ACLS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Levinson, S. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Levinson, S. S.
Annals of Clinical & Laboratory Science 32:123-136 (2002)
© 2002 Association of Clinical Scientists

High Density- and Beta-Lipoprotein Screening for Risk of Coronary Artery Disease in the Context of New Findings on Reverse Cholesterol Transport

Stanley S. Levinson
Department of Pathology and Laboratory Medicine, University of Louisville School of Medicine, and Laboratory Service, Department of Veterans Affairs Medical Center, Louisville, Kentucky

Address correspondence to Stanley S. Levinson, Ph.D., Laboratory Service, VAMC, 800 Zorn Avenue, Louisville, KY 40206, USA; tel 502 895 3401 ext 5565 or 5572; fax 502 894 6265; e-mail levinson{at}louisville.edu.


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Literature Review
 Discussion
 References
 
This article considers how high density lipoproteins (HDL) act as anti-arteriosclerotic agents, examines the usefulness of HDL indexes alone and in conjunction with other markers of coronary artery disease (CAD), and discusses how HDL markers compare with what one might expect from known metabolic mechanisms. This is accomplished by: (i) an overview of mechanisms associated with CAD, especially new findings regarding reverse cholesterol transport; (ii) a brief review of the clinical literature on biochemical markers for automated use; and (iii) analysis of data for persons with or without angiographically documented CAD. Based on these considerations, the ratio of optimized apo A-I/apo B appears superior to lipoprotein lipid markers for predicting the risk of CAD. Yet the ratio shows poor diagnostic accuracy by itself; it yields poorer diagnostic accuracy than would be expected from assessing the metabolic pathways. Discrimination is improved by using the ratio in conjunction with risk factors defined by the National Cholesterol Education Program (NCEP). Based on receiver-operator characteristic (ROC) curve data, this approach increases the accuracy by 13–14% above that obtained with current lipid markers; it improves discrimination more than the use of inflammatory markers. Apolipoprotein testing is better related to the mechanisms of cholesterol transport, is widely available, and requires only two tests, compared to three, to improve discrimination. However, inclusion of inflammatory markers may need to be considered in the future, when more information is available about their functions and clinical value.

(received 3 December 2001; accepted 22 December 2001)

Keywords: apolipoproteins, arteriosclerosis, C-reactive protein, reverse cholesterol transport


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Literature Review
 Discussion
 References
 
High density lipoproteins (HDL) are operationally defined as those within the density range of 1.063–1.21 g/ml. It has long been known that elevation of high density lipoprotein cholesterol (HDLC) is inversely associated with coronary artery disease (CAD), and a more powerful statistical predictor of CAD than low density lipoprotein cholesterol (LDL) [1]. This inverse association has led to the concept of reverse cholesterol transport (RCT). RCT is a process whereby cholesterol is removed from peripheral tissues directly by HDL and in conjunction with triglyceride-rich lipoproteins deposited in the liver for disposal in the bile. Accumulation of lipids in macrophages is considered an early event in arteriosclerosis, and RCT removes cholesterol from these cells. Although HDL metabolism is not as clearly understood as LDL metabolism, during the past decade considerable information has been acquired regarding HDL species, receptors, and transporters. This has resulted in better knowledge of HDL metabolism and RCT. Several assays for HDL markers are now available for the clinical laboratory, either as kits or in convenient procedures for automated or semi-automated testing. These markers include total HDLC, HDL2C, HDL3C, apo A-I, apo A-II, and apo E.

The purposes of this article are: (i) to determine which HDL marker(s) best predict the risk of CAD; (ii) to consider whether clinical studies support this premise; (iii) to identify which marker of beta-lipoproteins, used in conjunction with the selected HDL marker, best predicts the risk of CAD; (iv) to decide whether or not these markers meet theoretical expectations; and (v) to examine how the markers compare with the proposed inflammatory markers.

To achieve these goals, my initial step will be to present new information in the form of receiver-operator characteristic (ROC) curves and multivariate analyses from our studies of HDLC, subspecies of HDLC, apo A-I, and beta-lipoprotein markers. Second, I shall review current knowledge about the basic process of arteriosclerosis, about HDL metabolism and RCT, about the clinical usefulness of HDLC, the subspecies of HDLC, and the apolipoproteins associated with HDL as markers for CAD, and about the proposed anti-atherogenic attributes of HDL. Finally, I shall consider the meaning of the theoretical and actual relationships.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Literature Review
 Discussion
 References
 
Subjects, blood collection, and angiography.  Parameters regarding the patients, blood samples, and angiography are similar to those previously described [2]. Briefly, the patients were males, 40-to 70-yr old, who entered the hospital for clinically indicated angiographic studies. Blood samples were drawn from consecutively examined patients, excepting those with the following exclusion criteria: persons who received known lipid-altering medications (gemfibrozil, HMG CoA reductase inhibitors, bile acid resins, niacin, and heparin), diabetics, persons with chronic kidney disease, and persons who had myocardial infarction within 3 mo. The study was approved by the Hospital and University Committees on Protection of Human Rights.

The following demographic variables were also assessed: smoking, alcohol intake, family history of myocardial infarction, age, medications (ie, beta-adrenergic antagonists, calcium channel blockers, ACE inhibitors, diuretics), and body mass index (BMI = body weight (kg)/[height (m)]2). Nominal variables were classified as yes or no, except alcohol intake, which was classified as "none," "light-moderate," and "heavy," based on 0, 1 to 6, or >6 alcoholic drinks per day.

Patients were fasted over night. Blood for lipoprotein studies was drawn, without preservatives, from the femoral artery prior to infusion of angiographic contrast medium. Serum was obtained by centrifugation at room temperature. Cholesterol and triglyceride assays were performed within 48 hr on serum stored at 4°C. Aliquots were frozen at -70 °C for apo A-I and apo B measurements.

