University of Groningen

University of Groningen

The Clinical Value of HDL Function Measurements Ebtehaj, Sanam

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date:

2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Ebtehaj, S. (2019). The Clinical Value of HDL Function Measurements. University of Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

2

CHAPTER 2

HDL Cholesterol Efflux Does Not Predict Cardiovascular Risk in

Hemodialysis Patients

Chantal Kopecky¹*, Sanam Ebtehaj²*, Bernd Genser^3,4,5, Christiane Drechsler⁶, Vera Krane⁶, Marlies Antlanger¹, Johannes J. Kovarik¹, Christopher C. Kaltenecker¹, Mojtaba Parvizi⁷,

Christoph Wanner⁶, Thomas Weichhart⁸, Marcus D. Säemann¹*, and Uwe J.F. Tietge²*

1- Department of Internal Medicine III, Division of Nephrology and Dialysis, Medical University of Vienna, Vienna, Austria

2- Department of Pediatrics, Center for Liver, Digestive, and Metabolic Diseases, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

3- BGStats Consulting, Vienna, Austria

4- Mannheim Institute of Public Health, Social and Preventive Medicine, Medical Faculty Mannheim, Heidelberg University, Heidelberg, Germany

5- Instituto de Saúde Coletiva, Federal University of Bahia, Salvador, Brazil

6- Division of Nephrology, Department of Medicine 1 and Comprehensive Heart Failure Centre, University of Würzburg, Würzburg, Germany

7- Department of Pathology and Medical Biology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

8- Institute of Medical Genetics, Medical University of Vienna, Vienna, Austria

* These authors contributed equally to this study

Journal of the American Society of Nephrology. 2017 Mar; 28(3):769-775.

Chapter 2

Abstract

The cardioprotective effect of HDL is thought to be largely determined by its cholesterol efflux capacity, which was shown to inversely correlate with atherosclerotic cardiovascular disease in populations with normal kidney function. Patients with end-stage renal disease suffer an exceptionally high cardiovascular risk not fully explained by traditional risk factors. Here, we investigated in a post-hoc analysis in 1,147 type 2 diabetes mellitus patients on hemodialysis who participated in the 4D Study (The German Diabetes Dialysis Study), whether the HDL cholesterol efflux capacity is predictive for cardiovascular risk. Efflux capacity was quantified by incubating human macrophage foam cells with apolipoprotein B-depleted serum. During a median follow-up of 4.1 years N=423 patients reached the combined primary endpoint (composite of cardiac death, nonfatal myocardial infarction and stroke), N=410 experienced cardiac events and N=561 died. Strikingly, in Cox regression analyses we found no association of efflux capacity with the combined primary endpoint (hazard ratio [HR], 0.96; 95% confidence interval [CI], 0.88 – 1.06, p=0.417), cardiac events (HR, 0.92; CI, 0.83-1.02; p=0.108) or all-cause mortality (HR 0.96; 95% CI, 0.88 - 1.05; p=0.390). In conclusion, HDL cholesterol efflux capacity is not a prognostic cardiovascular risk marker in diabetic patients on hemodialysis.

Cholesterol Efflux in the 4D study

Introduction

Plasma levels of HDL cholesterol (HDL-C) are inversely correlated with the risk of atherosclerotic cardiovascular disease (CVD) in large population studies ¹, yet recent findings called the usefulness of static HDL-C measurements as predictive biomarker into question ^2-5. Rather, determining functional properties of the HDL particle represents an emerging concept in cardiovascular research ^{6, 7}. Accordingly, promoting HDL cholesterol efflux from macrophage foam cells plays a central role and is regarded crucial in reverse cholesterol transport (RCT) ^{8, 9}. In populations with normal kidney function a lower cholesterol efflux capacity was associated with increased cardiovascular morbidity and mortality, even independent of HDL-C concentrations ^{10, 11}.

End-stage renal disease (ESRD) patients represent a population with an excessively increased cardiovascular risk ¹², which is even potentiated by diabetes as a frequent comorbidity ¹³. Importantly, measurements of HDL-C levels have a rather limited predictive value in these patients ^14-16 making it conceivable that alterations in HDL function represent an underlying causative contributor to atherosclerotic risk. However, to date, the potential importance of this novel concept for ESRD patients has not been assessed. Therefore, we investigated in the present work, if cholesterol efflux as a key metric of HDL functionality is predictive for cardiovascular risk and overall mortality in ESRD patients participating in the 4D study (The German Diabetes Dialysis Study), a prospective trial exploring the efficacy of atorvastatin treatment in type 2 diabetes mellitus subjects on hemodialysis.

