ARTHRITIS & RHEUMATISM Vol. 54, No. 8, August 2006, pp 25412549 DOI 10.1002/art.

21976 2006, American College of Rheumatology

Proinflammatory High-Density Lipoprotein as a Biomarker for Atherosclerosis in Patients With Systemic Lupus Erythematosus and Rheumatoid Arthritis
Maureen McMahon,1 Jennifer Grossman,1 John FitzGerald,1 Erika Dahlin-Lee,1 Daniel J. Wallace,2 Bernard Y. Thong,3 Humeira Badsha,3 Kenneth Kalunian,4 Christina Charles,1 Mohamad Navab,1 Alan M. Fogelman,1 and Bevra H. Hahn1
Objective. Women with systemic lupus erythematosus (SLE) have a 750-fold increased risk of coronary artery disease (CAD). In the general population, oxidized low-density lipoprotein (ox-LDL) increases the risk for CAD. Normal high-density lipoproteins (HDLs) protect LDL from oxidation; proinflammatory HDLs do not. This study was undertaken to determine whether patients with SLE, who have chronic inflammation that causes oxidative damage, have more proinflammatory HDL and higher levels of ox-LDL, thus predisposing them to atherosclerosis. Methods. One hundred fifty-four women with SLE, 48 women with rheumatoid arthritis (RA), and 72 healthy controls were studied. The ability of the patients HDL to prevent oxidation of normal LDL was measured. Values >1.0 (the value assigned for LDL oxidation in the absence of HDL) after the addition of HDL indicated proinflammatory HDL. Plasma ox-LDL levels were measured as the amount of oxidation produced by the patients LDL after the removal of HDL.
Dr. McMahons work was supported by grants from the American College of Rheumatology Research and Education Foundation and the Lupus Research Institute. Drs. Navab and Fogelmans work was supported by Bruin Pharmaceuticals. Dr. Hahns work was supported by the Lupus Research Institute and by a Kirkland Award. 1 Maureen McMahon, MD, Jennifer Grossman, MD, John FitzGerald, MD, PhD, Erika Dahlin-Lee, AB, Christina Charles, MD, Mohamad Navab, PhD, Alan M. Fogelman, MD, Bevra H. Hahn, MD: University of California at Los Angeles David Geffen School of Medicine; 2Daniel J. Wallace, MD: Cedars Sinai Medical Center, Los Angeles, California; 3Bernard Y. Thong, MD, Humeira Badsha, MD: Tan Tock Seng Hospital, Singapore; 4Kenneth Kalunian, MD: University of California at San Diego, La Jolla. Address correspondence and reprint requests to Maureen McMahon, MD, Division of Rheumatology, UCLA David Geffen School of Medicine, 1000 Veteran Avenue 32-59, Los Angeles, CA 90095. E-mail: mmcmahon@mednet.ucla.edu. Submitted for publication August 26, 2005; accepted in revised form April 13, 2006. 2541

Results. SLE patients had more proinflammatory HDL (mean SD score 1.02 0.57, versus 0.68 0.28 in controls [P < 0.0001] and 0.81 0.22 in RA patients [P 0.001 versus SLE patients]). A higher proportion of SLE patients had proinflammatory HDL: 44.7% of SLE patients versus 4.1% of controls and 20.1% of RA patients had scores >1.0 (P < 0.006 between all groups). Levels of ox-LDL correlated with levels of proinflammatory HDL (r 0.37, P < 0.001). SLE patients with CAD had significantly higher proinflammatory HDL scores than patients without CAD (P < 0.001). Conclusion. HDLs are proinflammatory in a significant proportion of SLE patients and are associated with elevated levels of ox-LDL. Abnormal HDLs impair the ability to prevent LDL oxidation and may predispose to atherosclerosis. Premature atherosclerosis is a major comorbid condition in patients with systemic lupus erythematosus (SLE). Young women with SLE have an estimated 50-fold increased risk of myocardial infarction compared with age- and sex-matched controls (1). While SLE patients are subject to the same traditional risk factors as the general population (24), these factors do not adequately account for the significantly increased level of cardiovascular disease. Thus, there is a continuing search for new biomarkers of increased risk for atherosclerotic disease in SLE patients. One possible biomarker is proinflammatory high-density lipoprotein (HDL). Proinflammatory HDL is unable to perform its usual protective role in the prevention of atherosclerosis. Atherosclerotic lesions begin when low-density lipoproteins (LDLs) are trapped in artery walls and are seeded with reactive oxygen species (ROS) (5), resulting

2542

McMAHON ET AL

Table 1. Demographic and clinical data on the patients and controls* SLE patients (n 154) Age, mean SD years Total cholesterol, mean SD mg/dl HDL, mean SD mg/dl LDL, mean SD mg/dl Triglycerides, mean SD mg/dl Ethnicity, no. (%) Caucasian Asian African American Hispanic Ox-LDL, mean SD FU 40.2 13.6 180.3 44.7 56.3 18.7 99.9 36.4 116.1 80.4 72 (46.7) 33 (21.4) 15 (9.7) 33 (21.4) 6,713.7 3,674.5 Controls (n 72) 39.9 16.3 183.2 36.3 51.8 14.9 113.1 35.9 97.6 47.1 35 (48.6) 23 (31.9) 2.8 (2) 12 (16.7) 5,356.2 2,610.0 RA patients (n 48) 54.5 12.5 196.5 41.5 63.4 22.5 110.2 31.3 128.1 72.2 33 (68.8) 2 (4.2) 3 (6.3) 10 (20.8) 6,260.6 3,494.3 P 0.001 (controls vs. RA) 0.001 (SLE vs. RA) 0.07 (SLE vs. RA) 0.007 (controls vs. RA) 0.04 (controls vs. SLE) 0.03 (controls vs. RA) 0.02 (SLE vs. RA) 0.001 (controls vs. RA) 0.001 (SLE vs. RA) NS NS 0.05 (controls vs. SLE)