Angiography was performed by standard methods. Patients were categorized into two groups for ROC curve analysis and three groups for discriminant analysis [3,4]. Group NCA comprised patients with normal coronary arteries (<20% stenosis). In the great majority of cases, these patients showed no detectable stenosis. Group CAD comprised patients with severe CAD (>=70% stenosis in at least one major vessel). For discriminant analysis, a third group (Group ICS) comprised the subjects with intermediate degrees of coronary artery stenosis (ie, >20% and <70% in any vessel).

Assays.  Serum cholesterol, HDLC, and triglycerides were assayed by routine methods with a discrete sample analyzer (Eastman Kodak, Rochester, NY). HDLC was separated by precipitation with dextran sulfate-magnesium ion, with magnetic removal of LDL and VLDL (Polymedco, Courtland, NY) [5]. HDL3C was precipitated using dextran sulfate 50,000 (Genzyme, Boston MA), and HDL2C was calculated (HDL2C = total HDLC - HDL3C) [6]. LDLC was calculated by the Friedewald equation on samples with triglycerides <= 4 g/L.

Serum apo A-I and apo B levels were measured by automated rate immunonephelometry using kits with the Array analyzer (Beckman Instruments, Inc., Brea, CA) [7]. The apo A-I kit was modified by addition of 0.5% Tween 20 [8]. Although results obtained with this kit did not show a statistically significant difference with or without addition of Tween 20, the assay was modified to ensure comparability with our prior study [2].

Statistics.  The Kruskal-Wallis test and contingency tables (likelihood ratio) were used for determining possible significant confounders (JMP, SAS Institute, Cary, NC). Based on the contingency tables and Kruskal-Wallis test, the only variables that showed significant relationships with CAD were age and beta-adrenergic antagonists. Age showed a strong relationship (p <0.0001); beta-adrenergic antagonists showed a weak relationship (p 0.05). These variables were considered as possible confounders in stepwise discriminant analysis.

The Kolmogorov-Smirnov test was used for goodness-of-fit testing (Statview 4, SAS Institute). Results of goodness-of-fit testing indicated that HDL cholesterol assays and apo A-I results were significantly different from a Gaussian distribution. When these tests were transformed logarithmically, they fit well. Consequently, for discriminant analysis, log-converted values were used with HDL results.

Stepwise discriminant analysis (SPSS 6.1, SPSS, Inc., Chicago, IL) used Wilks’ lambda with a p for entry of 0.05 and a p for removal of 0.1. Receiver operating characteristic (ROC) curves were calculated by the Rockit program (kindly provided by Dr. Charles E. Metz, Department of Radiology, University of Chicago Medical Center, Chicago, IL; http://www-radiology.uchicago.edu/sections/roc). This program uses the maximum-likelihood-estimation technique for estimating the ROC curve. When logistic regression was combined with ROC analysis, coefficients for logistic regression were obtained from JMP (SAS Institute).

The ninety-fifth percentile significance levels (p <= 0.05) were based on two-tail cutoffs (JMP), except for the ROC areas, where one-tail 95% confidence intervals (CI) were used.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Literature Review
 Discussion
 References
 
ROC analysis.  Comparisons of the areas and confidence intervals (CI ) for the CAD and NCA groups and plots derived from ROC curve analysis are shown in Table 1Go and Fig. 1Go, respectively. Based on the areas and CI values in Table 1Go, HDL2C shows the poorest discrimination among the individual HDL markers. The ratio of HDL2C/HDL3C is the poorest marker, overall, and its CI substantially overlaps 0.5 (area of no discrimination), indicating that this ratio is nugatory. On the basis of the ROC areas, although apo A-I appears to be the best predictor, its area is similar to HDL3C; total HDLC is third. Based on the overlap of the respective CI values, these differences are not statistically significant.


View this table:
[in this window]
[in a new window]
  		Table 1. Areas under the receiver-operator characteristic (ROC) curves for the various lipoprotein tests and selected test ratios, based on data from patients in groups NCA (normal coronary arteries) and CAD (severe coronary disease).
 


View larger version (41K):
[in this window]
[in a new window]
  		Fig. 1. Receiver-operator characteristic (ROC) curves depicting HDL markers and combinations with beta-lipoprotein markers. The solid diagonal line is the line of no discrimination (area = 0.5).

 
Table 1Go indicates that apo B is a better discriminator than LDLC or total cholesterol. Although there is slight overlap between the upper CI of triglycerides and apo B, serum triglycerides have usually been shown by multivariate analysis to provide no additional predictive value when HDL and LDL markers are combined [9]. This feature of triglycerides can be seen from the entries near the bottom of Table 1Go, where ROC analysis from ratios of HDL and beta-lipoprotein markers were combined with triglycerides from the output of logistic regression analysis [10]. The combined area is no greater than the areas for HDL and beta-lipoprotein markers without triglycerides.

As seen in Table 1Go, ROC analysis indicates the ratio of apo A-I/apo B is a superior marker, compared to the ratios of HDLC/total cholesterol or HDLC/LDLC. Examination of plots in Fig. 1Go shows that the curve for the ratio of HDLC/LDLC crosses the curve for apo A-I, so that apo A-I by itself shows better discrimination in the useful range (ie, below a false positive frequency of about 20%).

Discriminant analysis.  The data were also analyzed by multivariate discriminant analysis. Logistic regression and ROC analysis permit comparisons of only two groups. On the other hand, discriminant analysis allows simultaneous comparison of the three groups (ie, NCA, ICS, and CAD), while also correcting for possible confounders. The results of discriminant analysis are listed in Table 2Go. The conclusions are similar to those obtained by ROC analysis. In comparison #1, apo A-I is the best marker. In comparison #2, the other markers were tested in the absence of apo A-I. Here, HDL3C was the better marker, but total cholesterol showed a similar "F to enter" value, indicating that the two tests give similar information. In comparison #3, total HDLC proved to be a better marker than HDL2C. These results show that apo A-I is the best discriminator. This confirms the conclusions from ROC analysis that apo A-I is the superior marker and indicates that the ROC analysis was not overly affected by confounding variables or by the absence of the intermediate (ICS) group.