In this post hoc analysis of the 4D study, we measured HDL cholesterol efflux in 1147 patients, who were divided into tertiles based on cholesterol efflux capacity: first tertile, median [IQR]

0.73 [0.67–0.77]; second tertile, 0.89 [0.86-0.94]; and third tertile, 1.08 [1.02-1.20]. Baseline characteristics of the study participants according to tertiles of cholesterol efflux capacity are shown in Table 1. The proportion of men decreased significantly with increasing HDL efflux capacity, as well as the levels of the inflammatory proteins CRP and HDL-bound serum amyloid A (SAA[HDL]). Higher cholesterol efflux was paralleled by increased plasma total cholesterol, LDL cholesterol (LDL-C), HDL-C, apoA-I, albumin, apoC-III, symmetric dimethylarginine (SDMA) and carbamylated albumin. Of note, patients in the first tertile had a shorter duration of hemodialysis treatment and lower phosphate levels. There was no difference in age, body mass index, systolic blood pressure, hypertension or previous history of CVD throughout the tertiles.

We next determined clinical parameters correlated with cholesterol efflux capacity in dialysis patients (Supplemental table I) and found strong positive correlations with plasma HDL-C (r=0.246, p<0.001), apoA-I (r=0.264, p<0.001) and albumin (r=0.161, p<0.001). There were weaker correlations with LDL-C (r=0.101, p=0.001) and dialysis duration (r=0.098, p=0.001). Of note, there was no significant relation between cholesterol efflux and the inflammation markers CRP and SAA(HDL).

Chapter 2

During a mean follow-up of 4.1 years, a total of N=423 of the study participants reached the combined primary endpoint (composite of cardiac death, nonfatal myocardial infarction and stroke), N=410 experienced cardiac events (cardiac death and non-fatal myocardial infarction) and N=561 died (all-cause mortality). In an univariate Cox regression analysis examining the prognostic effect of baseline cholesterol efflux capacity for selected endpoints (Figure 1), we found no association with cholesterol efflux and either the combined primary endpoint (hazard ratio [HR] per 1 SD increase: 0.96, 95% confidence interval [CI]: 0.88–1.06, p=0.417), or all cardiac events combined (HR per 1 SD increase: 0.92, 95% CI: 0.83–1.02, p=0.108) or all-cause mortality (HR per 1 SD increase: 0.96, 95% CI: 0.88–1.05, p=0.390). Multivariate Cox regression

Table 1. Baseline characteristics of the study participants according to tertiles of cholesterol efflux capacity

Parameter Tertile 1

(n=383) Tertile 2

(n=382) Tertile 3

(n=382) p value

Cholesterol efflux capacity 0.73 [0.67–0.77] 0.89 [0.86-0.94] 1.08 [1.02-1.20]

Age, years 66.7 (8.2) 66.2 (8.4) 66.0 (8.2) 0.486

Body mass index, kg/m2 27.5 (4.7) 27.7 (5.0) 27.4 (4.8) 0.723

Duration of diabetes, years 17.8 (8.2) 18.4 (8.9) 18.1 (8.5) 0.623

Duration of dialysis, months 5.5 [3.2-10.4] 5.6 [2.6-11.8] 6.7 [3.4-12.2] 0.033

Male, n (%) 233 (61.0) 207 (54.2) 188 (49.1) 0.004

Nonsmoker, n (%) 213 (55.8) 234 (61.3) 235 (61.4) 0.411

History, n (%)

Arrhythmia 81 (21.2) 69 (18.1) 62 (16.2) 0.196

Congestive heart failure 166 (43.5) 130 (34.0) 115 (30.0) <0.001

Stroke/TIA 68 (17.8) 66 (17.3) 68 (17.8) 0.978

Peripheral vascular disease 174 (45.6) 166 (43.5) 174 (45.4) 0.808

MI/CABG/PTCA/CAD 128 (33.5) 106 (27.8) 109 (28.5) 0.166

Hypertension 336 (88.0) 341 (89.3) 341 (89.0) 0.830

Systolic blood pressure, mm Hg 144.4 (23.2) 146.5 (20.9) 146.4 (21.8) 0.324

Diastolic blood pressure, mm

Hg 74.6 (10.7) 76.3 (10.5) 76.6 (11.4) 0.024

Total cholesterol, mg/dL 209.8 (42.0) 221.5 (41.8) 226.9 (40.8) <0.001

Triglycerides, mg/dL 219.5 (149.0-345.0] 227.5 [158.0-332.0] 210.0 [139.0-298.0] 0.008