* SLE systemic lupus erythematosus; RA rheumatoid arthritis; HDL high-density lipoprotein; ox-LDL oxidized low-density lipoprotein; FU fluorescence units. Only statistically significant P values are shown. NS not significant. Determined in 83 patients with SLE, 39 healthy controls, and 28 patients with RA.

in oxidized LDL (ox-LDL) phospholipids (6). After endothelial cells are exposed to these oxidized lipids, they release cytokines, which induce monocyte binding, chemotaxis, and differentiation of monocytes into macrophages (6). Ox-LDLs are phagocytized by infiltrating macrophages, leading to the formation of foam cells, a hallmark of atherosclerotic lesions (7). Normal HDL removes ROS from LDL, preventing both the oxidation of LDL and the recruitment of inflammation mediators (6,8). HDLs, however, are chameleon-like lipoproteins (7) with the capacity to be antiinflammation in the basal state and proinflammatory during acute-phase responses. Antiinflammatory HDL protects LDL from oxidation, while proinflammatory HDL does not (7). Navab et al have developed a novel assay that determines the functional properties of HDL by evaluating the ability of a subjects HDL to prevent lipid oxidation (9,10). A study using this assay in nonSLE patients with coronary artery disease (CAD) and normal lipid profiles revealed that 25 of 26 patients had proinflammatory HDL (11). HDL function improved with statin therapy, but not to normal levels (11). The idea that proinflammatory HDL is a risk factor for atherosclerosis in SLE is a novel hypothesis. In this study, we set out to determine if levels of proinflammatory HDL differ between individuals with chronic inflammatory conditions (SLE and rheumatoid arthritis [RA]) and healthy controls. We also investigated whether any traditional cardiac risk factors or diseasespecific factors are associated with the presence of proinflammatory HDL. We hypothesize that patients with SLE, who have ongoing oxidative damage due to

chronic inflammation, have proinflammatory HDL and high levels of ox-LDL. PATIENTS AND METHODS
Study population. Three groups of subjects were studied: pre-exclusion numbers were 161 SLE patients, 60 RA patients, and 80 healthy controls. The lupus patients are patients at the University of California, Los Angeles (UCLA), Cedars Sinai Medical Center in Los Angeles, and Tan Tock Seng Hospital in Singapore. All patients fulfilled at least 4 of the 1997 revised American College of Rheumatology (ACR; formerly, the American Rheumatism Association) classification criteria for SLE (12). The controls were women, healthy by self-report, with no clinical manifestations of SLE. The RA patients were patients attending UCLA clinics who met the ACR criteria for RA (13). Patients were excluded if they had taken any lipid-lowering agents within the previous 3 months. After exclusions for statin use, the final group consisted of 154 SLE patients, 48 RA patients, and 72 controls. Clinical data obtained on patients with SLE and RA included information regarding disease state, cardiac risk factors, and medication use (Tables 1 and 2). Because C-reactive protein (CRP) measurements were not available for many patients, these data are not included in the analysis. Subjects were included in the study after providing written consent. The study was reviewed and approved by the Institutional Review Board of UCLA. Laboratory and clinical assessments. Blood samples were collected and processed by sucrose cryopreservation, and HDL was isolated from plasma as previously described (11). Tests to measure plasma lipid concentrations, erythrocyte sedimentation rate (ESR), serum complement levels, and autoantibodies against DNA and cardiolipin were performed in our clinical laboratory using standard methods. On the day of serum sampling, the disease activity of SLE patients was assessed using the Safety of Estrogens in Lupus Erythematosus: National Assessment (SELENA) version of the Systemic

INFLAMMATORY HDL AS A BIOMARKER FOR ATHEROSCLEROSIS IN SLE AND RA

2543

Table 2.

Disease characteristics and traditional cardiovascular risk factors in SLE and RA patients* Characteristic SLE patients (n 154) 4 (2.7) 12 (7.8) 47 (30.5) 6 (3.9) 27 (17.5) 12 (7.8) 46 (30.0) 10.1 8.7 4.5 5.3 1.5 1.8 87 (56.5) 35 (22.7) 60 (38.9) 154 (100) 22.0 19.5 102.3 31.4 27 (17.5) 89 (57.8) 9 (5.8) 7 (4.5) 21 (13.6) 0 7.85 15.8 80 (51.9) 25 (16.7) 45 (29.2) RA patients (n 48) 0 0 13 (27.1) 3 (6.3) 9 (18.8) 3 (6.3) 0 13.5 11.1 0 0 0 0 0 0 20.0 15.6 NA 0 7 (14.6) 0 25 (52.1) 3 (6.3) 30 (62.5) 3.17 9.04 36 (75) 4 (8.3) 7 (14.6)

History of coronary artery disease, no. (%) History of cerebrovascular events, no. (%) History of hypertension, no. (%) History of diabetes, no. (%) Ever smoked, no. (%) Currently smoking, no. (%)# History of glomerulonephritis, no. (%) Disease duration, mean SD years SELENA-SLEDAI score, mean SD SDI score, mean SD History of dsDNA antibody positivity, no. (%) History of LAC positivity, no. (%) History of aCL positivity (IgG, IgM, IgA), no. (%) ANA positive, no. (%) ESR, mean SD mm/hour** C3 level, mean SD mg/dl Current medications Mycophenolate mofetil, no. (%) Hydroxychloroquine, no. (%) Cyclophosphamide, no. (%) Methotrexate, no. (%) Azathioprine, no. (%) Anti-TNF agents, no. (%) Current prednisone dosage, mean SD mg/day Cumulative lifetime prednisone dose, no. (%) 10 gm 1020 gm 20 gm