View this table:
[in this window]
[in a new window]
  		Table 2. Discriminant analysis of HDL markers, based on data from 130 patients in group NCA (normal coronary arteries), 225 patients in group CAD (severe coronary disease), and 55 patients in group ICS (intermediate coronary stenosis). The discriminate analysis was performed after logarithmic transformation of the data.
 

   Literature Review
Top
 Abstract
 Introduction
 Methods
 Results
 Literature Review
Discussion
 References
 
Current view of the process of arteriosclerosis.  The process of atherosclerosis that leads to myocardial infarction involves endothelial cell injury and dysfunction, lipid deposition, abnormal cell migration, chronic inflammation, and thrombosis [11,12]. This process, outlined in Fig. 2Go, occurs largely in the subendothelial space, with rupture into the blood stream. Injury causes trapping of lipoproteins within the vessel wall. Among other causes, injury to the vascular wall may be due to blood pressure fluctuations, toxic products from smoking, diabetic complications, bacterial or viral insults, and elevated lipoproteins. It appears that trapped LDL and other apo B containing lipoproteins are oxidized due to interactions with monocytes, other white blood cells, and oxidizing agents. Macrophages incorporate oxidized LDL in preference to normal LDL. Unlike normal LDL, which is taken up by the well-regulated LDL receptor, oxidized LDL is taken up by scavenger receptors (SR) that appear not to be subject to feedback inhibition. This group of multifunctional receptors, which are capable of taking up a variety of ligands, are homologs of CD36. They include CD36, SR-A, the LDL receptor-related protein (LRP), and SR-BI, of which, as will be discussed, the latter two appear most important in HDL metabolism [13–15]. As a result of this unregulated uptake, macrophages form lipid-laden foam cells; this early event in atherosclerosis initiates the fatty streak [16,17].



View larger version (40K):
[in this window]
[in a new window]
  		Fig. 2. A diagramatic outline of the process of arteriosclerosis. Cross-sectional and diagonal views of a blood vessel are illustrated for each stage. In the advanced stage, the vessel shows a plaque below and a thrombus above.

 
Endothelial cells and white blood cells release diverse cytokines, growth factors, and other proinflammatory mediators that stimulate cell migration and inflammation. This gives rise to a fibrous cap with cholesterol crystals and debris within the deep necrotic layer, while inflammatory cells form a dynamic outer edge. Oxidized lipoproteins may facilitate many of these processes. Consequently, a cycle of oxidation, inflammation, thrombosis, and further oxidation is perpetuated [11,12,16]. Mechanical forces predispose the soft outer layer of the atheromatous plaque to rupture at sites of structural weakness. Rupture of plaques causes thrombosis and incorporation of thrombi into the plaque. Ultimately, the appearance of a large thrombus in an obstructed vessel can lead to sudden ischemia and unstable coronary syndromes.

HDL species and subclasses.  Table 3Go lists the major and minor apolipoproteins that comprise serum HDL, along with their sites of synthesis, and their known functions. Apo A-I is the major protein in HDL, consisting of ~70% of the total protein. Apo A-II comprises approximately 20–25% of the protein, and several minor proteins, of which apo E is the largest, comprise the rest [18]. Several species of HDL have been identified. These include lipid-free nascent apo A-I chains, lipid-poor apo A-I chains containing cholesterol and phospholipid, lipid-rich discoid particles containing apo A-I, spherical particles containing only apo A-I, and spherical particles containing apo A-I and apo A-II [19]. Some spherical particles also contain apo E. Lipid-poor species that contain apo E and apo A-IV have also been described [19]. Most HDL migrates in the alpha region of the gel upon agarose electrophoresis, but lipid-free and lipid-containing apo AI chains, and discoid particles migrate in the pre-beta region and are referred to as "pre-beta HDL" [18–20].


View this table:
[in this window]
[in a new window]
  		Table 3. Apolipoproteins that comprise HDL.
 
Particles that contain apo A-I alone are called "LpA-I"; those that contain apo A-I and apo A-II are called "LpA-I:A-II." The LpA-I:A-II particles contain apo A-I (two to four copies per particle) and apo A-II in a ratio of 2:1 [21]. LpA-I particles appear to promote cholesterol efflux from cells, while LpA-I:A-II particles do not [22,23]. HDL can be assigned to subclasses based on particles of various density [24], of which the main two are the less dense, larger HDL2 (d. 1.063–1.125 g/mL) and the more dense, smaller HDL3 (d. 1.125–1.21 g/mL). HDL2 is richer in particles containing apo A-I without apo A-II, whereas HDL3 is richer in particles containing both apo A-I and apo A-II [25].

Mechanisms of HDL metabolism and reverse cholesterol transport.  Tangier disease is an autosomal dominant disorder in which homozygotes typically show HDLC and apo A-I levels of ~5% and 1% of normal and heterozygotes show levels of ~50% of normal. Homozygotes show 4- to 6-fold increase in the incidence of CAD [26,27]. These patients show increased accumulation of cholesterol esters in macrophages of the reticuloendothelial system [26,27]. It appears that a defect in the ATP-binding cassette transporter (ABC A1) is responsible for this disease [27]. Evidence indicates that ABCA1 transports cholesterol and phospholipid to nascent apo A-I chains from many types of cells and tissues, including liver, lung, adrenal glands, and macrophages [19,27]. This evidence, along with new findings on the roles of phospholipid transport protein (PLTP) and SR-BI in HDL metabolism, has enhanced our understanding of reverse cholesterol transport (RCT).