LDL cholesterol, mg/dL 119.1 (27.9) 126.9 (28.9) 131.9 (30.2) <0.001

HDL cholesterol, mg /dL 31.2 (9.9) 35.6 (12.5) 42.3 (15.6) <0.001

C-reactive protein, mg/L 6.2 [3.1-13.4] 5.4 [2.2-11.1] 7.1 [2.6-11.1] 0.014

Albumin, g/dL 3.8 (0.3) 3.8 (0.3) 3.9 (0.3) <0.001

Hemoglobin, g/dL 10.9 (1.4) 10.9 (1.3) 10.9 (1.3) 0.901

HbA1c, % 6.7 (1.3) 6.7 (1.3) 6.8 (1.2) 0.693

Phosphate, mg/dL 5.8 (1.7) 6.1 (1.6) 6.2 (1.6) 0.006

Apolipoprotein A-I, mg/dL 115.8 (19.9) 125.5 (20.0) 137.6 (25.5) <0.001

SAA(HDL) 6.7 [3.0-13.9] 5.6 [2.7-12.3] 5.3 [2.9-10.2] 0.013

SP-B(HDL) 7.3 [4.0-13.0] 6.6 [3.5-13.7] 5.6 [3.1-11.7] 0.166

Apolipoprotein C-III, mg/dL 19.3 (9.5) 20.8 (10.0) 20.9 (9.0) 0.027

Apolipoprotein C-II, mg/dL 6.0 (3.1) 6.5 (3.3) 6.4 (2.7) 0.068

ADMA 0.9 (0.2) 0.9 (0.2) 0.9 (0.2) 0.369

SDMA 2.5 (0.8) 2.6 (0.8) 2.6 (0.8) 0.031

Carbamylated albumin 0.6 (0.3) 0.6 (0.3) 0.7 (0.3) <0.001

Data shown are means (SDs) or medians [interquartile ranges] if not indicated otherwise. P values for comparisons of groups were calculated from ANOVA models (for continuous variables) or logistic regression models (for categorical variables). ADMA, asymmetric dimethylarginine; CABG, coronary artery bypass grafting surgery; CAD, coronary artery disease; MI, myocardial infarction; PTCA, percutaneous transluminal coronary angioplasty; SAA(HDL), HDL-bound serum amyloid A; SDMA, symmetric dimethylarginine; SP-B(HDL), HDL bound surfactant protein B; TIA, transitory ischemic attack.

Cholesterol Efflux in the 4D study

analyses with additional adjustment for a number of relevant clinical parameters (Table 2) further strengthened the conclusion that cholesterol efflux capacity was not associated with the risk for cardiovascular events or mortality in hemodialysis patients. The respective Cox models according to tertiles of cholesterol efflux are shown in Supplemental table II.

Table 2. Hazard ratios for combined primary endpoint, all cardiac events combined and all-cause mortality by cholesterol efflux capacity

combined primary endpoint

(498 events) all cardiac events combined

(534 events) all-cause mortality

(561 events) HR (95% CI) per 1-SD

increase p value HR (95% CI) per 1-SD

increase p value HR (95% CI) per 1-SD

increase p value

Model 1 0.96 (0.88 - 1.06) 0.417 0.92 (0.83 - 1.02) 0.108 0.96 (0.88 - 1.05) 0.390

Model 2 0.96 (0.88 - 1.05) 0.379 0.92 (0.82 - 1.03) 0.149 0.98 (0.90 - 1.06) 0.541

Model 3 0.89 (0.73 – 1.09) 0.271 0.77 (0.62 – 0.95) 0.015 0.91 (0.76 – 1.08) 0.263

Model 4 0.98 (0.91 - 1.06) 0.637 0.94 (0.85 - 1.04) 0.234 1.00 (0.95 - 1.06) 0.996

Model 5 0.98 (0.91 - 1.05) 0.550 0.94 (0.85 - 1.04) 0.216 0.99 (0.94 - 1.05) 0.853

Model 6 0.98 (0.91 - 1.05) 0.557 0.94 (0.85 - 1.04) 0.217 0.99 (0.94 - 1.05) 0.826

Continuous Cox regression model to assess the prognostic effect of cholesterol efflux on selected endpoints. Combined primary endpoint (composite of cardiac death, nonfatal myocardial infarction, or stroke). Model 1: univariate; Model 2: adjusted for age and sex; Model 3: adjusted for age, sex and CRP; Model 4: adjusted for traditional risk factors (age, sex, coronary artery disease, arrhythmia, transitory ischemic attack, congestive heart failure, peripheral vascular disease, smoking, systolic/diastolic blood pressure, body mass index, albumin, phosphate, hemoglobin, HbA1c, duration of dialysis); Model 5: adjusted for traditional risk factors, LDL-C, HDL-C, apoA-I; Model 6: adjusted for traditional risk factors, LDL-C, HDL-C, apoA-I, CRP.

Chapter 2

Figure 1. Kaplan-Meier curves according to tertiles of cholesterol efflux capacity. Prognostic effect of cholesterol efflux capacity on (A) combined primary endpoint (composite of cardiac death, nonfatal myocardial infarction and stroke), (B) cardiac events combined or (C) all-cause mortality. HR, hazard ratio; CI, confidence interval.

Next, participants were also stratified according to event occurrence (Supplemental table III).