* SLE systemic lupus erythematosus; RA rheumatoid arthritis; SELENA-SLEDAI Safety of Estrogens in Lupus Erythematosus: National Assessment version of the Systemic Lupus Erythematosus Disease Activity Index; SDI Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; dsDNA double-stranded DNA; LAC lupus anticoagulant; aCL anticardiolipin antibody; ANA antinuclear antibody; ESR erythrocyte sedimentation rate; NA not available; anti-TNF antitumor necrosis factor . Defined as a history of myocardial infarction, coronary artery disease documented on an angiogram or stress test, or angina. Cerebrovascular events included transient ischemic attacks and stroke. Defined as use of antihypertensive medication or systolic blood pressure 140 mm Hg or diastolic blood pressure 90 mm Hg. Defined as a fasting glucose level 7.0 mmoles/liter (126 mg/dl), or treatment with insulin or an oral hypoglycemic agent. # Patients were defined as smokers if they had smoked any cigarettes within the previous 3 months. ** Data available for 142 SLE and 47 RA patients. Adalimumab, etanercept, or infliximab.

Lupus Erythematosus Disease Activity Index (SLEDAI) (14). Organ damage was determined using the Systemic Lupus International Collaborating Clinics/ACR Damage Index (SDI) (15). Cell-free assay. The cell-free assay is a modification of a previously described method using LDL as the fluorescenceinducing agent (10). Normal HDL prevents oxidation of LDL and therefore the oxidation of dichlorofluorescein (DCFH), which releases a fluorochrome upon oxidation. To determine the functional properties of HDL, the change in fluorescence intensity resulting from oxidation of DCFH by LDL in the presence or absence of test HDL was measured. LDL was prepared from normal plasma as previously described (10). Twenty microliters of the normal LDL solution (final concentration 50 g/ml) and 90 l of test HDL (at a final concentra-

tion of 10 g/ml cholesterol) were incubated in 96-well plates for 1 hour. Ten microliters of DCFH solution (0.2 mg/ml) was then added to each well and incubated for 2 hours. Fluorescence was determined using a plate reader (Spectra Max, Gemini XS; Molecular Devices, Sunnyvale, CA). Values of DCFH activated by LDL in the absence of HDL (in fluorescence units [FU]) were normalized to 1.0 as the positive control. Values 1.0 FU after the addition of test HDL indicated proinflammatory HDL; values 1.0 indicated antiinflammatory HDL. Mean SD values for intra- and interassay variability were 5.3 1.7% and 7.1 3.2%, respectively. Oxidized LDL. Oxidized LDL was measured in an unselected subset of 150 of the patients from all 3 groups, by adding 200 l of DCFH solution (0.2 mg/ml) to the apolipo-

2544

McMAHON ET AL

protein B (Apo B)/LDLcontaining fraction of the patients plasma. This solution (90 l) was incubated for 2 hours in 96-well plates. Fluorescence was determined using a plate reader. In addition, ox-LDL was measured in 38 SLE patients or control subjects by sandwich enzyme-linked immunosorbent assay (ELISA; Mercodia, Uppsala, Sweden), using the monoclonal antibody 4E6. Statistical analysis. Results are expressed as the mean SD. Data were analyzed using SPSS 13.0 software (SPSS, Chicago, IL). Skewed continuous variables were logarithmically transformed to attain a normal distribution. For variables that did not attain a normal distribution by logarithmic transformation, nonparametric tests were used. Study groups were compared using analysis of variance/Students t-test for continuous variables and the chi-square test for categorical variables. Correlation coefficients were calculated using simple regression, or Spearmans rank correlation for abnormally distributed data. Logistic regression was used to build models that identify risk factors associated with the presence of proinflammatory HDL in SLE and RA patients. P values less than 0.05 were considered significant.

RESULTS Association of proinflammatory HDL with SLE and RA. HDL samples were assessed for their pro- and antiinflammatory properties using the cell-free assay. HDL from SLE patients was more proinflammatory than HDL from controls, with a mean SD score of 1.02 0.57 in the SLE patients versus 0.68 0.28 in the controls (P 0.0001). HDL from RA patients was also more inflammatory than HDL from controls, but less than the HDL from SLE patients, with a mean SD score of 0.81 0.22 (P 0.016 versus controls, P 0.001 versus SLE patients) (Figure 1A). Using a threshold of 1.0 as a cutoff to define proinflammatory HDL, 44.7% of SLE patients versus 20.1% of RA patients versus 4.1% of controls had proinflammatory HDL (P 0.006 between all groups, P 0.0001 between SLE patients and controls) (Figure 1B). Association of proinflammatory HDL with CAD. Fourteen patients in the cohort, all in the SLE group, had a history of documented clinical atherosclerosis. Four patients had a history of CAD, 12 had a history of cerebrovascular disease, and 2 patients had both. All 4 patients with CAD had proinflammatory HDL, with a mean SD score of 1.11 0.07, compared with 0.80 0.43 in patients without CAD (P 0.0001). Patients who had had a stroke had a mean SD score of 0.97 0.40, versus 0.80 0.43 (P 0.13) in patients who had not, and half (6 of 12) had proinflammatory HDL. Overall, the 14 patients with a documented history of atheroscle-

Figure 1. Association of proinflammatory high-density lipoprotein (HDL) and systemic lupus erythematosus (SLE). A, Individual HDL scores among subjects in the SLE, rheumatoid arthritis (RA), and healthy control groups. Bars show the group means. P values were determined by analysis of variance with Dunnetts multiple comparison test. B, Percentage of subjects with proinflammatory HDL. Note that the proportion of subjects with proinflammatory HDL is highest in the SLE group, but the proportion in the RA group is also increased compared with controls. P values were determined by chi-square analysis.

rosis had a mean SD score of 1.01 0.33, versus 0.79 0.43 (P 0.01) in patients without atherosclerosis, with 8 patients with atherosclerosis (57%) having HDL in the proinflammatory range. In contrast, only 43.4% of the 140 patients without atherosclerotic events had a score 1.0 (Figure 2A). Thus, in the SLE patients, the presence of an atherosclerotic event increased the likelihood of having proinflammatory HDL.