Fig. 3Go outlines current concepts regarding HDL metabolism. ABCA1 transports cholesterol and phospholipid from tissues to nascent apo A-I. This lipid-free (or lipid-poor) apo A-I may come from the liver or may dissociate from mature particles [28]. Especially interesting is an in vitro finding that ABC A1 from macrophage lines, but not from other tissues, can be up-regulated [29]. Lipid-poor HDL takes up additional cholesterol from tissues, probably by diffusion, facilitated by the phospho-lipid-rich surface of HDL [21,28]. Lipid-rich apo A-I forms discoid particles that are associated with lecithin:cholesterol acyltransferase (LCAT) which catalyses esterfication of free cholesterol for deposit in the particle core [21,30]. Since apo A-I is a powerful activator of LCAT, discoid and apo A-I-rich spherical particles probably accumulate cholesterol ester by diffusion of free cholesterol from tissue to the outer layer with esterfication into the core [21,23,28].



View larger version (36K):
[in this window]
[in a new window]
  		Fig. 3. A diagram of reverse cholesterol transport (RCT). Apo A-I is released from the liver into the circulation. ABC A1 transports cholesterol and phospholipid from tissue to apo A-I, which binds the cholesterol to form lipid-poor LpA-I. The uptake of more cholesterol by efflux from peripheral cells gives rise to discoid LpA-I, which is rich in LCAT. Discoid HDL may become spherical HDL LpA-I particles, losing LCAT, then forming spherical LpA-I:A-II particles. PLTP completes a loop by recycling lipid-poor apo A-I from spherical particles and giving rise to larger HDL3-like HDL from fusion of the unstable remnants. CETP mediates the exchange of cholesterol for triglyceride from beta-lipoproteins. Lipids transferred to beta-lipoproteins are removed from the circulation by the liver. Removal of spherical HDL from the circulation is unclear, except that SR-BI removes cholesterol from HDL, which may be recycled to a prior form. Solid arrows represent the more certain reactions; dashed arrows indicate the hypothesized reactions; a question mark indicates that the role of apo E is unknown.

 
LCAT is not associated with larger spherical particles, but these particles can exchange cholesterol with phospholipid from triglyceride-rich lipoproteins, such as very low density lipoprotein (VLDL), and various remnants [21,28]. The lipid exchange is mediated by cholesterol ester transfer protein (CETP) and phospholipid transfer protein (PLTP) [31]. These triglyceride-rich particles containing apo E may be converted to intermediate density lipoprotein (IDL), which is rapidly taken up by the LDL receptor, and apo E-containing remnants that can be taken up by the LRP [32]. In either case, cholesterol can be removed from HDL by exchange with triglyceride-rich lipoproteins and disposed of through the liver [23,31]. PLTP also mediates the splitting of lipid-poor apo A-I from smaller HDL3-like spherical particles and the subsequent fusion of the unstable residues into a larger HDL2-like particle [19,33]. Hepatic lipase can also split apo A-I from larger HDL particles, but this remodeling does not give rise to larger HDL3-like particles; rather, in conjunction with CETP larger particles can be converted to smaller ones [19]. Thus, as illustrated by the solid arrows in Fig. 3Go, a cycle is perpetuated whereby cholesterol can be transferred from tissue to lipid-poor apo A-I and then to discoid and spherical particles, and finally back to lipid-poor apo A-I, while cholesterol is removed for elimination through triglyceride-rich lipoproteins. A final event in this process seems to be the selective uptake of cholesterol ester from spherical HDL2-like particles by SR-B1, which is highly expressed in liver and steroidogenic tissues [19,27]. In this two-step process, high affinity lipoprotein binding is followed by a transfer of lipid from HDL through the receptor to the cell membrane, after which the depleted particle is released [14]. The evacuated cholesterol may be eliminated in the bile [34].

Although the foregoing information advances our understanding of how RCT proceeds, it is less than complete. The dashed arrows in Fig. 3Go illustrate some of many points that remain unclear. What is the role of apo E in RCT? Apo E in HDL has been hypothesized to be transferred from macrophages, providing a mechanism by which HDL can bind to an apo E receptor. It has been suggested that in this way whole HDL particles can be removed from the circulation. This proposed holo-HDL mechanism may proceed through an apo E receptor or binding to proteoglycans, or by another mechanism [28,31]. In any case, the role of apo E in HDL remains unclear. In fact, there is evidence that apo E-containing particles play little or no role in normal RCT [21]. How does apo A-II enter spherical particles and what is its role in RCT? Lipid-poor apo A-I is reclaimed by the kidney through cubilin, a protein that also acts to reclaim vitamin B12 [19]. What role does this play?

Clinical findings associated with RCT. HDL subtypes.  In the 1980’s, it was suggested that the more dense HDL2 accepts cholesterol esters and is converted to the less dense HDL3 because the HDL3 subtype accepts cholesterol more readily than HDL2 [35,36]. It was suggested that apo E from macrophages might facilitate cholesterol transfer to the less dense HDL subtypes [36]. Studies suggested that HDL2C was a more powerful predictor of CAD than total HDLC or HDL3C [37–41], and that the ratio of HDL2C/HDL3C might be an even more powerful marker for risk of CAD. These findings seemed to support the notion that heavier HDL3 takes up apo E and cholesterol from peripheral tissue and macrophages to form HDL2. Consequently, a high ratio of HDL2C/HDL3C would suggest more active RCT [38,39]. As found here (Table 1Go, Fig. 1Go), subsequent studies did not confirm these findings, but showed that total HDLC and HDL3C are equal or more powerful markers for CAD than HDL2C [42–45]. These reports included one study by investigators who had previously concluded that HDL2C was better, but altered their conclusions based on additional data [44].