Baseline HDL cholesterol efflux, total cholesterol and HDL-C levels were not different between patients that reached the defined endpoints and those that did not. For all three analyzed endpoints, patients with an event during follow-up had a higher prevalence of preexisting CVD, longer diabetes duration as well as lower hemoglobin and phosphate levels. Notably, patients who died displayed pronounced signs of wasting and inflammation, indicated by higher age, lower body mass index, higher prevalence of cardiovascular complications, higher plasma LDL- C and higher levels of CRP and SAA(HDL).

Furthermore, we attempted to determine a potential role for cholesterol efflux capacity in the treatment efficacy of atorvastatin in the 4D study population. Interestingly, stratified subgroup analysis by tertiles indicated a potential effect modification of atorvastatin on cardiac events (Supplemental Figure I) although the overall effect was not significant (P=0.190). In patients with the lowest cholesterol efflux capacity atorvastatin reduced the risk for all cardiac events combined (HR per 1 SD increase 0.66, 95% CI: 0.47–0.92, p=0.015), which remained significant following further adjustment with a number of relevant clinical parameters (Table 3).

Table 3. Cox regression models assessing treatment efficacy of Atorvastatin on all cardiac events combined stratified by tertiles of cholesterol efflux Model 1

Model 2 Model 3 Model 4 Model 5 HR (95% CI) p value HR (95% CI) p value HR (95% CI) p value HR (95% CI) p value HR (95% CI) p value Tertile 1 0.66 (0.47 - 0.92) 0.015 0.65 (0.47 - 0.92) 0.014 0.68 (0.49 - 0.94) 0.021 0.70 (0.50 - 0.98) 0.036 0.70 (0.50 - 0.97) 0.034 Tertile 2 0.85 (0.61 - 1.19) 0.336 0.85 (0.61 - 1.18) 0.330 0.85 (0.61 - 1.18) 0.326 0.85 (0.61 - 1.18) 0.322 0.85 (0.61 - 1.18) 0.329 Tertile 3 0.85 (0.56 - 1.27) 0.415 0.84 (0.56 - 1.26) 0.400 0.70 (0.48 - 1.02) 0.066 0.69 (0.47 - 1.01) 0.056 0.69 (0.47 - 1.01) 0.055 Model 1: univariate Model 2: adjusted for age and sex Model 3: adjusted for traditional risk factors (age, sex, coronary artery disease, arrhythmia, transient ischemic attack, congestive heart failure, peripheral vascular disease, smoking, systolic/diastolic blood pressure, body mass index, albumin, phosphate, hemoglobin, HbA1c, duration of dialysis) Model 4: adjusted for traditional risk factors, LDL-C, HDL-C, apoA-I Model 5: adjusted for traditional risk factors, LDL-C, HDL-C, apoA-I, CRP

Cholesterol Efflux in the 4D study

By contrast, in the higher tertiles of cholesterol efflux no differential effect of statin treatment was observed.

This post-hoc analysis of the 4D study demonstrates that HDL cholesterol efflux capacity is not associated with cardiovascular events or mortality in a large and sufficiently powered cohort of diabetic ESRD patients on hemodialysis. These data are consistent with our previous observation that HDL cholesterol efflux did not predict all-cause or specific CVD mortality in kidney transplant recipients. Intriguingly though, higher efflux was independently associated with graft survival, thus cholesterol efflux might represent a meaningful predictor of graft outcome ¹⁷.

Hence, our results extend previous observations that plasma HDL-C levels have little association with the risk of cardiovascular morbidity or mortality in ESRD ¹⁸ to the emerging area of HDL function studies. However, our findings are in apparent contrast to the prevailing view in the cardiovascular field, that, at least in cohorts with normal or only mildly impaired kidney function, the cholesterol efflux capacity of HDL might be an even stronger predictor for CVD risk than the mere measurement of circulating levels of HDL-C or apoA-I ^10,11. Although it should be noted that not all data on risk prediction by measuring cholesterol efflux are consistent with such a concept ¹⁹, we do believe that specific characteristics of ESRD patients impact the outcome of our study. Smaller cross-sectional studies carried out in the setting of renal failure indicated several functional impairments of HDL, which are not restricted to decreased cholesterol efflux but affect its anti-inflammatory ¹⁹, antioxidative ²⁰ or endothelial health-promoting activities ²¹ to a similar degree. Potential underlying reasons of the dysfunctional HDL in kidney disease could be the high inflammatory and oxidative stress burden of ESRD that contributes to the excessive cardiovascular risk and is accelerated by the presence of diabetes ^{22, 23}. In fact, certain modifications in the protein composition of uremic HDL linked to chronic systemic inflammation conceivably contribute to render the particle ineffective in providing sufficient anti-atherogenic protection ^{19, 20}. In this respect, we have previously found that HDL of ESRD patients has lost its anti-inflammatory property and even acts proinflammatory as a direct consequence of enrichment with the acute phase protein SAA

19. Subsequently, we could show that high HDL-bound SAA levels were predictive of cardiac events in the 4D population ²⁴. Such data suggest that the protein cargo on HDL might provide useful biomarkers to predict clinical events.