INFLAMMATORY HDL AS A BIOMARKER FOR ATHEROSCLEROSIS IN SLE AND RA

2545

Figure 2. Association of HDL function and traditional and disease risk factors. A, Association of HDL function and history of cardiovascular (CV) disease. Values are the mean and SD in 14 subjects. P values were determined by Students t-test. B, Association of HDL function and traditional risk factors for cardiac disease. Patients taking lipid-lowering medications were excluded from this study. Values are the mean and SD in 14 subjects. P values were determined by Students t-test. C, Correlation of HDL function and levels of oxidized lowdensity lipoprotein (LDL), determined by Spearmans rank test. Data were analyzed in a subset of 83 SLE patients, 28 RA patients, and 39 healthy controls. D, Correlation of HDL function and erythrocyte sedimentation rate (ESR; measured using the Westergren method), as determined by Spearmans rank test. CAD coronary artery disease; CVA cerebrovascular accident; HTN hypertension; FU fluorescence units (see Figure 1 for other definitions).

Association of proinflammatory HDL with traditional cardiac risk factors. The relationship between traditional cardiac risk factors and HDL function in SLE and RA patients was examined in a bivariate model. There was significant association with a history of hypertension (P 0.0001) and with nonwhite race (P 0.003). There was no association with a history of smoking (currently or ever) or with diabetes (Figure 2B). There was also no association between HDL function and levels of total cholesterol, HDL cholesterol, LDL cholesterol, or triglycerides, or age. We performed logistic regression to examine the relationship of SLE and RA with proinflammatory HDL, controlling for traditional cardiac risk factors. After controlling for these risk factors, the odds ratio (OR) for proinflammatory HDL was 19.3 (95% confidence interval [95% CI] 4.485.3) in SLE patients and 6.18 (95% CI 1.232.1) in RA patients compared with controls (Table 3).

Correlation with ox-LDL. The relationship between HDL function and levels of ox-LDL was examined in SLE, RA, and control subjects. There was a positive correlation between HDL function and levels of ox-LDL, using the fluorescence assay method to detect ox-LDL (r 0.37, P 0.0001) (Figure 2C). This correlation was confirmed in a subset of patients in whom levels of ox-LDL were measured by ELISA (r 0.47, P 0.003) (data not shown). The correlations between HDL function and ox-LDL were similar in each of the 3 patient groups: SLE (r 0.42, P 0.0001), controls (r 0.364, P 0.021), and RA (r 0.355, P 0.064). Correlation with markers of disease activity. The relationships between HDL function and markers of disease activity were also examined in SLE and RA patients. There was a positive correlation between HDL function and ESR (mm/hour) (r 0.32, P 0.001) (Figure 2D). Among SLE patients, however, there were no other relationships between HDL function and measures of disease activity or damage, such as positivity (ever) for antidouble-stranded DNA antibody, lupus anticoagulant (LAC), or anticardiolipin antibody (aCL), a history of antiphospholipid antibody (aPL) syndrome (with a history of arterial or venous thrombosis), or glomerulonephritis. Additionally, there were no significant correlations between HDL function and SELENASLEDAI score (r 0.11, P 0.07), SDI score, C3 levels, or disease duration at the time of blood draw. There were also no associations between HDL function and most current treatment regimens, including methotrexate, azathioprine, mycophenolate mofetil, tumor necrosis factor inhibitors, or cyclophosphamide, in RA or SLE patients. There was, however, a statistically significant association between HDL function and

Table 3. Logistic regression of the relationship of SLE and RA with proinflammatory HDL, controlling for traditional cardiac risk factors* Variable SLE (yes, no) RA (yes, no) Age (years) Hypertension (yes, no) Diabetes (yes, no) Smoking (ever, never) Nonwhite (yes, no) HDL (mg/dl) LDL (mg/dl) OR 19.30 6.18 1.01 1.78 2.64 1.28 3.04 1.01 1.00 95% CI 4.3785.26 1.1932.12 0.991.04 0.883.61 0.5213.37 0.562.93 1.585.83 0.991.02 0.991.00 P 0.001 0.03 0.26 0.11 0.24 0.55 0.001 0.54 0.34

* Proinflammatory high-density lipoprotein (HDL) was defined as a score of 1.0 fluorescence units in the cell-free assay. SLE systemic lupus erythematosus; RA rheumatoid arthritis; OR odds ratio; 95% CI 95% confidence interval; LDL low-density lipoprotein.

2546

McMAHON ET AL

Table 4.