Apolipoprotein A-I.  Automated assay kits for apo A-I are available with standardized calibrators [46], but it is unclear that results obtained by all methods give comparable results. While some studies indicated that apo A-I is a better marker for CAD than HDLC [3,4,47], other studies showed no difference between HDLC and apo A-I [43,48,49]. An important reason for this disagreement may be that the antigenic sites of apo A-I are difficult to access by immunological methods, because as much as 75% is hidden by lipids [50]. The problems of immunological non-reactivity and diagnostic inaccuracy seem related. Studies indicated that a detergent, Tween 20, was effective for exposing the antigenic groups of apo A-I [4,51,52]. We showed that treatment with Tween 20 substantially and significantly improved the discriminating power with one assay kit, but caused only small improvement with another kit [8]. After treatment with Tween 20, the apo A-I results obtained by both kits were better than HDLC as a marker for CAD. We concluded that the nature of the detergents used is important in ensuring optimal diagnostic performance with apo A-I testing. Recent studies suggest that optimized apo A-I may be better than lipoprotein lipids for predicting recurrent CAD events in patients treated with statin therapy [53] and may predict CAD in persons without other major risk factors [54].

Apo A-II and apo E.  Based on current thinking regarding RCT, neither apo A-II nor apo E would be expected to be good markers for CAD. Apo A-II inhibits cholesterol uptake from tissue, and LpA-I generally appears to be a better marker for CAD [55]. Although the role of apo E in RCT is unclear, apo E may possibly facilitate cholesterol uptake by the liver. Only a small fraction of the total apo E is found in HDL, so even if HDL-apo E proved to be a useful marker, it would be necessary to develop new tests to measure HDL-apo E. Both of these apolipoproteins appear inferior to apo A-I or HDLC as a risk marker [42,43,56].

Other mechanisms by which HDL may act.  Besides RCT, other potential anti-atherogenic mechanisms have been attributed to HDL [18,57]. Some of these are outlined in Table 4Go. The antioxidant activity of HDL has been well studied. Lipid peroxide accumulation due to copper-mediated oxidation of LDL was shown to be reduced in the presence of HDL [58]. Apo A-I and LCAT were also found to have some antioxidant potential, and paraoxinase-1 (PON1), an aryldialkylphosphatase, had substantially more.


View this table:
[in this window]
[in a new window]
  		Table 4. Anti-arteriosclerotic activities other than reverse cholesterol transport (RCT) that are attributed to HDL.
 
PON1 is a enzyme (43 kDa, 354 amino acids) that splits organophosphate compounds [59]. It is synthesized in the liver and is largely associated with HDL. The ability of PON1 to protect LDL from oxidation involves a cysteine residue at position 283 and is separate from its organophosphatase activity. PON1 has two amino acid polymorphisms–one at position 55 (L->M) and the other at 192 (Q->R). Although it was thought that these polymorphisms are associated with CAD, recent evidence indicates this may not be so [60]. A mechanism was suggested whereby HDL diminishes the accumulation of lipid peroxides by enzymatic hydrolysis of phospholipid hydroperoxides from unsaturated fatty acids of LDL This produces fragments that may be bound to HDL; such a mechanism would remove toxic products from LDL and inhibit autooxidation [59]. It is uncertain whether non-RCT mechanisms are clinically important. For example, paraoxonase is decreased in persons with Tangier disease, yet these persons have a lower frequency of CAD than would be expected [26].


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Literature Review
 Discussion
References
 
In the present study, ROC analysis indicated that HDL3C is nearly equal to apo A-I in predicting the risk of CAD, whereas discriminant analysis, which examined a more diverse group of patients, indicated that apo A-I is a more powerful marker than HDLC or its subfractions. Consequently, the data presented here point to apo A-I as the superior marker for assessing CAD risk. This conclusion is supported by other clinical studies using optimized apo A-I assays [3,4,47,53,54].

This conclusion is also supported by our current understanding of RCT mechanisms (Fig. 3Go). Apo A-I appears largely responsible for cholesterol uptake from the ABCAI transporter. Apo A-I activates LCAT, recycling of spherical particles involves the splitting off of apo A-I, and LpAI particles promote cellular cholesterol efflux, while LpAI:LpAII do not. Yet apo A-I gave an ROC curve area of 0.63, which provides poor diagnostic accuracy [10]. This result is surprising, in view of the widespread participation of apo A-I in RCT. Consequently, despite the recent advances in understanding HDL metabolism, some important mechanisms still need to be elucidated.

Among these, the role of LpAI:LpAII particles remains unclear. It has been suggested that measuring apo A-I in LpAI particles, but not LpAI:LpAII, may enhance diagnostic discrimination [55]. Tall et al [61] suggested that the flux of cholesterol from macrophage foam cells is the critical index in RCT. An increase of apo A-I, HDLC, or ABCAI that leads to decrease in foam cell cholesterol would be only a small change in the overall transport of cholesterol between HDL particles and peripheral tissues. This "macrophage cholesterol efflux" would be too slight to be detected by ordinary methods [61]. The flux of cholesterol through the CETP pathway and its relationship to HDL metabolism is unclear. Men with null mutations in the CETP gene have massively elevated HDL levels and heterozygotes have moderately elevated levels. Men with very high HDL levels showed reduced risk of heart disease irrespective of the mutation, but those with moderately elevated levels showed increased risk [27]. Consequently, it appears that understanding the exact interactions between the pathways may be important, and it may be necessary to measure exact fluxes of different components.

In practice, the problem of poor discrimination of markers for CAD can be moderated by assessing several risk factors at once. The additional inclusion of the Framingham risk factors (age, sex, blood pressure, glucose intolerance, and smoking) significantly improved the area under the ROC curve, compared to that obtained using only HDLC and LDLC [62]. With a computer-assisted model, the area under the ROC curve was increased above 0.8. The recent report of the Adult Treatment Panel-III (ATP-III) of the National Cholesterol Education Program estimates the 10-yr risk of CAD on the basis of age, smoking status, and blood pressure, along with HDLC and total cholesterol [63].