Another interesting outcome of our study was a potential effect modification of statin treatment. We found that atorvastatin reduced the risk for all cardiac events in patients of the assay, our assay system mainly differs in the choice of cell line from comparable studies in non- CKD cohorts ^{10, 11}. Such studies used cholesterol-equilibrated murine J774 cells with limited baseline expression of ABCA1, which upon induction with cAMP becomes the dominant efflux pathway. In contrast, we employed a human cell line that was loaded with cholesterol to reflect foam cells, and in which all efflux pathways are active ⁷. These differences might impact the results to a certain extent, however, it has to be noted that at present no gold standard or comparative studies are available for efflux assays. In addition, albumin was shown to facilitate cholesterol efflux ⁷; since in our cohort plasma albumin levels were within the normal range throughout the tertiles, we do not expect this parameter to have a major impact on the results, but can also not formally excluded this. Further, cholesterol efflux, as evaluated in the present Chapter 2

Cholesterol Efflux in the 4D study

study, does not take ESRD-related changes in cells and tissues into account, which also play a role in RCT. Here, specifically increased ACAT-1 activity in e.g. macrophages is worth mentioning ^28,29. In addition, modifications of SR-BI by reactive oxygen species, myeloperoxidase and glycation occur, that negatively impact its cholesterol uptake function ³⁰. Such alterations possibly lead to an even more substantial reduction in overall RCT in ESRD.

In conclusion, our study is the first to investigate the predictive potential of cholesterol efflux capacity in dialysis patients. Remarkably, there was no association of this key metric of HDL function and cardiovascular outcomes in this specific patient population. Nonetheless, we believe that our study significantly contributes to clarify the role of HDL in ESRD and adds important information to the understanding of the underlying pathophysiology of CVD in ESRD.

The ultimate clinical goal remains the identification of novel and reliable risk predictors for CVD in ESRD patients that also offer the possibility to be influenced by targeted therapies.

CONCISE METHODS

Study participants

The design of The German Diabetes Dialysis Study (4D Study) has been described previously ²⁵ The 4D study was a double-blind, randomized, multicenter trial including 1,255 patients with type 2 diabetes mellitus that were 18 to 80 years of age and had a previous duration of hemodialysis of <2 years. Patients were recruited between March 1998 and October 2002 and randomly assigned to receive daily treatment with 20 mg of atorvastatin (N=619) or placebo (N=636). Participants were followed-up at 4 weeks and every 6 months after randomization. At each follow-up visit information about any suspected endpoint or serious adverse event was recorded. The study was approved by the local ethics committees and performed according to the Declaration of Helsinki and informed consent was obtained from all participants.

Laboratory procedures

HDL cholesterol efflux capacity was determined from THP-1 macrophages towards apoB- depleted plasma as acceptor following a previously reported protocol ¹⁷. Briefly, THP-1 human monocytes (ATCC via LGC Promochem, Teddington, UK) were seeded in 48-well plates in RPMI 1640 Glutamax medium containing 10% fetal bovine serum and penicillin (100 U/mL)/streptomycin (100 μg/mL) and differentiated into macrophages with 100 nM phorbol myristate acetate for 24h. Then macrophages were loaded for 24h with acetylated LDL (50 μg protein/mL) and 1 μCi/mL 3H-cholesterol (Perkin Elmer, Boston, MA) followed by overnight equilibration with RPMI 1640 Glutamax medium containing 2% bovine serum albumin.

Thereafter cells were washed with PBS, and 2% of individual apoB-depleted plasma samples were added in RPMI 1640 Glutamax medium containing penicillin/streptomycin. After 5h of efflux the medium was collected, centrifuged in a table-top centrifuge (Hettich, Tuttlingen, Germany) for 5 minutes at 10 000 rpm to pellet cellular debris and radioactivity was determined in an aliquot by liquid scintillation counting (Packard 1600CA Tri-Carb, Packard, Meriden, CT).

To the cells 0.1 M NaOH was added for at least 30 min, then radioactivity remaining in the cells was determined. Cholesterol efflux was calculated as the percentage of radiolabel recovered in the medium related to the total added dose of radioactivity (counts from the medium added to counts from cells). Values obtained from negative control cells without added apoB-depleted

Chapter 2

patient plasma were subtracted to correct for unspecific efflux. Finally, values were normalized to a plasma pool from healthy controls present on every plate. Thus, values given for efflux do not have a specific unit. All determinations were carried out in duplicates at the same time using the same reagents to reduce variation due to different assay conditions. The intra-assay CV of this method is 5.4%, the interassay CV is 7.9%.

Endpoints

For our post-hoc analysis, we evaluated the following predefined and centrally adjudicated outcome measures: (1) combined primary endpoint (composite of cardiac death, non-fatal myocardial infarction, or stroke), (2) all cardiac events combined (cardiac death and non-fatal myocardial infarction) and (3) all-cause mortality.