Logistic regression of variables associated with proinflammatory HDL in women with SLE* Variable OR 3.73 2.46 2.95 2.82 1.15 0.96 1.31 0.84 0.98 1.00 1.43 0.48 95% CI 0.87515.9 1.523.97 1.078.18 1.196.68 0.671.98 0.881.06 0.453.84 0.651.09 0.921.04 0.981.01 0.593.45 0.181.31 P 0.08 0.001 0.04 0.02 0.62 0.42 0.62 0.18 0.48 0.76 0.43 0.15

Age (years) ESR (mm/hour) Current prednisone dosage 7.5 mg/day (yes, no) Nonwhite (yes, no) Cumulative lifetime prednisone dose (gm) SELENA-SLEDAI score History of lupus glomerulonephritis (yes, no) SDI score Disease duration (years) C3 level (mm/dl) Anticardiolipin antibody (yes, no) Anti-dsDNA antibody (yes, no)

* Proinflammatory high-density lipoprotein (HDL) was defined as a score of 1.0 fluorescence unit in the cell-free assay. OR odds ratio; 95% CI 95% confidence interval; ESR erythrocyte sedimentation rate; SELENA-SLEDAI Safety of Estrogens in Lupus Erythematosus; National Assessment version of the Systemic Lupus Erythematosus Disease Activity Index; SDI Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; dsDNA double-stranded DNA.

current prednisone dosage; HDL function was more proinflammatory in those currently receiving prednisone at a dosage of 7.5 mg/day (mean SD 1.03 0.52 versus 0.83 0.49) (P 0.002). There was also a relationship between cumulative prednisone dose and HDL scores; i.e., SLE and RA patients who had received a lifetime prednisone dose of 10 gm had lower mean HDL scores when compared with those who had received 1020 gm prednisone (0.81 0.36 versus 1.07 0.42) (P 0.006). Patients who had received a lifetime prednisone dose of 10 gm had a mean SD score of 0.94 0.41, which was not significantly different from scores in the other 2 groups (data not shown). Multivariate analysis was performed to determine which variables were most consistently associated with proinflammatory HDL in SLE patients. In this analysis, the only significant factors were ESR (OR 2.46, 95% CI 1.523.97) (P 0.001), current prednisone dosage 7.5 mg/day (OR 2.95, 95% CI 1.078.18) (P 0.04), and nonwhite race (OR 2.82, 95% CI 1.196.68) (P 0.02) (Table 4). When RA patients were included in the model, ESR, nonwhite race, and current prednisone dosage all remained significant. Stability over time. We obtained a second blood draw from a convenience sample of 11 control subjects and 14 SLE patients at least 6 months after the initial sampling, to measure the stability of HDL function over time. The overall mean SD HDL score in the group was 0.82 0.31 at time 2, versus 0.83 0.28 at time 1. The correlation between the 2 time points was excellent (r 0.88, P 0.001). The SELENA-SLEDAI score changed by 2 units in 10 of the 14 SLE patients; there

was no correlation between the change in the SELENASLEDAI score from time 1 to time 2 and the change in HDL function score. In only 1 patient in the SLE group did an antiinflammatory level of HDL become proinflammatory, and proinflammatory levels of HDL did not became antiinflammatory in any patient. DISCUSSION SLE and RA are associated with an increased risk of atherosclerosis, with the greater risk in the SLE group. To our knowledge, the data presented here are the first to describe proinflammatory HDL in patients with SLE or RA. In patients with no SLE or RA, these abnormal HDLs are known to correlate with CAD or CAD equivalents (11). In this cohort, all 4 patients with known CAD had proinflammatory HDL, and all had SLE. Normal antiinflammatory HDLs play several critical roles in the prevention of atherosclerosis. HDL removes and transports excess cholesterol from peripheral cells to the liver for removal from the body (1619). HDL also plays an antiinflammatory role by removing ROS from LDL, thus preventing the oxidation of LDL and the subsequent recruitment of inflammation mediators to the vessel wall subendothelial space (6,8). In addition, HDL inhibits the expression of adhesion molecules (20,21) and cytokines, such as monocyte chemoattractant protein 1, in endothelial cells (22), which prevents monocyte movement into the vessel wall. Proinflammatory HDLs are unable to prevent the oxidation of LDL and the recruitment of monocytes, and

INFLAMMATORY HDL AS A BIOMARKER FOR ATHEROSCLEROSIS IN SLE AND RA

2547

may enhance the inflammatory response (10,23). A flow chart depicting a conceptual model of the interaction between HDL and ox-LDL is available at the authors Web site (http://rheumatology.med.ucla.edu/). Links between inflammation, autoimmunity, oxidative damage, and atherosclerosis have been increasingly reported. Recently, levels of proinflammatory oxidized phospholipids have been described as a significant predictor of coronary artery stenosis in the general population (24). Our findings, that levels of both proinflammatory HDL and ox-LDL are elevated in SLE and RA patients, and that proinflammatory HDL correlates with higher levels of ox-LDL, provide further evidence of a link between autoimmunity, proinflammatory oxidized phospholipids, and atherosclerosis. Previous studies have also shown that higher levels of ox-LDL in adult (25,26) and pediatric (27) patients with SLE are associated with arterial disease. Our data suggest that impaired functional capacity of HDL may be responsible, at least in part, for the increase in LDL oxidation in SLE patients. Long-term oxidative damage may load both LDL and HDL with oxidized products, thus overwhelming normal clearance mechanisms and hindering the ability of HDL to prevent oxidation of LDL. Although other studies have demonstrated lower HDL levels in SLE patients than in controls (28), mean levels of HDL were normal in our cohort, and HDL functional capacity was independent of serum levels. HDL particles contain several components that can prevent LDL oxidation, including Apo A-I (6,18), paraoxonase (29), and platelet-activating factoracetylhydrolase (PAF-AH) (30). During periods of acute or chronic inflammation, HDLs that were previously antiinflammatory can become proinflammatory (7,31), with decreased fractions of paraoxonase, PAF-AH, and Apo A-I (32 35). Abnormal levels of several HDL components have previously been described in SLE patients compared with healthy controls, although none of these studies has demonstrated abnormal HDL function (36). Antibodies against HDL (37), Apo A-I (38), and PAF-AH (11) have also been described in SLE patients. We tested whether these antibodies play a role in HDL function by removing Ig that was isolated with HDL. Proinflammatory effects were strictly confined to the HDL particles, and were not influenced by Ig (data not shown). Therefore, although antibodies to ox-LDL and to HDL may play a protective role in atherosclerosis or induce lipid abnormalities, they are unlikely to influence HDL function in this system. We chose to exclude patients who were taking statins, because previous studies of proinflammatory