Apo B is a better predictor of coronary risk than total cholesterol or LDLC [64–66]. Theoretically, this would be expected because apo B is found in all fractions of atherogenic lipoproteins; apo B is better than LDLC as a measure of the number of particles of potentially atherogenic small, dense LDL [63,65,66]. Despite the less than desirable area under the ROC curve for apo A-I/Apo B (Fig. 1Go), shifting from the use of HDLC and total cholesterol or LDLC to the use of apo A-I and apo B would improve diagnostic sensitivity by 13–14% at false positive frequencies of 15–20%. This would identify thousands of additional persons at risk.

There has recently been a search for markers to identify more persons at high risk for CAD [67]. Because of evidence that links inflammation to arteriosclerosis, inflammatory markers have been examined [67–69]. Acute phase proteins, especially C-reactive protein (CRP), have been suggested as a marker [68–70], although their linkage to arteriosclerosis is far less clear than apo A-I and apo B. In one study, the addition of CRP to the ratio of total cholesterol/HDLC increased the ROC curve area from 0.64 to 0.68 [71]. In the present study, substitution of apo A-I/apo B for HDLC/total cholesterol (Table 1Go) increases the ROC area from 0.62 to 0.70. This is a comparable or greater increase than that reported by addition of CRP.

The present author suggests that a collaborative study should be initiated to compare apo A-I and apo B kits from different manufacturers, to define the conditions for optimal antigen exposure and performance, and to normalize methods if necessary. This would be a simple way to improve discrimination since the procedures are widely available and only two tests are required. In the future, when more information is known about the value of inflammatory markers, a further evaluation could be performed to see if the addition of a third test (eg, CRP) enhances the discrimination of CAD risk.


   Acknowledgments
 
This work was supported by the Department of Veterans Affairs and by a generous gift of reagents from Beckman Instruments, Inc.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Literature Review
 Discussion
 References
 

