of renewed flushing may occur even with long-term administration as a result of insufficient food intake, overconsumption of alcohol, overexposure to heat, and aerobic activity.
In this regard, a selective prostaglandin D 2 receptor 1 antagonist (laropiprant) that is associated with reduced flushing is currently being studied in a clinical outcome trial (see also later). Other less frequent niacin-based side effects include dyspepsia, gout, acanthosis nigricans, toxic amblyopia, and elevation of plasma glucose concentration. However, studies suggest that diabetic subjects receiving up to 3 g IR or 2 g ER niacin had relatively modest (5% to 7%) but clinically insignificant increases in fasting glucose levels. In fact, a CDP post hoc analysis of diabetic and metabolic syndrome patients found that niacin therapy improved CHD outcomes.
The most disconcerting side effect, hepatic toxicity, has not been encountered with the ER regimen. Nevertheless, the results of the AIM-HIGH trial and the Heart Protection Study 2–Treatment of HDL to Reduce the Incidence of Vascular Events (HPS2-THRIVE) comparing the combination of nicotinic acid/laropiprant and statin therapy versus statin-based therapy alone will in large part dictate the future of this therapy in patients with vascular disease.
Fibrates
Fibrate therapy generally results in a 10% to 25% increase in HDL-C levels. Fibrates are synthetic agonists of peroxisome proliferator-activated receptor α (PPARα), which stimulates expression of the hepatic apolipoprotein A-I gene and modulates the transcription of other genes involved in reverse cholesterol transport, including SR-B1 and ABCA1. However, the relative increase in HDL-C is largely driven by the associated reductions in triglyceride levels, the primary lipid-mediated effect of fibrates. In primary prevention, the Helsinki Heart Study of 4081 asymptomatic middle-aged men with elevation of non HDL-C ( > 200 mg/dL) showed the fibrate gemfibrozil to be associated with a 34% overall reduction in incident CHD. However, lower median HDL-C at baseline (i.e., <42 mg/dL) was associated with improved CHD outcomes despite more modest raising of HDL-C compared with higher baseline HDL-C levels.
Moreover, in the secondary prevention Veterans Affairs HDL Intervention Trial (VA-HIT), recurrent CHD events were reduced by 11% with every increase of 5 mg/mL in HDL-C when patients were treated with gemfibrozil. In this study, the HDL subfraction HDL 3 was correlated to a greater degree than total HDL-C levels with CHD. Moreover, gemfibrozil treatment resulted in a 10% increase in HDL particle number that correlated with a 29% reduction in CHD risk.
Statins
Although statins exert a more modest effect on HDL-C levels (5% to 15% increase), they are especially effective in patients with low HDL-C, in whom they may attenuate the elevated risk associated with reduced levels (see Fig. 13-7 ). For example, in the Lipoprotein and Coronary Atherosclerosis Study, patients with reduced HDL-C levels (on placebo) evidenced the highest rates of arteriographic progression. However, statin treatment resulted in greater decrease in progression in low versus high HDL patients that was likely a consequence of reduction in atherogenic lipoproteins and inflammation rather than raising of HDL-C.
Statins raise HDL-C levels in part by reducing CETP activity and increasing apo A-I synthesis. Among the different statins, rosuvastatin may raise HDL-C levels at the upper end of the spectrum (15%). Similarly, simvastatin raises HDL-C and apo A-I levels, with the most robust increases (10% to 15%) observed at the highest dose (80 mg).
Drugs That Reduce HDL-C
Several classes of drug therapies have also been found to lower HDL-C levels. In one comprehensive review, 474 trials investigated the effects of 85 antihypertensive drugs on lipids and blood pressure. Nonselective beta blockers were shown to have a negative effect on HDL-C (10% to 20%) levels. Subjects taking benzodiazepine derivatives have been shown to have HDL-C levels 3.3 mg/dL lower than that of nonusers. However, these lipid effects may be secondary to the weight gain and accompanying increases in triglycerides.