HDL in CAD patients without SLE have demonstrated that HDL function can be improved, though not normalized, by statin therapy (1,4,39). It is likely that the number of patients with CAD included in our analysis was small because most patients with atherosclerosis are currently treated with statins. Even with this small sample size, however, there was a significant increase in proinflammatory HDL in SLE patients with known CAD over those with no known CAD. There was a correlation between HDL function and ESR, which was found in both bivariate and multivariate analyses. This suggests that proinflammatory HDL could be linked to an acute-phase response. Interestingly, however, despite similar mean ESR levels between our cohorts with SLE and RA, proinflammatory HDL was seen only half as frequently in patients with RA. These data are consistent with the lower frequencies of cardiovascular events in patients with RA compared with patients with SLE (40). Proinflammatory HDL function also did not significantly correlate with disease activity, as measured by the SELENA-SLEDAI score, although the P value was borderline (0.08) in bivariate analysis. It is possible that longitudinal data measuring the area under the curve of the SELENA-SLEDAI scores would correlate more accurately with HDL function. Alternatively, the lack of correlation between SELENA-SLEDAI scores and proinflammatory HDL could mean that our cohort was not adequately powered, or that the SELENASLEDAI score is a less accurate measure of active inflammation than the ESR, or that factors other than inflammation are responsible for the correlations with proinflammatory HDL. The fact that the HDL scores were stable over time in patients with changing SELENA-SLEDAI scores and prednisone dosages also suggests that elements other than chronic inflammation may be affecting HDL function. The most likely association would be genetic risk factors, and studies to correlate certain genes or gene regions with proinflammatory HDL are currently in progress. There is some speculation that the increased risk of thrombotic and atherosclerotic events seen in patients with SLE may be due in part to a cross-reactivity between antibodies to phospholipids and ox-LDL (41). Interestingly, there was no association in our cohort between either proinflammatory HDL or levels of oxLDL and the presence of aCL, LAC, or a history of thrombosis. Although some lupus cohorts have shown a significant association between aPL and atherosclerosis (42,43), several other cohorts have shown no correlation (40,44). The association between LAC and atheroscle-

2548

McMAHON ET AL

rosis is also unclear; for example, patients who were positive for LAC in the prospective Hopkins Lupus Cohort were more likely to develop myocardial infarction than those who were negative (45). There was no association, however, with aCL, and neither LAC nor aCL was associated with subclinical atherosclerosis by carotid ultrasound (45). Thus, aPL alone are unable to fully account for the development of atherosclerosis in SLE. The lack of association between proinflammatory HDL and aPL seen in our study suggests that the mechanisms by which proinflammatory HDLs contribute to atherosclerosis differ from those of aPL. For example, it is possible that aPL contribute to an increase in cardiovascular events through thrombosis, while proinflammatory HDL contributes to the formation of atherosclerotic plaques. Interestingly, 10 of 12 SLE patients in our cohort with a history of stroke (83%) were aPL positive, compared with 2 of 4 (50%) with a history of myocardial infarction. This may also help to explain the finding that proinflammatory HDL had a weaker association with a history of stroke than with a history of CAD in our cohort. In conclusion, proinflammatory HDL is a novel biomarker for increased risk of atherosclerosis in patients with SLE and RA. This abnormality is particularly evident in the SLE patients. Other potential biomarkers in SLE include elevated levels of ox-LDL (46), antibodies against lipoprotein lipase (26), and antiox-LDL antibodies (47). Our data, along with those of others, suggest that any one or a combination of these factors will be highly predictive of risk for atherosclerosis. Targeting HDL with interventions that restore protective capacity, such as treatment with an Apo A-I mimetic peptide (48), might be considered for prevention of atherosclerosis in patients with proinflammatory HDL and high levels of ox-LDL. ACKNOWLEDGMENTS
The authors thank Betty Tsao, PhD, for providing assistance in the collection and preparation of patient samples, and Fanny Ebling, PhD, for assistance in the isolation of pure HDL from patient samples. Additionally, we thank Brian Skaggs, PhD, for his careful reading of the text and editorial suggestions.

2. 3.

4.

5.

6.

7.

8.

9.

10.

11.

12. 13.

14. 15.

16.

REFERENCES
1. Manzi S, Meilahn EN, Rairie JE, Conte CG, Medsger TA Jr, Jansen-McWilliams L, et al. Age-specific incidence rates of myocardial infarction and angina in women with systemic lupus 17.