Statistical Analyses

We examined characteristics of subgroups defined by tertiles of cholesterol efflux by calculating descriptive statistics (means and SDs for continuous variables and frequency tables for categorical variables). We compared the distribution of important cardiovascular parameters across quartiles by using ANOVA (for continuous variables) or chi-squared tests (for categorical variables). Further we conducted a correlation analysis calculating Pearson Correlation coefficients and scatter plots to investigate which clinical parameters were correlated with cholesterol efflux capacity in dialysis patients. We used time-to-event analysis (extended Cox regression model allowing for multiple events) aimed to (1) evaluate the effect of cholesterol efflux concentrations on endpoint occurrence and (2) to investigate the effect of cholesterol efflux concentrations on the efficacy of atorvastatin. For (1) we conducted a pooled analysis in both randomization groups including a dummy variable for randomization group and the efflux concentrations as continuous covariates, standardized by the sample standard deviation. For (2) we conducted efficacy analysis calculating hazard ratios (HRs) for atorvastatin versus placebo stratified by efflux tertiles. For both analyses we conducted sensitivity analysis fitting different models each including different covariates: Model 1: no additional covariate adjustment; Model 2: adjusted for age and sex; Model 3: adjusted for age, sex and CRP; Model 4: adjusted for traditional risk factors (age, sex, coronary artery disease, arrhythmia, transitory ischemic attack, congestive heart failure, peripheral vascular disease, smoking, systolic/diastolic blood pressure, BMI, albumin, phosphate, hemoglobin, HbA1c, duration of dialysis); Model 5: adjusted for traditional risk factors, LDL-C, HDL-C, apoA-I; Model 6: adjusted for traditional risk factors, LDL-C, HDL-C, apoA-I, CRP.

All statistical analyses were performed using the statistical software package STATA (StataCorp 2011, Stata Statistical Software: Release 13; StataCorp LP, College Station, TX).

References

1. Gordon DJ, Probstfield, JL, Garrison, RJ, Neaton, JD, Castelli, WP, Knoke, JD, Jacobs, DR, Jr., Bangdiwala, S, Tyroler, HA: High-density lipoprotein cholesterol and cardiovascular disease. Four prospective American studies.

Circulation, 79: 8-15, 1989.

2. van Capelleveen JC, Bochem, AE, Motazacker, MM, Hovingh, GK, Kastelein, JJ: Genetics of HDL-C: a causal link to atherosclerosis? Curr Atheroscler Rep, 15: 326, 2013.

3. Toth PP, Barter, PJ, Rosenson, RS, Boden, WE, Chapman, MJ, Cuchel, M, D'Agostino, RB, Sr., Davidson, MH, Davidson, WS, Heinecke, JW, Karas, RH, Kontush, A, Krauss, RM, Miller, M, Rader, DJ: High-density lipoproteins: a consensus statement from the National Lipid Association. J Clin Lipidol, 7: 484- 525, 2013.

4. Ray K, Wainwright, NW, Visser, L, Witteman, J, Breteler, M, Ambegaonkar, B, Hofman, A, Stricker, B, Wareham, N, Khaw, KT, Sandhu, M: Changes in HDL cholesterol and cardiovascular outcomes after lipid modification therapy. Heart, 98: 780-785, 2012.

5. Voight BF, Peloso, GM, Orho- Melander, M, Frikke-Schmidt, R, Barbalic, M, Jensen, MK, Hindy, G, Holm, H, Ding, EL, Johnson, T, Schunkert, H, Samani, NJ, Clarke, R, Hopewell, JC, Thompson, JF, Li, M, Thorleifsson, G, Newton-Cheh, C, Musunuru, K, Pirruccello, JP, Saleheen, D, Chen, L, Stewart, A, Schillert, A, Thorsteinsdottir, U, Thorgeirsson, G, Anand, S, Engert, JC, Morgan, T, Spertus, J, Stoll, M,