  1. Gordon T, Castelli WP, Hjortland MC, et al. High density lipoprotein as a protective factor against coronary heart disease. Am J Med 1977;62:707–714.[Medline]
  2. Levinson SS, Wagner S. Measurement of apolipoprotein B containing lipoproteins for routine clinical laboratory use in cardiovascular disease. Arch Path Lab Med 1992;116:1350–1354.
  3. Kottke BA, Zinsmeister AR, Holmes DR Jr, et al. Apolipoproteins and coronary artery disease. Mayo Clin Proc 1986;61:313–320.[Medline]
  4. Maciejko JJ, Homes DR, Kottke BA, Zinsmeister AB, Dinh DM, Mao SJT. Apolipoprotein AI as a marker of angiographically assessed coronary artery disease. N Engl J Med 1983;309:385–389.[Abstract]
  5. Harris N, Galpchian V, Rifai N. Three routine methods for measuring high-density lipoprotein cholesterol compared with the reference method. Clin Chem 1996;42:738–743.[Abstract/Free Full Text]
  6. Warnick GR, Benderson JM, Albers JJ. Quantitation of high-density lipoprotein subclasses after separation by dextran sulfate and Mg2+ precipitation (abstract). Clin Chem 1982;28:1574.
  7. Maciejko JJ, Levinson SS. Markyvech L, Smith M. Evaluation of a new method for assaying apo A-I and apo B lipoproteins by rate nephelometry. Clin Chem 1987;11:2065–2069.
  8. Levinson SS, Hobbs GA. Optimized automated apolipoprotein A-I assays as markers for coronary artery disease. Arch Path Lab Med 1997;121:678–684.
  9. Austin MA. Plasma triglycerides and coronary heart disease. Arterioscler Thromb 1991;11:2–11.[Abstract/Free Full Text]
  10. Zweig MN, Broste SK, Reinhart RA. ROC curve analysis: an example showing the relationships among serum lipid and apolipoprotein concentrations in identifying patients with coronary artery disease. Clin Chem 1992;38:1425–1428.[Abstract/Free Full Text]
  11. Ross R. Atherosclerosis - an inflammatory disease. N Engl J Med 1999;340:115–126.[Free Full Text]
  12. Zaman AG, Helft G, Worthley SG, Badimon JJ. The role of plaque rupture and thrombosis in coronary artery disease. Atherosclerosis 2000;149:251–266.[Medline]
  13. Herz J, Strickland DK. LRP: a multifunctional scavenger and signaling receptor. J Clin Invest 2001;108:779–784.[Medline]
  14. Krieger M. Scavenger receptor class B type I is a multiligand HDL receptor that influences diverse physiologic systems. J Clin Invest 2001;108:793–797.[Medline]
  15. Febbraio M, Hajjar DP, Silverstein RL. CD36: a class B scavenger receptor involved in angiogenesis, atherosclerosis, inflammation, and lipid metabolism. J Clin Invest 2001;108:785–791.[Medline]
  16. Steinberg D, Parthasarathy S, Carew TE, Khoo JC, Witzum JL. Beyond cholesterol. N Engl J Med 1989;320:915–924.[Medline]
  17. Witzum JL, Steinberg D. Role of oxidized low density lipoprotein in atherosclerosis. J Clin Invest 1991;88:1785–1792.
  18. Brouillette CG, Anantharamaiah GM, Engler JA, Borhani DW. Structural models of human apolipoprotein A-I: a critical analysis and review. Biochim Biophys Acta 2001;1531:4–46.[Medline]
  19. von Eckardstein A, Nofer JR, Assmann G. High density lipoproteins and arteriosclerosis. Role of cholesterol efflux and reverse cholesterol transport. Arterioscler Thromb Vasc Biol. 2001;21:13–27.[Abstract/Free Full Text]
  20. Francone OL, Fielding CJ. Initial steps in reverse cholesterol transport: the role of short-lived cholesterol acceptors. European Heart J 1990;11(supp E):218–224.
  21. Fielding CJ, Feilding PE. Molecular physiology of reverse cholesterol transport. J Lipid Res 1995;36: 211–228.[Abstract]
  22. Barbaras R, Puchois P, Fruchart J, Ailhaud G. Cholesterol efflux from cultured adipose cells is mediated by LpAI Particles but not by LpAI:AII particles. Biochem Biophys Res Commun 1987;142:63–69.[Medline]
  23. Stein O, Stein Y. Atheroprotective mechanisms of HDL. Atherosclerosis 1999;144:285–301.[Medline]
  24. Gofman JW, de Lalla O, Glazier F, Freeman NK, Lindgren FT, et al. The serum lipoprotein transport system in health, metabolic disorders, atherosclerosis, and coronary artery disease. Plasma 1954;2:413–484.
  25. Gotto AM. Low high-density lipoprotein cholesterol as a risk factor in coronary heart disease. A working group report. Circulation 2001;103:2213–2218.[Free Full Text]
  26. Asztalos BF, Brousseau ME, McNamara JR, Horvath KV, Roheim PS, Schaefer EJ. Subpopulations of high density lipoproteins in homozygous and heterozygous Tangier disease. Atherosclerosis 2001;156: 217–225.[Medline]
  27. Tall AR, Wang N. Tangier disease as a test of the reverse cholesterol transport hypothesis. J Clin Invest 2000;106:1205–1207.[Medline]
  28. Oram JH, Yokoyama S. Apolipoprotein-mediated removal of cellular cholesterol and phospholipids. J Lipid Res 1996;37:2473–2491.[Abstract]
  29. Bortnick AE, Rothblat GH, Stoudt G, Hoppe KL, Royer LJ, McNeish J, Francone OL. The correlation of ATP-binding cassette-1 mRNA levels with cholesterol efflux from various cell lines. J Biol Chem 2000;275:28634–28640.[Abstract/Free Full Text]
  30. Frank PG, Marcel YL. Apolipoprotein A-I: structure-function relationships. J Lipid Res 2000;41:853–872.[Abstract/Free Full Text]
  31. Tall AR, Jiang X, Luo Y, Silver D. 1999 George Lyman Duff memorial lecture: lipid transfer proteins, HDL metabolism, and atherogenesis. Arterioscler Thromb Vasc Biol 2000:20;1185–1188.[Abstract/Free Full Text]
  32. Mahley RW, Ji ZS. Remnant lipoprotein metabolism: key pathways involving cell-surface heparin sulfate proteoglycans and apolipoprotein E. J Lipid Res 1999;40:1–16.[Abstract/Free Full Text]
  33. Huuskonen J, Olkkonen VM, Jauhiainen M, Ehnholm C. The impact of phospholipid transfer protein (PLTP) on HDL metabolism. Atherosclerosis 2001;155:269–281.[Medline]
  34. Mardones P, Quinones V, Amigo L, Moreno M, Miquel JF, Schwarz M, Miettinen HE, Trigatti B, Krieger M, VanPatten S, Cohen DE, Rigotti A. Hepatic cholesterol and bile acid metabolism and intestinal cholesterol absorption in scavenger receptor class B type I-deficient mice. J Lipid Res 2001;42: 170–180.[Abstract/Free Full Text]
  35. Eisenberg S. High density lipoprotein metabolism. J Lipid Res 1984;25:1017–1057.[Medline]
  36. Gavish D, Oschry Y, Eisenberg S. In vivo conversion of human HDL3 to HDL2 and apo E-rich HDL in the rat:effects of lipid transfer protein. J Lipid Res 1987;28:257–267.[Abstract]
  37. Miller NE, Hammett F, Saltissi S, Roa S, et al. Relationship of angiographically defined coronary artery disease to plasma lipoprotein subfractions and apolipoproteins. Brit Med J 1981;282:1741–1744.
  38. Noma A, Yokosuka T, Kitamura K. Plasma lipids and apolipoproteins as discriminators for the presence and severity of angiographically defined coronary artery disease. Atherosclerosis 1983;49:1–7.[Medline]
  39. Naito H. The association of serum lipids, lipoproteins, and apolipoproteins with coronary artery disease assessed by angiography. Ann NY Acad Sci 1985;454:230–238.[Medline]