erythematosus: comparison with the Framingham Study. Am J Epidemiol 1997;145:40815. Petri M. Hopkins Lupus Cohort: 1999 update. Rheum Dis Clin North Am 2000;26:199213. Cervera R, Khamashta MA, Font J, Sebastiani GD, Gil A, Lavilla P, et al, for the European Working Party on Systemic Lupus Erythematosus. Morbidity and mortality in systemic lupus erythematosus during a 5-year period: a multicenter prospective study of 1,000 patients. Medicine (Baltimore) 1999;78:16775. Esdaile JM, Abrahamowicz M, Grodzicky T, Li Y, Panaritis C, du Berge R, et al. Traditional Framingham risk factors fail to fully account for accelerated atherosclerosis in systemic lupus erythematosus. Arthritis Rheum 2001;44:23317. Navab M, Imes SS, Hama SY, Hough GP, Ross LA, Bork RW, et al. Monocyte transmigration induced by modification of low density lipoprotein in cocultures of human aortic wall cells is due to induction of monocyte chemotactic protein 1 synthesis and is abolished by high density lipoprotein. J Clin Invest 1991;88: 203946. Navab M, Hama SY, Cooke CJ, Anantharamaiah GM, Chaddha M, Jin L, et al. Normal high density lipoprotein inhibits three steps in the formation of mildly oxidized low density lipoprotein: step 1. J Lipid Res 2000;41:148194. Navab M, Berliner JA, Watson AD, Hama SY, Territo MC, Lusis AJ, et al. The Yin and Yang of oxidation in the development of the fatty streak: a review based on the 1994 George Lyman Duff Memorial Lecture [review]. Arterioscler Thromb Vasc Biol 1996; 16:83142. Navab M, Hama SY, Anantharamaiah GM, Hassan K, Hough GP, Watson AD, et al. Normal high density lipoprotein inhibits three steps in the formation of mildly oxidized low density lipoprotein: steps 2 and 3. J Lipid Res 2000;41:1495508. Navab M, Berliner JA, Subbanagounder G, Hama S, Lusis AJ, Castellani LW, et al. HDL and the inflammatory response induced by LDL-derived oxidized phospholipids [published erratum appears in Arterioscler Throm Vasc Biol 2001; 21:880]. Arterioscler Thromb Vasc Biol 2001;21:4818. Navab M, Hama SY, Hough GP, Subbanagounder G, Reddy ST, Fogelman AM. A cell-free assay for detecting HDL that is dysfunctional in preventing the formation of or inactivating oxidized phospholipids. J Lipid Res 2001;42:130817. Ansell BJ, Navab M, Hama S, Kamranpour N, Fonarow G, Hough G, et al. Inflammatory/antiinflammatory properties of high-density lipoprotein distinguish patients from control subjects better than high-density lipoprotein cholesterol levels and are favorably affected by simvastatin treatment. Circulation 2003;108:27516. Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus [letter]. Arthritis Rheum 1997;40:1725. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:31524. Kim MY, Buyon JP, Petri M, Skovron ML, Shore RE. Equivalence trials in SLE research: issues to consider. Lupus 1999;8:6206. Gladman D, Ginzler E, Goldsmith C, Fortin P, Liang M, Urowitz M, et al. The development and initial validation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index for systemic lupus erythematosus. Arthritis Rheum 1996;39:3639. Remaley AT, Thomas F, Stonik JA, Demosky SJ, Bark SE, Neufeld EB, et al. Synthetic amphipathic helical peptides promote lipid efflux from cells by an ABCA1-dependent and an ABCA1independent pathway. J Lipid Res 2003;44:82836. Shah PK, Yano J, Reyes O, Chyu KY, Kaul S, Bisgaier CL, et al. High-dose recombinant apolipoprotein A-I(milano) mobilizes tissue cholesterol and rapidly reduces plaque lipid and macrophage

INFLAMMATORY HDL AS A BIOMARKER FOR ATHEROSCLEROSIS IN SLE AND RA

2549

18.

19.

20.

21.

22. 23.

24. 25.

26.

27.

28.

29.

30.

31.

32.

33.

content in apolipoprotein e-deficient mice: potential implications for acute plaque stabilization. Circulation 2001;103:304750. Martinez LO, Agerholm-Larsen B, Wang N, Chen W, Tall AR. Phosphorylation of a pest sequence in ABCA1 promotes calpain degradation and is reversed by ApoA-I. J Biol Chem 2003;278: 3736874. Zhang Y, Zanotti I, Reilly MP, Glick JM, Rothblat GH, Rader DJ. Overexpression of apolipoprotein A-I promotes reverse transport of cholesterol from macrophages to feces in vivo. Circulation 2003;108:6613. Cockerill GW, Rye KA, Gamble JR, Vadas MA, Barter PJ. High-density lipoproteins inhibit cytokine-induced expression of endothelial cell adhesion molecules. Arterioscler Thromb Vasc Biol 1995;15:198794. Calabresi L, Franceschini G, Sirtori CR, de Palma A, Saresella M, Ferrante P, et al. Inhibition of VCAM-1 expression in endothelial cells by reconstituted high density lipoproteins. Biochem Biophys Res Commun 1997;238:615. Barter PJ, Baker PW, Rye KA. Effect of high-density lipoproteins on the expression of adhesion molecules in endothelial cells [review]. Curr Opin Lipidol 2002;13:2858. Van Lenten BJ, Wagner AC, Anantharamaiah GM, Garber DW, Fishbein MC, Adhikary L, et al. Influenza infection promotes macrophage traffic into arteries of mice that is prevented by D-4F, an apolipoprotein A-I mimetic peptide. Circulation 2002;106: 112732. Tsimikas S, Brilakis ES, Miller ER, McConnell JP, Lennon RJ, Kornman KS, et al. Oxidized phospholipids, Lp(a) lipoprotein, and coronary artery disease. N Engl J Med 2005;353:4657. Frostegard J, Svenungsson E, Wu R, Gunnarsson I, Lundberg IE, Klareskog L, et al. Lipid peroxidation is enhanced in patients with systemic lupus erythematosus and is associated with arterial and renal disease manifestations. Arthritis Rheum 2005;52:192200. Svenungsson E, Jensen-Urstad K, Heimburger M, Silveira A, Hamsten A, de Faire U, et al. Risk factors for cardiovascular disease in systemic lupus erythematosus. Circulation 2001;104: 188793. Posadas-Romero C, Torres-Tamayo M, Zamora-Gonzalez J, Aguilar-Herrera BE, Posada-Sanchez R, Cardoso-Saldana G, et al. High insulin levels and increased low-density lipoprotein oxidizability in pediatric patients with systemic lupus erythematosus. Arthritis Rheum 2004;50:1605. Svenungsson E, Fei GZ, Jensen-Urstad K, de Faire U, Hamsten A, Frostegard J. TNF-: a link between hypertriglyceridaemia and inflammation in SLE patients with cardiovascular disease. Lupus 2003;12:45461. Watson AD, Berliner JA, Hama SY, La Du BN, Faull KF, Fogelman AM, et al. Protective effect of high density lipoprotein associated paraoxonase: inhibition of the biological activity of minimally oxidized low density lipoprotein. J Clin Invest 1995;96: 288291. Watson AD, Navab M, Hama SY, Sevanian A, Prescott SM, Stafforini DM, et al. Effect of platelet activating factor-acetylhydrolase on the formation and action of minimally oxidized low density lipoprotein. J Clin Invest 1995;95:77482. Van Lenten BJ, Hama SY, de Beer FC, Stafforini DM, McIntyre TM, Prescott SM, et al. Anti-inflammatory HDL becomes proinflammatory during the acute phase response: loss of protective effect of HDL against LDL oxidation in aortic wall cell cocultures. J Clin Invest 1995;96:275867. Cederholm A, Svenungsson E, Stengel D, Fei GZ, Pockley AG, Ninio E, et al. Platelet-activating factor-acetylhydrolase and other novel risk and protective factors for cardiovascular disease in systemic lupus erythematosus. Arthritis Rheum 2004;50:286976. Delgado Alves J, Kumar S, Isenberg DA. Cross-reactivity between anti-cardiolipin, anti-high-density lipoprotein and anti-apolipo-