Berger, K, Martinelli, N, Girelli, D, McKeown, PP, Patterson, CC, Epstein, SE, Devaney, J, Burnett, MS, Mooser, V, Ripatti, S, Surakka, I, Nieminen, MS, Sinisalo, J, Lokki, ML, Perola, M, Havulinna, A, de Faire, U, Gigante, B, Ingelsson, E, Zeller, T, Wild, P, de Bakker, PI, Klungel, OH, Maitland-van der Zee, AH, Peters, BJ, de Boer, A, Grobbee, DE, Kamphuisen, PW, Deneer, VH, Elbers, CC, Onland-Moret, NC, Hofker, MH, Wijmenga, C, Verschuren, WM, Boer, JM, van der Schouw, YT, Rasheed, A, Frossard, P, Demissie, S, Willer, C, Do, R, Ordovas, JM, Abecasis, GR, Boehnke, M, Mohlke, KL, Daly, MJ, Guiducci, C, Burtt, NP, Surti, A, Gonzalez, E, Purcell, S, Gabriel, S, Marrugat, J, Peden, J, Erdmann, J, Diemert, P, Willenborg, C, Konig, IR, Fischer, M, Hengstenberg, C, Ziegler, A, Buysschaert, I, Lambrechts, D, Van de Werf, F, Fox, KA, El Mokhtari, NE, Rubin, D, Schrezenmeir, J, Schreiber, S, Schafer, A, Danesh, J, Blankenberg, S, Roberts, R, McPherson, R, Watkins, H, Hall, AS, Overvad, K, Rimm, E, Boerwinkle, E, Tybjaerg-Hansen, A, Cupples, LA, Reilly, MP, Melander, O, Mannucci, PM, Ardissino, D, Siscovick, D, Elosua, R, Stefansson, K, O'Donnell, CJ, Salomaa, V, Rader, DJ, Peltonen, L, Schwartz, SM, Altshuler, D, Kathiresan, S: Plasma HDL cholesterol and risk of myocardial infarction: a mendelian randomisation study. Lancet, 380:

572-580, 2012.

6. Heinecke JW: The not-so-simple HDL story: A new era for quantifying HDL and cardiovascular risk? Nat Med, 18: 1346-1347, 2012.

7. Triolo M, Annema, W, Dullaart, RP, Cholesterol Efflux in the 4D study

Chapter 2

Tietge, UJ: Assessing the functional properties of high-density lipoproteins: an emerging concept in cardiovascular research.

Biomark Med, 7: 457-472, 2013.

8. Annema W, Tietge, UJ: Regulation of reverse cholesterol transport - a comprehensive appraisal of available animal studies. Nutr Metab (Lond), 9: 25, 2012.

9. Khera AV, Rader, DJ: Future therapeutic directions in reverse cholesterol transport. Curr Atheroscler Rep, 12: 73-81, 2010.

10. Rohatgi A, Khera, A, Berry, JD, Givens, EG, Ayers, CR, Wedin, KE, Neeland, IJ, Yuhanna, IS, Rader, DR, de Lemos, JA, Shaul, PW: HDL Cholesterol Efflux Capacity and Incident Cardiovascular Events. N Engl J Med, 371: 2383-2393, 2014.

11. Khera AV, Cuchel, M, de la Llera- Moya, M, Rodrigues, A, Burke, MF, Jafri, K, French, BC, Phillips, JA, Mucksavage, ML, Wilensky, RL, Mohler, ER, Rothblat, GH, Rader, DJ:

Cholesterol efflux capacity, high- density lipoprotein function, and atherosclerosis. N Engl J Med, 364:

127-135, 2011.

12. Ortiz A, Covic, A, Fliser, D, Fouque, D, Goldsmith, D, Kanbay, M, Mallamaci, F, Massy, ZA, Rossignol, P, Vanholder, R, Wiecek, A, Zoccali, C, London, GM: Epidemiology, contributors to, and clinical trials of mortality risk in chronic kidney failure. Lancet, 383: 1831-1843, 2014.

13. Jardine MJ, Hata, J, Woodward, M, Perkovic, V, Ninomiya, T, Arima, H, Zoungas, S, Cass, A, Patel, A, Marre, M, Mancia, G, Mogensen, CE, Poulter, N, Chalmers, J: Prediction of kidney-related outcomes in patients with type 2 diabetes. Am J

Kidney Dis, 60: 770-778, 2012.

14. Zewinger S, Speer, T, Kleber, ME, Scharnagl, H, Woitas, R, Lepper, PM, Pfahler, K, Seiler, S, Heine, GH, Marz, W, Silbernagel, G, Fliser, D: HDL Cholesterol Is Not Associated with Lower Mortality in Patients with Kidney Dysfunction. J Am Soc Nephrol, 25: 1073-1082, 2014.

15. Silbernagel G, Genser, B, Drechsler, C, Scharnagl, H, Grammer, TB, Stojakovic, T, Krane, V, Ritz, E, Wanner, C, Marz, W: HDL Cholesterol, Apolipoproteins, and

Cardiovascular Risk in

Hemodialysis Patients. J Am Soc Nephrol, 26: 484-492, 2014.

16. Moradi H, Streja, E, Kashyap, ML, Vaziri, ND, Fonarow, GC, Kalantar- Zadeh, K: Elevated high-density lipoprotein cholesterol and cardiovascular mortality in maintenance hemodialysis patients. Nephrol Dial Transplant, 29: 1554-1562., 2014.

17. Annema W, Dikkers, A, de Boer, JF, Dullaart, RP, Sanders, JF, Bakker, SJ, Tietge, UJ: HDL Cholesterol Efflux Predicts Graft Failure in Renal Transplant Recipients. J Am Soc Nephrol, 27: 595-603, 2015.

18. Vaziri ND: Risk factors: HDL- cholesterol levels and mortality in patients with ESRD. Nat Rev Nephrol, 10: 621-623, 2014.

19. Weichhart T, Kopecky, C, Kubicek, M, Haidinger, M, Doller, D, Katholnig, K, Suarna, C, Eller, P, Tolle, M, Gerner, C, Zlabinger, GJ, van der Giet, M, Horl, WH, Stocker, R, Saemann, MD: Serum amyloid A in uremic HDL promotes inflammation. J Am Soc Nephrol, 23:

934-947, 2012.

20. Holzer M, Birner-Gruenberger, R, Stojakovic, T, El-Gamal, D, Binder, V,

Cholesterol Efflux in the 4D study

Wadsack, C, Heinemann, A, Marsche, G: Uremia alters HDL composition and function. J Am Soc Nephrol, 22: 1631-1641, 2011.

21. Speer T: Abnormal high-density lipoprotein induces endothelial dysfunction via activation of Toll- like receptor-2. Immunity, 38: 754- 768, 2013.

22. Gansevoort RT, Correa-Rotter, R, Hemmelgarn, BR, Jafar, TH, Heerspink, HJ, Mann, JF, Matsushita, K, Wen, CP: Chronic kidney disease and cardiovascular risk:

epidemiology, mechanisms, and prevention. Lancet, 382: 339-352, 2013.

23. Chang YT, Wu, JL, Hsu, CC, Wang, JD, Sung, JM: Diabetes and end-stage renal disease synergistically contribute to increased incidence of cardiovascular events: a nationwide follow-up study during 1998-2009. Diabetes Care, 37: 277- 285, 2014.

24. Kopecky C, Genser, B, Drechsler, C, Krane, V, Kaltenecker, CC, Hengstschlager, M, Marz, W, Wanner, C, Saemann, MD, Weichhart, T: Quantification of HDL proteins, cardiac events, and mortality in patients with type 2 diabetes on hemodialysis. Clin J Am Soc Nephrol, 10: 224-231, 2015.

25. Wanner C, Krane, V, Marz, W, Olschewski, M, Mann, JF, Ruf, G, Ritz, E: Atorvastatin in patients with type 2 diabetes mellitus

undergoing hemodialysis. N Engl J Med, 353: 238-248, 2005.

26. Fellstrom BC, Jardine, AG, Schmieder, RE, Holdaas, H, Bannister, K, Beutler, J, Chae, DW, Chevaile, A, Cobbe, SM, Gronhagen- Riska, C, De Lima, JJ, Lins, R, Mayer, G, McMahon, AW, Parving, HH, Remuzzi, G, Samuelsson, O, Sonkodi, S, Sci, D, Suleymanlar, G, Tsakiris, D, Tesar, V, Todorov, V, Wiecek, A, Wuthrich, RP, Gottlow, M, Johnsson, E, Zannad, F:

Rosuvastatin and cardiovascular events in patients undergoing hemodialysis. N Engl J Med, 360:

1395-1407, 2009.

27. Heymann EP, Kassimatis, TI, Goldsmith, DJ: Dyslipidemia, statins, and CKD patients' outcomes - review of the evidence in the post- sharp era. J Nephrol, 25: 460-472, 2012.

28. Liang K, Vaziri, ND: Upregulation of

acyl-CoA: cholesterol

acyltransferase in chronic renal failure. Am J Physiol Endocrinol Metab, 283: E676-681, 2002.

29. Moradi H, Yuan, J, Ni, Z, Norris, K, Vaziri, ND: Reverse cholesterol transport pathway in experimental chronic renal failure. Am J Nephrol, 30: 147-154, 2009.

30. Norata GD, Pirillo, A, Catapano, AL:

Modified HDL: biological and physiopathological consequences.

Nutr Metab Cardiovasc Dis, 16:

371-386, 2006.

Chapter 2

Supplement

Supplemental Table I. Variables which are associated with cholesterol efflux capacity

r p value

Apolipoprotein A-I 0.264 <0.001

HDL-C 0.246 <0.001

Albumin 0.161 <0.001

LDL-C 0.101 0.001

Duration of dialysis 0.098 0.001

Total cholesterol 0.096 0.001

Serum creatinine 0.056 0.046

Diastolic blood pressure 0.055 0.063

Systolic blood pressure 0.052 0.079

Phosphate 0.046 0.123

Ultrafiltration volume 0.044 0.140

HbA1c 0.034 0.258

Age -0.044 0.138

C-reactive protein -0.044 0.140

SAA(HDL) -0.030 0.316

Duration of diabetes -0.023 0.445

BMI -0.022 0.466

Hemoglobin -0.008 0.785

SP-B(HDL) 0.005 0.857

Variables are listed in decreasing order of strength of association. r, Pearson correlation coefficient;

BMI, body mass index; SAA(HDL), HDL-bound serum amyloid A; SP-B(HDL), HDL-bound surfactant protein B.