Among the most potent HDL-C–lowering agents are anabolic steroids. Testosterone increases hepatic expression of hepatic lipase and SR-B1, with resulting reductions of HDL-C up to 90%. Fortunately, this effect can be reversed within 1 month of steroid discontinuation. Taken together, beta blockers and anabolic steroids are the two classes of drugs most associated with reductions in HDL-C.
Novel Targets of High-Density Lipoprotein Metabolism
CETP Inhibition
As described before, CETP mediates the transfer of cholesteryl esters from HDL to apo B–containing lipoproteins, and inheritable CETP deficiency is associated with intrinsically high total and HDL 2 cholesterol levels. Although vaccine-based strategies are in development, early testing has yielded only modest increases in HDL-C (5% to 10%). In contrast, several oral compounds have demonstrated more appreciable increases (>25%). The most well studied, torcetrapib, irreversibly bound to CETP and resulted in 50% to 100% increases in HDL-C levels.
Unfortunately, the clinical trial Investigation of Lipid Level Management to Understand Its Impact in Atherosclerotic Events (ILLUMINATE) was prematurely terminated because of off-target toxicity related to aldosterone stimulation. The increase in electrolyte abnormalities (e.g., increased plasma bicarbonate and reduced potassium) was believed to contribute in part to the higher mortality rate in the torcetrapib-treated patients.
Another CETP inhibitor, dalcetrapib (JTT-705), regressed aortic atherosclerosis in rabbits. Dalcetrapib raises HDL-C more moderately (25% to 40%) than torcetrapib does, but it does not irreversibly bind to CETP or upregulate aldosterone secretion and is currently under evaluation in a large clinical outcomes trial. A third compound, anacetrapib, also exhibits potent HDL-C–raising properties without affecting aldosterone. Pending the outcome of ongoing clinical studies, CETP inhibitors hold potential promise as adjuvant therapy in high-risk patients with dyslipidemia.
Apolipoprotein A-I–Targeted Therapies
On the basis of experimental evidence that intravenous HDL infusions or apo A-I overexpression reduced atherosclerosis in animals, there has been great interest in human-based therapy targeting HDL. Several different strategies include administration of intravenous apo A-I, reconstituted phospholipid–apo A-I complexes (rHDL), oral apo A-I mimetic compounds, and phospholipid-based therapy.
The first human study to gain extensive media attention was the reduction in atheroma volume (assessed by intravenous ultrasound) after five weekly infusions of apo A-I Milano/phospholipid complex in post-ACS patients. In the randomized study Effect of rHDL on Atherosclerosis Safety and Efficacy (ERASE), 183 ACS patients received four weekly infusions of saline or 40 mg/kg or 80 mg/kg of HDL mimetic. Although atheroma burden was not different among the groups, there was improvement in the plaque characterization index and coronary score on quantitative coronary angiography in favor of rHDL.
Apo A-I mimetic peptides are much smaller compounds than mature apo A-I (e.g., 18 versus 243 amino acids) and possess similar apo A-I/lipid-binding domains to promote cholesterol efflux, to decrease inflammation, and to improve endothelial function. For example, the mimetic D-4F improved the anti-inflammatory capacity of HDL after a single dose without altering HDL-C levels.
Phospholipid administration may also be a valuable HDL-targeted therapy. For example, the synthetic compound 1,2-dimyristoyl- sn -glycero-3-phosphocholine (DMPC) raised HDL and apo A-I levels in association with reduced aortic lesions in a murine study. Moreover, the soy-derived phospholipid phosphatidylinositol was found to raise HDL-C levels 13% to 18% during a 2-week period in normolipidemic subjects.
Liver X Receptor Agonists
The liver X receptor (LXR), a nuclear hormone receptor ( Fig. 13-9 ), forms a heterodimer with the retinoid X receptor and regulates transcription of ABCA1 and ABCG1, thereby serving an integral role in cholesterol efflux and reverse cholesterol transport. Not surprisingly, therefore, synthetic activators of LXR have been pursued as a potential target to improve HDL functionality. Unfortunately, complicating the early-stage testing of LXR agonists was the induction of sterol regulatory element–binding protein 1 (SREBP-1) expression, resulting in enhanced hepatic VLDL production and hypertriglyceridemia.