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45. 46. 47.

48.

protein A-I IgG antibodies in patients with systemic lupus erythematosus and primary antiphospholipid syndrome. Rheumatology (Oxford) 2003;42:8939. Soep JB, Mietus-Snyder M, Malloy MJ, Witztum JL, von Scheven E. Assessment of atherosclerotic risk factors and endothelial function in children and young adults with pediatric-onset systemic lupus erythematosus. Arthritis Rheum 2004;51:4517. Lahita RG, Rivkin E, Cavanagh I, Romano P. Low levels of total cholesterol, high-density lipoprotein, and apolipoprotein A1 in association with anticardiolipin antibodies in patients with systemic lupus erythematosus. Arthritis Rheum 1993;36:156674. Delgado Alves J, Ames PR, Donohue S, Stanyer L, NouroozZadeh J, Ravirajan C, et al. Antibodies to high-density lipoprotein and 2-glycoprotein I are inversely correlated with paraoxonase activity in systemic lupus erythematosus and primary antiphospholipid syndrome [published erratum appears in Arthritis Rheum 2003;48:284]. Arthritis Rheum 2002;46:268694. Dinu AR, Merrill JT, Shen C, Antonov IV, Myones BL, Lahita RG. Frequency of antibodies to the cholesterol transport protein apolipoprotein A1 in patients with SLE. Lupus 1998;7:35560. Tektonidou MG, Petrovas CA, Ioannidis JP, Vlachoyiannopoulos PG, Moutsopoulos HM. Clinical importance of antibodies against platelet activating factor in antiphospholipid syndrome manifestations. Eur J Clin Invest 2000;30:64652. Solomon DH, Karlson EW, Rimm EB, Cannuscio CC, Mandl LA, Manson JE, et al. Cardiovascular morbidity and mortality in women diagnosed with rheumatoid arthritis. Circulation 2003;107: 13037. Manzi S, Selzer F, Sutton-Tyrrell K, Fitzgerald SG, Rairie JE, Tracy RP, et al. Prevalence and risk factors of carotid plaque in women with systemic lupus erythematosus. Arthritis Rheum 1999; 42:5160. Vaarala O, Alfthan G, Jauhiainen M, Leirisalo-Repo M, Aho K, Palosuo T. Crossreaction between antibodies to oxidised lowdensity lipoprotein and to cardiolipin in systemic lupus erythematosus. Lancet 1993;341:9235. Toloza SM, Uribe AG, McGwin G Jr, Alarcon GS, Fessler BJ, Bastian HM, et al, for the LUMINA Study Group. Systemic lupus erythematosus in a multiethnic US cohort (LUMINA). XXIII. Baseline predictors of vascular events. Arthritis Rheum 2004;50: 394757. El-Magadmi M, Bodill H, Ahmad Y, Durrington PN, Mackness M, Walker M, et al. Systemic lupus erythematosus: an independent risk factor for endothelial dysfunction in women. Circulation 2004;110:399404. Buyon JP, Petri MA, Kim MY, Kalunian KC, Grossman J, Hahn BH, et al. The effect of combined estrogen and progesterone hormone replacement therapy on disease activity in systemic lupus erythematosus: a randomized trial. Ann Intern Med 2005;142: 95362. Petri M. The lupus anticoagulant is a risk factor for myocardial infarction (but not atherosclerosis): Hopkins Lupus Cohort. Thromb Res 2004;114:5935. Reichlin M, Fesmire J, Quintero-Del-Rio AI, Wolfson-Reichlin M. Autoantibodies to lipoprotein lipase and dyslipidemia in systemic lupus erythematosus. Arthritis Rheum 2002;46:295763. Navab M, Anantharamaiah GM, Hama S, Garber DW, Chaddha M, Hough G, et al. Oral administration of an Apo A-I mimetic peptide synthesized from D-amino acids dramatically reduces atherosclerosis in mice independent of plasma cholesterol. Circulation 2002;105:2902. Navab M, Anantharamaiah GM, Reddy ST, van Lenten BJ, Datta G, Garber D, et al. Human apolipoprotein A-I and A-I mimetic peptides: potential for atherosclerosis reversal. Curr Opin Lipidol 2004;15:6459.