  40. Drexel H, Amann FW, Rentsch K, Neuenschwander C, Luethy A, Khan SI, Follath F. Relationship of the level of high-density lipoprotein subfractions to the presence and extent of coronary artery disease. Am J Cardiol 1992;70:436–440.[Medline]
  41. Salonen JT, Salonen R, Seppanen K, Rauramaa R, Tuomilehto J. HDL, HDL2, and HDL3 subfractions, and the risk of acute myocardial infarction. A prospective population study in eastern Finnish men. Circulation 1991;84:129–139.[Abstract/Free Full Text]
  42. Buring JE, O’Connor GT, Goldhaber SZ, Rosner B, Herbert PN, et al. Decreased HDL2 and HDL3 cholesterol, apo A-I and apo A-II, and increased risk of myocardial infarction. Circulation 1992;85:22–29.[Abstract/Free Full Text]
  43. Stamper MJ, Sacks FM, Salvini S, Willett WC, Hennekens CH. A prospective study of cholesterol, apolipoproteins, and the risk of myocardial infarction. N Engl J Med 1991;325:373–381.[Abstract]
  44. Sweetnam PM, Bolton CH, Yarnell JW, Bainton D, Baker IA, Elwood PC, Miller NE. Associations of the HDL2 and HDL3 cholesterol subfractions with the development of ischemic heart disease in British men. Circulation 1994;90:769–774.[Abstract/Free Full Text]
  45. Levinson SS, Wagner S. The relative usefulness of automated apolipoprotein A-I and high density lipoprotein cholesterol assays as markers for coronary artery disease. Am J Clin Pathol 1994;101:104–108.[Medline]
  46. Marcovia SM, Albers JJ, Henderson LO, Hannon WH. International federation of clinical chemistry standardization project for measurement of apolipoproteins A-I and B. Comparability of apolipoprotein A-I values by use of international reference material. Clin Chem 1993;39:773–781.[Abstract/Free Full Text]
  47. Kweiterovich PO Jr, Coresh J, Smith HH, Bachorik PS, Derby CA, Pearson TA. Comparison of the plasma levels of apolipoproteins B and A1, and other risk factors in men and women with premature coronary artery disease. Am J Cardiol 1992;69:1015–1021.[Medline]
  48. Hamsten A, Walldius G, Szamoski A, Dahlen G, DE Faire U. Relationship of angiographically defined coronary artery disease to serum lipoproteins and apoproteins in young survivors of myocardial infarction. Circulation 1986;73:1097–1110.[Abstract/Free Full Text]
  49. Johansson S, Bondjers G, Fager G, Wedel H, et al. Serum lipids and apolipoprotein levels in women with acute myocardial infarction. Arteriosclerosis 1988;8:742–749.[Abstract/Free Full Text]
  50. Mao SJT, Miller J., Gotto Jr AM, Sparrow JT. The antigenic structure of apolipoprotein AI in high density lipoproteins. J Biol Chem 1980;255:3448–3453.[Free Full Text]
  51. Maciejko JJ, Mao STJ. Radioimmunoassay of apolipoprotein A-I. Application of a non-ionic detergent (Tween-20) and solid-phase staphylococcus. Clin Chem 1982;28:199–204.[Abstract/Free Full Text]
  52. Rothwell TC, Kamanna VS, Jin F, Koren E, Foley T, Kashyap ML. Characterization of a monoclonal antibody (HB-22) and development of an ELISA for human apolipoprotein A-I. Clin Chem 1995;41: 1150–1158.[Abstract/Free Full Text]
  53. van Lennep JE, Westerveld HT, van Lennep HW, Zwinderman AH, Erkelens DW, van der Wall EE. Apolipoprotein concentrations during treatment and recurrent coronary artery disease events. Arterioscler Thromb Vasc Biol 2000;20:2408–2413.[Abstract/Free Full Text]
  54. Francis MC, Frohlich JJ. Coronary artery disease in patients at low risk: apolipoprotein AI as an independent risk factor. Atherosclerosis 2001;155: 165–170.[Medline]
  55. Duriez P, Fruchart JC. High-density lipoprotein subclasses and apolipoprotein A-I. Clin Chim Acta 1999;286:97–114.[Medline]
  56. Wald NJ, Law M, Watt HC, Wu T, Bailey A, Johnson AM, Craig WY, Ledue TB, Haddow JE. Apolipoproteins and ischaemic heart disease: implications for screening. Lancet 1994;343:75–79.[Medline]
  57. O’Connell BJ, Genest, Jr J. High density lipoproteins and endothelial function. Circulation 2001; 104:1978–83.[Abstract/Free Full Text]
  58. Mackness MI, Abbott C, Arrol S, Durrington PN. The role of high-density lipoprotein and lipid-soluble antioxidant vitamins in inhibiting low-density lipoprotein oxidation. Biochem J 1993;294: 829–834.
  59. Durrington PN, Mackness B, Mackness MI. Paraoxonase and atherosclerosis. Arterioscler Thromb Vasc Biol 2001;21:473–480.[Abstract/Free Full Text]
  60. Mackness B, Davies GK, Turkie W, Lee E, Roberts DH, Hill E, Roberts C, Durrington PN, Mackness MI. Paraoxonase status in coronary heart disease: are activity and concentration more important than genotype? Arterioscler Thromb Vasc Biol 2001;21: 1451–1457.[Abstract/Free Full Text]
  61. Tall AR, Wang N, Mucksavage P. Is it time to modify the reverse cholesterol transport model? J Clin Invest 2001;108:1273–1275.[Medline]
  62. Grover SA, Coupal L, Xiao-Ping H. Identifying adults at increased risk of coronary artery disease. JAMA 1995;274:801–806.[Abstract/Free Full Text]
  63. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the Third Report (Adult Treatment Panel III). JAMA 2001;285:2486–2497.[Free Full Text]
  64. Havel RJ, Rapaport E. Management of primary hyperlipidemia. N Engl J Med 1995;332:1491–1498.[Free Full Text]
  65. Sniderman AD, Clanflone K. Measurement of apolipoproteins: time to improve the diagnosis and treatment of atherogenic dyslipoproteinemias. Clin Chem 1996;42:489–491.[Free Full Text]
  66. Lamarche B, Tchernof A, Mauiége P, Cantin B, Dagenais GL, et al. Fasting insulin and apolipoprotein B levels and low-density lipoprotein particle size as risk factors for ischemic heart disease. JAMA 1998;279:1955–1961.[Abstract/Free Full Text]
  67. Vorchheimer DA, Fuster V. Inflammatory markers in coronary artery disease. JAMA 2001;286:2154–2156.[Free Full Text]
  68. Libby P, Ridker PM. Novel inflammatory markers of coronary risk: theory versus practice. Circulation 1999;100:1148–1150.[Free Full Text]
  69. Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive proteins and other markers of inflammation in prediction of cardiovascular disease in women. N Engl J Med 2000;342:836–843.[Abstract/Free Full Text]
  70. Tracy RP. Inflammation in cardiovascular disease: cart, horse, or both (Editorial). Circulation 1998;97: 2000–2002.[Free Full Text]
  71. Rifai N, Ridker PM. Proposed cardiovascular risk assessment algorithm using high sensitivity C-reactive protein and lipid screening. Clin Chem 2001;47:28–30.[Free Full Text]



This article has been cited by other articles:


Home page
 J Am Coll CardiolHome page
D. N. Tziakas, J. C. Kaski, G. K. Chalikias, C. Romero, S. Fredericks, I. K. Tentes, A. X. Kortsaris, D. I. Hatseras, and D. W. Holt
Total Cholesterol Content of Erythrocyte Membranes Is Increased in Patients With Acute Coronary Syndrome: A New Marker of Clinical Instability?
J. Am. Coll. Cardiol., May 29, 2007; 49(21): 2081 - 2089.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Levinson, S. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Levinson, S. S.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS