lable at ScienceDirect

Atherosclerosis 241 (2015) 433e442

Contents lists avai

Atherosclerosis

journal homepage: www.elsevier .com/locate/atherosclerosis

Statin therapy reduces plasma endothelin-1 concentrations: Ameta-analysis of 15 randomized controlled trials

Amirhossein Sahebkar a, b, 1, Kazuhiko Kotani c, 1, Corina Serban d, Sorin Ursoniu e,Dimitri P. Mikhailidis f, Steven R. Jones g, Kausik K. Ray h, i, Michael J. Blaha g,Jacek Rysz h, i, Peter P. Toth g, j, Paul Muntner k, Gregory Y.H. Lip l, Maciej Banach m, *, Lipidand Blood Pressure Meta-analysis Collaboration (LBPMC) Groupa Biotechnology Research Center, Mashhad University of Medical Sciences, Mashhad, Iranb Metabolic Research Centre, Royal Perth Hospital, School of Medicine and Pharmacology, University of Western Australia, Perth, Australiac Department of Clinical Laboratory Medicine, Jichi Medical University, Shimotsuke City, Tochigi, Japand Department of Functional Sciences, Discipline of Pathophysiology, “Victor Babes” University of Medicine and Pharmacy, Timisoara, Romaniae Department of Functional Sciences, Discipline of Public Health, “Victor Babes” University of Medicine and Pharmacy, Timisoara, Romaniaf Department of Clinical Biochemistry, Royal Free Campus, University College London Medical School, University College London (UCL), London, UKg The Johns Hopkins Ciccarone Center for the Prevention of Heart Disease, Baltimore, MD, USAh Cardiovascular Sciences Research Centre, St George's University of London, Cranmer Terrace, London, UKi Department of Nephrology, Hypertension and Family Medicine, Medical University of Lodz, Polandj Preventive Cardiology, CGH Medical Center, Sterling, IL, USAk Department of Epidemiology, University of Alabama at Birmingham, Birmingham, AL, USAl University of Birmingham Centre for Cardiovascular Sciences, City Hospital, Birmingham, UKm Department of Hypertension, Chair of Nephrology and Hypertension, Medical University of Lodz, Poland

a r t i c l e i n f o

Article history:Received 6 March 2015Received in revised form26 April 2015Accepted 19 May 2015Available online 3 June 2015

Keywords:Endothelin-1Endothelial dysfunctionLipophilicityStatinsTherapy

* Corresponding author. Department of Hypertensiin Lodz, Medical University of Lodz, Zeromskiego 113

E-mail address: maciejbanach@aol.co.uk (M. Bana1 Drs Sahebkar and Kotani contributed equally to t

http://dx.doi.org/10.1016/j.atherosclerosis.2015.05.0220021-9150/© 2015 Elsevier Ireland Ltd. All rights rese

a b s t r a c t

Objective: Raised plasma endothelin-1 (ET-1) levels may be a risk factor for vascular dysfunction andcardiovascular (CV) disease. This meta-analysis assessed the effect of statins on circulating ET-1concentrations.Methods and results: The search included PUBMED, Cochrane Library, Web of Science, Scopus, andEMBASE up to September 30, 2014 to identify randomized controlled trials (RCTs) with ET-1 measure-ment during statin therapy. Quantitative data synthesis was performed using a random-effects model,with weighed mean difference (WMD) and 95% confidence interval (CI) as summary statistics. Data from15 RCTs showed that statin therapy significantly reduces plasma ET-1 concentrations (WMD: �0.30 pg/mL, 95%CI: �0.47, �0.13; p < 0.01). This effect was robust in sensitivity analysis, and not largely affectedby the duration of statin therapy (<12 weeks e WMD: �0.51 pg/mL, 95%CI: �0.89, �0.14, p < 0.01; >12week eWMD: �0.19 pg/mL, 95%CI: �0.36, �0.02; p < 0.05) or by dose of statins (<40 mg/day e

WMD: �0.27 pg/mL, 95%CI: �0.49, �0.05; p ¼ 0.01; >40 mg/day e WMD: �0.38 pg/mL, 95%CI: �0.68, �0.08; p ¼ 0.01). Lipophilic (atorvastatin, simvastatin, fluvastatin, and cerivastatin e

WMD: �0.34 pg/mL, 95%CI: �0.55, �0.13; p < 0.01), but not a hydrophilic statin (pravastatin e

WMD: �0.18 pg/mL, 95%CI: �0.44 �0.08; p > 0.05) had a significant effect in promoting ET-1 reduction.Conclusions: Statin therapy significantly reduces circulating ET-1 concentrations, regardless of treatmentduration or dose of statins. This effect of statins may be influenced by statin lipophilicity. There is a needto establish whether lowering ET-1 levels has a beneficial effect on CV events.

© 2015 Elsevier Ireland Ltd. All rights reserved.

on, WAM University Hospital, 90-549, Lodz, Poland.ch).his meta-analysis.

rved.

1. Background

Atherosclerosis leads to cardiovascular diseases (CVD), a majorcause of morbidity andmortality [1,2]. The endothelium plays a rolein atherogenesis, and endothelial dysfunction is considered to be

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442434

involved in the onset of CVD and its progression [2]. Endothelialdysfunction results in reduced nitric oxide and prostacyclinbioavailability, vasoconstriction, oxidative stress, inflammation,and platelet activation [3,4].

Among molecules that may modulate endothelial function,endothelin-1 (ET-1) is a peptide, which is primarily produced byvascular endothelial cells [5]. ET-1 was first identified as a vaso-constrictor [6]. The synthesis of ET-1 starts from precursor pep-tides; endothelin-converting enzyme converts pro-endothelin toET-1 [7]. ET-1 is multifunctional, and promotes inflammation andcell proliferation within arterial vessel walls [5]. The synthesis ofET-1 is mediated by various factors, including oxidized low-densitylipoprotein (LDL), platelet activation, and hypoxia [8e10].Conversely, ET-1 may also induce LDL oxidation and platelet acti-vation [11]. Thus, over-production of ET-1 may be associated withincreased risk for CVD [5]. The control of ET-1 expression mightprovide benefits against the development of atherosclerosis andCVD events. Consistent with this is the observation that antagonismof the ET-1 system can modify atherogenesis [12].

Many clinical trials have reported the beneficial effects of statinsin CVD prevention [13e16]. Recently, attention has been paid to thepleiotropic actions of statins beyond simple cholesterol-lowering[17e19]. In experimental studies, statins can inhibit ET-1 produc-tion [20]; however, findings concerning changes in ET-1 concen-trations following statin therapy have been inconsistent. Therefore,in the present meta-analysis we evaluated the impact of statintherapy on circulating ET-1 concentrations.

2. Methods

2.1. Search strategy

This study was designed according to the guidelines of the2009 preferred reporting items for systematic reviews and meta-analysis (PRISMA) statement [21]. Our search included SCOPUS(http://www.scopus.com), Medline (http://www.ncbi.nlm.nih.gov/pubmed), Web of Science (http://apps.webofknowledge.com), and Cochrane Library (www.thecochranelibrary.com/) da-tabases. It was limited to randomized controlled trials (RCTs)carried out from January 1, 1970 to September 30, 2014, investi-gating the potential effects of statin therapy on ET-1 concentra-tions. The databases were searched using the following searchterms in titles and abstracts (also in combination with MESHterms): (rosuvastatin or pravastatin or fluvastatin or simvastatinor atorvastatin or pitavastatin or lovastatin or cerivastatin or“statin therapy” or statins) and (endothelin-1 or endothelin or ET-1). The wild-card term “*” was used to increase the sensitivity ofthe search strategy. No language restriction was used in theliterature search. The search was limited to studies in human. Tworeviewers (CS and AS) evaluated each article separately. Dis-agreements were resolved by discussion with a third party (MB).

2.2. Study selection

Original studies were included if they met the following inclu-sion criteria: (i) a randomized controlled trial (RCT) in either parallelor cross-over design, (ii) investigating the impact of statin therapyon plasma/serum levels of ET-1, (iii) treatment duration of at leasttwo weeks, and, (iv) presentation of sufficient information on ET-1concentrations at baseline and at the end of study in both statinand control groups or providing the net changes in each group.

Exclusion criteria were: (i) non-clinical studies, (ii) uncontrolledtrials, (iii) lack of sufficient information on baseline or follow-up ET-1 levels, (iv) inability to obtain adequate details of study method-ology or results from the article or the investigators, and, (v) the

study was ongoing. Exclusion of an article for the latter reason wascarried out if no feedback was received after contacting theauthor(s).

2.3. Data extraction

Eligible studies were reviewed and the following data wereabstracted: 1) first author's name, 2) year of publication, 3) studylocation, 4) number of participants in the statin and control groups,5) age, gender and body mass index (BMI) of study participants, 6)baseline levels of total cholesterol, LDL cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglycerides, high-sensitivity C-reactive protein (hsCRP) and glucose, 7) systolic anddiastolic blood pressure, and, 8) data regarding baseline and follow-up concentrations of ET-1. In case the values were only presented asgraphs, the GetData Graph Digitizer 2.24 (http://getdata-graph-digitizer.com/) software was used to digitize and extract the data.

2.4. Quality assessment

The quality of included studies was assessed using the Jadadscale. This scale encompasses randomization (0e2 points), blinding(0e2 points), and dropouts and withdrawals (0e1 point). Theoverall score of a study according to this scale ranges between 0 and5, with higher scores indicative of a better quality [22]. Studies withJadad scores of �2 and � 3 were considered as low- and high-quality, respectively [23].

2.5. Quantitative data synthesis

Meta-analysis was conducted using Review Manager, version5.2 (Cochrane Collaboration), and Comprehensive Meta-Analysis(CMA) V2 software (Biostat, NJ) [24]. Standard deviations (SDs)of the mean difference were calculated using the following for-mula: SD ¼ square root [(SDpre-treatment)2 þ (SDpost-treatment)2 e

(2R � SDpre-treatment � SDpost-treatment)], assuming a correlationcoefficient (R) ¼ 0.5. In case of reporting SEM, SD was estimatedusing the following formula: SD ¼ SEM � sqrt (n), where n is thenumber of subjects.

Net changes in measurements (change scores) were calculatedfor parallel and cross-over trials, as follows: (measure at end offollow-up in the treatment group � measure at baseline in thetreatment group) � (measure at end of follow-up in the controlgroup � measure at baseline in the control group). A random-effects model (using DerSimonian-Laird method) and the genericinverse variance method were used to compensate for the het-erogeneity of studies in terms of statin type, statin dose, studydesign, treatment duration, and the characteristics of populationsbeing studied [25]. Effect sizes were expressed as weighed meandifference (WMD) and 95% confidence interval (CI). Post-hoc sub-group analyses were carried out to explore the impact of dose(<40 mg/day vs > 40 mg/day), duration (<12 weeks vs > 12 weeks),and type (lipophilic vs hydrophilic) of statin therapy on plasma ET-1concentrations. In order to evaluate the influence of each study onthe overall effect size, sensitivity analysis was conducted using theone-study remove (leave-one-out) approach [26]. The power ofanalysis to detect statistically significant difference between statinand control groups was performed using the PS software [27].

In the absence of trials making head-to-head comparison ofhydrophilic versus lipophilic statins, the effect of these two types ofstatins on plasma ET-1 levels were compared using adjusted indi-rect comparison according to the method proposed by Song et al.[28] and Bucher et al. [29]. In this method, treatment effects esti-mated for each type of statins in the random-effects model could becompared indirectly through common controls.

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442 435

2.6. Meta-regression

Random-effects meta-regression was performed using unre-stricted maximum likelihood method to evaluate the associationbetween calculated WMD in plasma ET-1 concentrations withduration and dose of treatment with statins, as well as age, genderand changes in plasma LDL-C concentrations as potential modera-tors of treatment response.

2.7. Publication bias

Potential publication bias was explored using visual inspectionof Begg's funnel plot asymmetry, and Begg's rank correlation andEgger's weighted regression tests. Duval and Tweedie “trim and fill”and “fail-safe N” methods were used to adjust the analysis for theeffects of publication bias [30].

2.8. Heterogeneity analysis

Heterogeneity analysis was performed using the Cochrane's Qtest and I2 index. Another attempt to explore heterogeneity wasmade via the Galbraith plot, a scatter plot of WMD divided by itsstandard error (Z-statistic) against the reciprocal of the standarderror in the included studies.

3. Results

The initial screening for potential relevance removed the articlesin whose titles and/or abstracts were obviously irrelevant. Amongthe 30 full text articles assessed for eligibility, 15 studies wereexcluded because of: lack of assessment of plasma ET-1

Fig. 1. Flow chart of the number of studies iden

concentrations (n ¼ 1), insufficient data on plasma ET-1 levels(n ¼ 4), not being an original research study (n ¼ 1), not having anappropriate RCT design (n ¼ 4), short (<2 weeks) duration oftreatment (n ¼ 2) and non-English language (n ¼ 3) (Fig. 1).

3.1. Characteristics of included studies

After final assessment, 15 RCTs [31e45] met the inclusioncriteria and were considered for the final meta-analysis. In total,810 participants were randomized, of whom 421 were allocatedto statin intervention and 389 to controls. The number of par-ticipants in these trials ranged from 32 to 82. Included studieswere published between 1999 and 2013, and were conducted inEgypt, Norway, Russian Federation, Canada, USA (2 trials), Italy,Taiwan (3 trials), Poland, China, Japan, Sweden, and India. Thefollowing statin doses were administered in the included trials:10e80 mg atorvastatin/day, 10e40 mg pravastatin/day, 40 mg/day simvastatin and fluvastatin, and 0.15 mg/day cerivastatin.Duration of statin intervention ranged between 2 weeks and 12months. 12 trials were designed as parallel-group studies and 3as crossover, comprising a total of 16 treatment arms. The mea-surements of ET-1 concentrations were based on the immuno-assays in all the included studies. Demographic and baselineparameters of the included studies are shown in Table 1. Thesystematic assessment of bias in the included studies is shown inSupplemental Table 1.

3.2. Quantitative data synthesis

The meta-analysis of data from 15 RCTs (comprising 16 treat-ment arms) [31e45] showed a significant effect of statin therapy in

tified and included into the meta-analysis.

Table 1Demographic characteristics of the included studies.

Study Abou Raya et al.[31]

Asberg et al. [32] Barsuk et al. [33] Dupuis et al. [34] Economides et al. [45] Glorioso et al. [35]

Year 2007 2003 2013 1999 2004 1999Jadad score 3 4 3 3 4 4Location Egypt Norway Russian Federation Canada USA ItalyDesign Randomized,

placebo-controlledparallel group trial

Randomizeddouble-blindplacebo-controlledparallel group trial

Randomizedplacebo-controlledparallel group trial

Randomizeddouble-blindplacebo-controlledparallel group trial

Randomizeddouble-blindplacebo-controlledparallel group trial

Randomizeddouble-blindplacebo-controlledcrossover grouptrial

Duration of trial 6 months 12 weeks 30 days 6 weeks 12 weeks 32 weeksInclusion criteria Patients fulfilling

the ACRpreliminaryclassificationcriteria forSystemicSclerosis

Kidneytransplantrecipients over18 yr of agewith a totalcholesterol of4.0e9.0 mmol/L

Patients withterminal chronicrenal failure,included inhemodialysisprogram notless than 6months, withdyslipidemia,aged 35e70 yrs,hemoglobinlevel �100 g/L,urea reductionrate(URR) �65%,dialysis dose�1.2

Patientsadmitted to thecoronary careunit of theMontreal HeartInstitute with adiagnosis ofacute myocardialinfarction orunstable anginaif they hadadmission totalserumcholesterol�5.2 mmol/L orLDL cholesterol�3.4 mmol/Land serumtriglyceride�4.5 mmol/L

Patients aged 21e80 yearsand at risk for type 2diabetes (either having afirst-degree relative withtype 2 diabetes and normalglucose tolerance or impairedglucose tolerance definedas a 2-h blood glucose valuebetween 140 and 199 mg/dlduring a 75-g oral glucosetolerance test) or had type1 or type 2 diabetes

Patients aged40e70 yearswho haddiastolichypertensionand primaryhypercholesterolemia andwho were nottaking anylipid-lowering orantihypertensivedrugs

Statin form atorvastatin fluvastatin atorvastatin pravastatin atorvastatin pravastatinStatin intervention 40 mg/day 40 mg/day 20 mg/day 40 mg/day 20 mg/day 20 to 40 mg/day**Participants Case 20 37 28 28 15# 19## 25

Control 20 35 26 27 15# 18## 25Age (years) Case 59.9 ± 10.1 52 ± 13 NS 55 ± 2.1 48 ± 13# 51 ± 14## 53 ± 2

Control 58.7 ± 9.2 51 ± 17 NS 56 ± 2.3 49 ± 11# 55 ± 11##

Male (%) Case 22.5 81.1 70.59 81.48.0 54.05# 57.5## 43.0

Control 12.5 71.4 76.47 92.85BMI (kg/m2) Case NS NS NS NS 29.5 ± 5.8# 29.8 ± 9.4## 25.8 ± 4.2

Control NS NS NS NShs-CRP (mg/L) Case 3.79 (1.8) NS NS NS 0.24 (0.07e0.35)# 0.30 (0.11e0.62)## NS

Control 3.85 (1.4) NS NS NS 0.20 (0.06e0.53)# 0.41 (0.17e0.76)##

Total cholesterol(mg/dL)

Case 198.3 ± 25.5 201.49 ± 46.32 NS 246.65 ± 7.33 193 ± 42# 205 ± 40## 242.79 ± 20.07

Control 189.4 ± 25.9 201.11 ± 52.49 NS 247.43 ± 5.4 216 ± 32# 208 ± 47##

LDL-C (mg/dL) Case 112.6 ± 21.4 132.4 ± 39.76 NS 160.19 ± 5.4 115 ± 31# 124 ± 36## 166.37 ± 18.91

Control 111.8 ± 20.9 132.4 ± 45.55 NS 167.52 ± 6.95 129 ± 28# 125 ± 37##

HDL-C (mg/dL) Case 59.9 ± 19.6 38.6 ± 11.58 NS 39.76 ± 2.32 53 ± 9# 63 ± 14## 56.36 ± 11.97

Control 60.1 ± 19.5 35.51 ± 10.42 NS 43.23 ± 3.09 66 ± 24# 59 ± 14##

Triglycerides(mg/dL)

Case NS 173.46 ± 117.7 NS 193.81 ± 15.93 132 ± 96# 101 ± 78## 116.82 ± 36.28

Control NS 147.79 ± 80.53 NS 193.81 ± 17.7 116 ± 71# 106 ± 79##

Glucose (mg/dL) Case NS NS NS NS 92 ± 12# 163 ± 72## 81.9 ± 5.94

Control NS NS NS NS 89 ± 10# 182 ± 79##

SBP (mmHg) Case NS 148 ± 19 NS 116 ± 3.4 126 ± 12# 128 ± 14## 149 ± 6

Control NS 142 ± 17 NS 122 ± 3.6 123 ± 21# 125 ± 13##

DBP (mmHg) Case NS 89 ± 7 NS 74 ± 4.3 80 ± 7# 80 ± 9## 97 ± 2

Control NS 90 ± 9 NS 72 ± 2.5 79 ± 10# 78 ± 9##

Endothelin-1(pg/mL)

Case 3.2 ± 1.7 1.19 ± 1.02* 2.24 ± 0.32 1.46 ± 0.7 0.82 ± 0.25# 1.19 ± 0.42## 4.5 ± 2.1

Control 2.98 ± 1.9 1.19 ± 0.87* 2.24 ± 0.32 1.48 ± 0.8 0.89 ± 0.40# 1.02 ± 0.29##

Values are expressed as mean ± SD or median (25e75 percentiles). ABBREVIATIONS: BMI: body mass index; NS: not stated; LDL-C: low-density lipoprotein cholesterol;HDL-C: high-density lipoprotein cholesterol; SBP: systolic blood pressure; DBP: diastolic blood pressure; hs-CRP: high-sensitivity C-reactive protein; BMI: body massindex; *the value was provided following 12 week treatment, # denotes at risk of type 2 diabetes arm; ## denotes diabetic patients arm; **If total plasma cholesterol was>5.46mmol/L 8weeks after randomization or 8weeks after crossover, the drug dose was doubled to 40mg/day; $ denotes value after statin; $$ denotes value after placebo;x denotes statin group xx denotes NCB-02 group (two capsules containing curcumin 150 mg twice daily).

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442436

Lee et al. [36] Lee et al. [38] Lee et al. [37] Lewandowski et al.[39]

Li & Hui et al. [40] Mozaffarian et al.[42]

Nakamura et al.[41]

Tehrani et al. [43] Usharani et al. [44]

2002 2005 2009 2010 2005 2005 2001 2013 20085 5 5 3 3 5 4 3 3Taiwan Taiwan Taiwan Poland China USA Japan Sweden India

Randomizeddouble-blindplacebo-controlledparallel-group trial

Randomizeddouble-blindplacebo-controlledparallel-group trial

Randomizeddouble-blindplacebo-controlledparallel-group trial

Randomizeddouble-blindplacebo-controlledparallel-group trial

Randomizeddouble-blindplacebo-controlledparallel-group trial

Randomizeddouble-blindplacebo-controlledcrossover grouptrial

Randomizeddouble-blindplacebo-controlledparallel-group trial

Randomizeddouble-blindplacebo-controlledcrossover grouptrial

Randomizedplacebo-controlledparallel-group trial

6 months 12 months 6 months 8 weeks 2 weeks 16 weeks 6 months 2 months 8 weeksPatients withproteinuriaand stable,well-controlledhypertensionwith a seateddiastolic bloodpressure of90 mm Hgand systolic bloodpressure of140 mmHg at a3-monthscreening period

Consecutiveproteinuricpatients withstable,well-controlledhypertension witha seated diastolicblood pressureof 90 mm Hgand systolicblood pressureof 140 mmHg at a3-monthscreening period

Patients withchronic obstructivepulmonary diseasein whom aroutineechocardiogramshowed pulmonaryhypertension,stable for atleast 3 monthsand between 40and 80 yrsof age

Men withhypercholesterolemia(total andLDL cholesterollevels �90and � 115 mg/dl,respectively), andmild-to moderateessentialhypertension(bloodpressures:140e179/90e109 mmHg).

Patients withangiographically-documentedcoronaryartery disease

Patients withheart failureand atleast New YorkHeart Associationclass II symptomsand left ventricularejection fraction<40%

ormotensive type2 diabetic patientswithmicroalbuminuria(20e200 mg/min)and dyslipidemia(total cholesterol>200 mg/dl, LDLcholesterol>160 mg/dl,HDL cholesterol<35 mg/dl, andtriglyceride>150 mg/dl)

Patients agedbetween 30and 70 yrs withtype 1 diabetesand elevated levelsof low-densitylipoprotein (LDL)(>2.5 mmol/L) and/ortotal cholesterol(>4.5 mmol/L)

Patients with type2 diabetes,21e80 yrs,with fastingplasma glucoseof �130 mg/dL,a glycosylatedhaemoglobin(HbA1c)range of between7% and 11%, andtaking stableantihyperglycaemicmedications fora 2-month period

pravastatin pravastatin pravastatin simvastatin simvastatin atorvastatin cerivastatin atorvastatin atorvastatin10 mg/day 10 mg/day 40 mg/day 40 mg/day 40 mg/day 10 mg/day 0.15 mg/day 80 mg/day 10 mg/day31 42 27 15 16 22 30 20 23x

23xx

32 40 26 16 16 22 30 20 2150 ± 9 50 ± 9 71 ± 8 39.0 ± 10.2 NS 51 ± 11 58 ± 10 44 (39e61) 50.47 ± 10.35x

55.52 ± 10.76xx

47 ± 8 48 ± 8 72 ± 6 38.3 ± 10.1 NS 55 ± 9 49.75 ± 8.1858.06 69.05 74.07 100.0 NS 86.36 60.0 50.0 52.17x

52.17xx

65.62 67.5 73.08 100.0 NS 66.6 52.3824.7 ± 1.1 24.8 ± 1.4 22 ± 2 28.7 ± 4.5 NS 31.8 ± 6.4 NS 25 ± 3 25.03 ± 1.83x

24.66 ± 2.42xx

24.5 ± 1.2 24.6 ± 1.3 23 ± 1 27.3 ± 4.1 NS NS 23.98 ± 2.35NS NS NS NS NS 3.2 ± 0.9$ NS NS NSx

NSxx

NS NS NS NS NS 4.8 ± 1.3$$ NS NS210 ± 23 208 ± 23 240 ± 43 249 ± 30.6 NS 202 ± 37 262 ± 42 185.28 ± 19.3 196.78 ± 35.28x

195.0 ± 41.16xx

205 ± 23 202 ± 23 245 ± 39 232 ± 22.1 NS 258 ± 42 196.95 ± 35.72125 ± 23 121 ± 24 145 ± 46 169 ± 32.5 NS 129 ± 28 208 ± 46 119.66 ± 19.3 123.50 ± 38.73x

120.35 ± 42.13xx

123 ± 25 123 ± 25 148 ± 50 157 ± 29.1 NS 210 ± 40 125.29 ± 34.9436 ± 4 38 ± 5 65 ± 15 44 ± 1.6 NS 35 ± 8 22 ± 12 46.32 (46.32

e54.04)36.82 ± 5.45x

38.78 ± 7.69xx

37 ± 4 38 ± 5 61 ± 13 43 ± 12.0 NS 24 ± 14 36.38 ± 7.67241 ± 42 249 ± 42 146 ± 70 156 ± 52.1 NS 218 ± 226 202 ± 38 61.95 ± 26.55 182.26 ± 43.85x

176.39 ± 27.61xx

226 ± 52 237 ± 56 180 ± 66 159 ± 52.0 NS 198 ± 32 170.14 ± 47.54NS NS NS NS NS NS NS NS 161.21 ± 19.74x

155.04 ± 17.94xx

NS NS NS NS NS NS 161.19 ± 19.97117 ± 10 118 ± 10 133 ± 16 142 ± 11.8 NS 107 ± 16 122 ± 14 130 ± 15 127.73 ± 11.96x

130.43 ± 18.59xx

123 ± 10 124 ± 10 134 ± 15 136 ± 9.5 NS 124 ± 12 126.38 ± 15.4372 ± 5 72 ± 6 76 ± 10 91 ± 10.8 NS NS 78 ± 10 74 ± 8 80.95 ± 7.93x

81.82 ± 10.02xx

74 ± 5 74 ± 5 75 ± 9 86 ± 11.0 NS 76 ± 12 80.71 ± 7.481.80 ± 0.60 1.87 ± 0.55 2.03 ± 1.18 1.38 ± 1.56 3.2 ± 1.4 1.7 ± 0.2$ 1.9 ± 1.0 1.28 ± 0.38 1.31 ± 0.33x

1.38 ± 0.51xx

1.84 ± 0.60 1.84 ± 0.56 2.15 ± 0.94 0.57 ± 0.44 3.0 ± 1.2 1.7 ± 0.1$$ NS 1.17 ± 0.33 1.21 ± 0.49

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442 437

Fig. 2. Forest plot detailing weighted mean difference and 95% confidence intervals for the impact of statin therapy on plasma endothelin-1 concentrations. Meta-analysis wasperformed using a random-effect model with inverse variance weighting.

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442438

reducing plasma ET-1 concentrations (WMD: �0.30 pg/mL, 95%CI:�0.47,�0.13; p¼ 0.0004; power¼ 100%) (Fig. 2). This effect sizewas robust in sensitivity analysis and omission of a single study didnot significantly change the overall estimated effect size(Supplemental Figure 1). When the analysis was repeated using thefixed-effects model, significant results were again obtained(WMD: �0.16 pg/mL, 95%CI: �0.23, �0.09; p < 0.00001). In thesubgroup analysis, the effect of statins on plasma ET-1 was signif-icant in both subsets of studies with treatment durations >12weeks (WMD:�0.19 pg/mL, 95%CI:�0.36,�0.02; p¼ 0.03) and<12weeks (WMD: �0.51 pg/mL, 95%CI: �0.89, �0.14, p ¼ 0.008)

Fig. 3. Forest plot detailing weighted mean difference and 95% confidence intervals for the idurations of <12 weeks (above) and >12 weeks (below). Meta-analysis was performed usin

(Fig. 3). With respect to statin dose, a significant reduction ofplasma ET-1 levels was observed with both statin doses <40 mg/day (WMD: �0.27 pg/mL, 95%CI: �0.49, �0.05; p ¼ 0.01) and>40 mg/day (WMD: �0.38 pg/mL, 95%CI: �0.68, �0.08, p ¼ 0.01)(Fig. 4).

3.3. Adjusted indirect meta-analysis

In order to compare the effects of hydrophilic versus lipophilicstatins on plasma ET-1 levels, a subgroup analysis was first con-ducted to estimate the effect size. In the subgroup analysis,

mpact of statin therapy on plasma endothelin-1 concentrations in trials with treatmentg a random-effect model with inverse variance weighting.

Fig. 4. Forest plot detailing weighted mean difference and 95% confidence intervals for the impact of statin therapy on plasma endothelin-1 concentrations in trials with statindoses of <40 mg/day (above) and >40 mg/day (below). Meta-analysis was performed using a random-effect model with inverse variance weighting.

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442 439

lipophilic (comprising 7 treatment arms with atorvastatin, 2 armswith simvastatin, 1 arm with fluvastatin and 1 arm with cer-ivastatin) (WMD:�0.34 pg/mL, 95%CI:�0.55,�0.13; p¼ 0.001) but

Fig. 5. Forest plot detailing weighted mean difference and 95% confidence intervals for the im(above) and hydrophobic (below) statins. Meta-analysis was performed using a random-eff

not hydrophilic (comprising 5 treatment arms with pravastatin)(WMD: �0.18 pg/mL, 95%CI: �0.44, 0.08, p ¼ 0.17) statins had asignificant effect in lowering plasma ET-1 levels (Fig. 5). A superior

pact of statin therapy on plasma endothelin-1 concentrations in trials with hydrophilicect model with inverse variance weighting.

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442440

effect of lipophilic compared with hydrophilic statins was alsoconfirmed in the adjusted indirect comparison, where the effectsize was estimated to be �0.16 pg/mL (95%CI: �0.20, �0.12,Z ¼ 7.62, p < 0.05; power ¼ 100%).

3.4. Meta-regression

The meta-regression analysis was conducted to assess the as-sociation of changes in plasma ET-1 concentrations with dose andduration of statin therapy as potential moderator variables.Consistent with the results of subgroup analysis, the impact ofstatins on plasma ET-1 concentrationswas found to be independentof administered dose (slope: 0.001; 95%CI: �0.008 to 0.010;p ¼ 0.808) and duration of supplementation (slope: �0.002; 95%CI: �0.030 to 0.026; p ¼ 0.888). In addition, no significant associ-ation was found between changes in plasma LDL-C concentrations(slope: 0.004; 95%CI: �0.007 to 0.016; p ¼ 0.439), baseline age(slope: �0.010; 95%CI: �0.039 to 0.019; p ¼ 0.502), and sex (fre-quency of male subjects in each study) (slope: 0.005; 95%CI:�0.007to 0.017; p¼ 0.449) with the changes in plasma ET-1 concentrations(Supplemental Figure 2).

3.5. Publication bias

The funnel plot of the study precision (inverse standard error)by effect size (mean difference) was asymmetric and suggestedpotential publication bias. This observation was further supportedby the results of Begg's rank correlation (Kendall's Tau with conti-nuity correction ¼ �0.39, Z ¼ 2.12, two-tailed p-value ¼ 0.034) andEgger's linear regression (intercept ¼ �2.49, standard error ¼ 0.83;95%CI ¼ �4.26, �0.71, t ¼ 3.01, df ¼ 14.00, two-tailed p ¼ 0.009)tests. The observed publication bias was imputed using trim-and-fill correction. Two potentially missing studies were imputedleading to a corrected effect size that was still significant �0.25 pg/mL (95%CI: �0.42, �0.09). The “fail safe N” method indicated that156 theoretically missing studies would be required to make theoverall estimated effect size non-significant. Funnel plot of theimpact of statins on plasma ET-1 concentrations is illustrated inSupplemental Figure 3.

3.6. Heterogeneity analysis

The meta-analysis indicated a significant heterogeneity basedon the calculated I2 value of 75%, thus supporting the choice ofrandom-effects model. A Galbraith plot was used to identify RCTsthat are outside the pooled 95%CI estimate and might serve aspotential outliers causing heterogeneity. According to the plot, 4RCTs [39e41,46] resided outside the limits of the 95%CI. A secondanalysis excluding these 4 RCTs showed a low inter-study hetero-geneity (I2 ¼ 10%); yet the pooled effect turned out to be marginallysignificant (WMD: �0.08 pg/mL, 95%CI: �0.16e0.00; p ¼ 0.06)(Supplemental Figure 4). This latter analysis yielded significantresults under the fixed-effects model (WMD: �0.08 pg/mL, 95%CI: �0.15e0.00; p ¼ 0.05).

4. Discussion

The present meta-analysis suggests that statin therapy reducesplasma ET-1 concentrations. The efficacy of statins was indepen-dent of therapeutic duration or dose. Since there have been nointervention studies specific for the reduction of ET-1 levels inrelation to the CVD outcomes, the relevance of themean level of thereduction (�0.30 pg/mL) on CVD prevention still remains to bedetermined. Even if so, these findings are of large interest since ET-1 may be a potential therapeutic target for atheroprotection [5].

The robustness of our combined analysis was verified in sensi-tivity analysis and it was found that the significance of the pooledestimate is the result of all studies rather than a single study. Ouranalysis included a study with cerivastatin that is a statin with-drawn from the market. However, when the analysis was repeatedafter excluding the cerivastatin arm, the result remained significant(WMD: �0.26 pg/mL, 95%CI: �0.41, �0.10; p ¼ 0.002). In thisanalysis, although a greater effect size was calculated for the subsetof trials with <12 weeks treatment duration compared with thesubset lasting >12 weeks, no association between treatmentduration and effect size was found in the meta-regression analysis.It may be hypothesized that the greater effect in the subset of trialswith <12 weeks duration might be due to the fact that all studies eexcept one [34] - in this subset used lipophilic statins, which werefound to have a greater effect compared with hydrophilic statins interms of reducing plasma ET-1 levels. In contrast, there were fourtrials [35e38] with hydrophilic statins in the subset of trials with>12 weeks treatment duration.

The biological mechanisms involved in the reduction of ET-1 bystatins are not completely known. Some experimental studiesreport that statins may inhibit ET-1 expression at the transcrip-tional level in vascular endothelial cells [20]. In addition, ET-1 issynthesized by conditions inwhich oxidized LDL, platelet activationand oxidative stress exist [9,10,47]. Statins inhibit these conditions[17,18,48,49]. More studies are required.

ET-1 in the circulation mainly stems from vascular cells [5],while urinary ET-1 is thought to reflect kidney derived production[50]. It has been suggested that urinary ET-1 reflects overallendogenous production of this protein [51]. Besides plasma ET-1measurement, 4 studies included in this meta-analysis measuredurinary ET-1 levels; they confirmed a significant reduction of uri-nary ET-1 during statin therapy [36e38,41].

In addition to cholesterol-lowering effects, the so-called pleio-tropic effects of statins have been the subject of increasing debate[17,18]. These effects may be mainly due to LDL-C reduction (and inturn, plaque stabilization, reduced inflammation, and oxidativestress, etc.) [52]. Variations in these pleiotropic effects mightdecrease residual CVD risk. While there is a correlation between ET-1 and LDL oxidation [8,11], ET-1 concentrations are not clearlycorrelated with LDL-C levels. Future clinical studies are needed todetermine to what degree the reduction of ET-1 is independent ofan anti-oxidative pleiotropic effect of statins.

We show a possible superior effect of lipophilic compared withhydrophilic statins on the reduction of ET-1. There is also additionalevidence (besides data on pravastatin), which shows no significanteffect of rosuvastatin on plasma ET-1 concentrations [52,53]. Adebate exists about the clinical impact of statin lipophilicity[54e59], as disposition of hydrophilic statins could be mediated viaactive transporters [60]. The reasons for the different effects oflipophilic/hydrophilic statins on ET-1 may be due to a wider tissuedistribution with lipophilic statins [61]. Unlike the hepatic tissue,uptake of hydrophilic statins by non-hepatic tissues such asvascular cells and myocardial tissue, as the sources of ET-1, is lessthan lipophilic statins [62,63]. A recent meta-analysis of 13 RCTsindicated that lipophilic statins are better than hydrophilic statinsfor the treatment of heart failure [64]. The issue of statin lip-ophilicity requires further investigation; in our opinion, until thattime lipophilicity should not influence the choice of statin.

This meta-analysis has limitations. The studies included hadrelatively short follow-up durations (2 weekse12 months) andmost had a small number of participants (32e82). Furthermore, inrelation to the study durations, they did not assess long-term CVDoutcomes. The variations in study durations and statin doses mayhave not been of sufficient diversity to assess the impact of thesefactors on the ET-1-lowering effect of statins. Therefore, there is still

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442 441

a need for data from additional trials to identify determinants of ET-1 response to statin therapy, and also the impact of novel LDL-lowering agents [65,66] on plasma ET-1 levels. To address inter-study heterogeneity, a conservative random-effects model wasapplied. In addition, sensitivity analysis confirmed that the pooledestimate is not significantly deviated by a single study.

5. Conclusions

In conclusion, the findings of the present meta-analysis suggestthat statin therapy significantly reduces plasma ET-1 concentra-tions, regardless of treatment duration or statin dose. Statin prop-erties, such as lipophilicity, may affect the level of ET-1 reduction.Larger, well-designed studies with longer follow-up are needed tovalidate our findings, and to determine the parameters that coulddetermine ET-1 response to statin therapy. Whether reduction ofplasma ET-1 concentrations can prevent atherosclerosis and CVevents also requires further investigations.

Authors' contribution

AS e designed the study, made the statistical analysis, correctedthe draft of the paper; KK, CS e designed the study, made theliterature search, drafted the manuscript; SU emade the statisticalanalysis, drafted the manuscript; DPM, SRJ, KKR, MJB, JR, PPT, PM,GYHL e discussed the design of the study, corrected the draft of thepaper;MBe designed the study, made the literature search, draftedthe manuscript, prepared the revised version, submitted the paper.

Acknowledgment

The authors declare that they have no competing interests. Thismeta-analysis was written independently; no company or institu-tion supported it financially. Some of the authors have given talks,attended conferences and participated in trials and advisory boardssponsored by various pharmaceutical companies. No professionalwriter was involved in the preparation of this meta-analysis.

The meta-analysis has been selected for an oral presentation atthe 83rd European Atherosclerosis Society (EAS) Congress inGlasgow during the FREE COMMUNICATIONS e CLINICAL STUDIESsession (Tuesday, March 24, 2015).

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.atherosclerosis.2015.05.022.

References

[1] S. Negi, A. Anand, Atherosclerotic coronary heart disease epidemiology, clas-sification and management, Cardiovasc. Haematol. Disorders-Drug Targets(Formerly Curr. Drug Targets-Cardiovasc. Hematol. Disord. 10 (2010)257e261.

[2] I. Mordi, N. Tzemos, Is reversal of endothelial dysfunction still an attractivetarget in modern cardiology? World J. Cardiol. 6 (2014) 824.

[3] M. Piechota, M. Banach, A. Jaco�n, et al., Natriuretic peptides in cardiovasculardiseases, Cell. Mol. Biol. Lett. 13 (2008) 155e181.

[4] U. F€orstermann, Oxidative stress in vascular disease: causes, defense mecha-nisms and potential therapies, Nat. Clin. Pract. Cardiovasc. Med. 5 (2008)338e349.

[5] J. Pernow, A. Shemyakin, F. B€ohm, New perspectives on endothelin-1 inatherosclerosis and diabetes mellitus, Life Sci. 91 (2012) 507e516.

[6] M. Yanagisawa, H. Kurihara, S. Kimura, et al., A novel potent vasoconstrictorpeptide produced by vascular endothelial cells, Nature 332 (1988) 411e415.

[7] K. Shimada, M. Takahashi, K. Tanzawa, Cloning and functional expression ofendothelin-converting enzyme from rat endothelial cells, J. Biol. Chem. 269(1994) 18275e18278.

[8] R. Irzmanski, M. Banach, M. Piechota, et al., Atrial and brain natriuretic peptideand endothelin-1 concentration in patients with idiopathic arterial hyper-tension: the dependence on the selected morphological parameters, Clin. Exp.

Hypertens. 29 (2007) 149e164.[9] L. Lavie, Obstructive sleep apnoea syndromeean oxidative stress disorder,

Sleep. Med. Rev. 7 (2003) 35e51.[10] I.A. Jagroop, S.S. Daskalopoulou, D.P. Mikhailidis, Endothelin-1 and human

platelets, Curr. Vasc. Pharmacol. 3 (2005) 393e399.[11] H. Morawietz, N. Duerrschmidt, B. Niemann, et al., Induction of the oxLDL

receptor LOX-1 by endothelin-1 in human endothelial cells, Biochem. Biophys.Res. Commun. 284 (2001) 961e965.

[12] K.E. Meyers, C. Sethna, Endothelin antagonists in hypertension and kidneydisease, Pediatr. Nephrol. 28 (2013) 711e720.

[13] R. Schaefer, Lipid management for the prevention of cardiovascular disease,Curr. Pharm. Des. 17 (2011) 852e860.

[14] P.P. Toth, Low-density lipoprotein reduction in high-risk patients: how low doyou go? Curr. Atheroscler. Rep. 6 (2004) 348e352.

[15] M. Banach, D.P. Mikhailidis, S.E. Kjeldsen, et al., Time for new indications forstatins? Med. Sci. Monit. 15 (2009) MS1eMS5.

[16] M. Banach, W.S. Aronow, C. Serban, et al., Lipids, blood pressure and kidneyupdate 2014, Pharmacol. Res. 95 (2015) 111e125.

[17] M. Ii, D.W. Losordo, Statins and the endothelium, Vasc. Pharmacol. 46 (2007)1e9.

[18] A. Blum, R. Shamburek, The pleiotropic effects of statins on endothelialfunction, vascular inflammation, immunomodulation and thrombogenesis,Atherosclerosis 203 (2009) 325e330.

[19] A. Sahebkar, G.T. Chew, G.F. Watts, Recent advances in pharmacotherapy forhypertriglyceridemia, Prog. Lipid Res. 56 (2014) 47e66.

[20] O. Hern�andez-Perera, D. P�erez-Sala, J. Navarro-Antolín, et al., Effects of the 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors, atorvastatin and simva-statin, on the expression of endothelin-1 and endothelial nitric oxide synthasein vascular endothelial cells, J. Clin. Investig. 101 (1998) 2711.

[21] D. Moher, A. Liberati, J. Tetzlaff, et al., Preferred reporting items for systematicreviews and meta-analyses: the PRISMA statement, BMJ 339 (2009) b2535.

[22] A.R. Jadad, R.A. Moore, D. Carroll, et al., Assessing the quality of reports ofrandomized clinical trials: is blinding necessary? Control. Clin. Trials 17(1996) 1e12.

[23] D. Moher, D. Cook, A. Jadad, et al., Assessing the quality of reports of rando-mised trials: implications for the conduct of meta-analyses, Health Technol.Assess. Winch. Engl. 3 (1998) ieiv, 1e98.

[24] M. Borenstein, L. Hedges, J. Higgins, et al., Comprehensive Meta-analysisVersion 2, Biostat, Englewood, NJ, 2005, 104e104.

[25] A.J. Sutton, K.R. Abrams, D.R. Jones, et al., Methods for Meta-analysis inMedical Research, J. Wiley, 2000.

[26] A. Sahebkar, Does PPARg2 gene Pro12Ala polymorphism affect nonalcoholicfatty liver disease risk? Evidence from a meta-analysis, DNA Cell Biol. 32(2013) 188e198.

[27] W.D. Dupont, W.D. Plummer, Power and sample size calculations for studiesinvolving linear regression, Control. Clin. Trials 19 (1998) 589e601.

[28] F. Song, D.G. Altman, A.-M. Glenny, et al., Validity of indirect comparison forestimating efficacy of competing interventions: empirical evidence frompublished meta-analyses, Bmj 326 (2003) 472.

[29] H.C. Bucher, G.H. Guyatt, L.E. Griffith, et al., The results of direct and indirecttreatment comparisons in meta-analysis of randomized controlled trials,J. Clin. Epidemiol. 50 (1997) 683e691.

[30] S. Duval, R. Tweedie, Trim and fill: a simple funnel-plotebased method oftesting and adjusting for publication bias in meta-analysis, Biometrics 56(2000) 455e463.

[31] A. Abou-Raya, S. Abou-Raya, M. Helmii, Statins as immunomodulators insystemic sclerosis, Ann. N. Y. Acad. Sci. 1110 (2007) 670e680.

[32] A. Åsberg, H. Holdaas, A.G. Jardine, et al., Fluvastatin reduces atherogeniclipids without any effect on native endothelial function early after kidneytransplantation, Clin. Transplant. 17 (2003) 385e390.

[33] A.L. Barsuk, VAM, E.V. Malinok, L.V. Lovtsova, I.E. Okrut, V.B. Kuzin, The dy-namics and interaction of endothelium state and immune status variations inprogram hemodialysis patients against the background of atorvastatin ther-apy, Mod. Thechnol. Med. 5 (2013) 78e83.

[34] J. Dupuis, J.-C. Tardif, P. Cernacek, et al., Cholesterol reduction rapidly im-proves endothelial function after acute coronary syndromes the RECIFE(reduction of cholesterol in ischemia and function of the endothelium) trial,Circulation 99 (1999) 3227e3233.

[35] N. Glorioso, C. Troffa, F. Filigheddu, et al., Effect of the HMG-CoA reductaseinhibitors on blood pressure in patients with essential hypertension andprimary hypercholesterolemia, Hypertension 34 (1999) 1281e1286.

[36] T.-M. Lee, S.-F. Su, C.-H. Tsai, Effect of pravastatin on proteinuria in patientswith well-controlled hypertension, Hypertension 40 (2002) 67e73.

[37] T. Lee, C. Chen, H. Shen, et al., Effects of pravastatin on functional capacity inpatients with chronic obstructive pulmonary disease and pulmonary hyper-tension, Clin. Sci. 116 (2009) 497e505.

[38] T.-M. Lee, M.-S. Lin, C.-H. Tsai, et al., Add-on and withdrawal effect of pra-vastatin on proteinuria in hypertensive patients treated with AT1 receptorblockers, Kidney Int. 68 (2005) 779e787.

[39] J. Lewandowski, M. Si�nski, J. Bidiuk, et al., Simvastatin reduces sympatheticactivity in men with hypertension and hypercholesterolemia, Hypertens. Res.33 (2010) 1038e1043.

[40] J.-J. Li, C.-H. Fang, C. Wang, et al., Effects of simvastatin on exercise-inducedmyocardial ischemia and plasma endothelin-1 concentrations in patientswith stable angina, Clin. Chim. acta 354 (2005) 205e208.

A. Sahebkar et al. / Atherosclerosis 241 (2015) 433e442442

[41] T. Nakamura, C. Ushiyama, K. Hirokawa, et al., Effect of cerivastatin on urinaryalbumin excretion and plasma endothelin-1 concentrations in type 2 diabetespatients with microalbuminuria and dyslipidemia, Am. J. Nephrol. 21 (2001)449e454.

[42] D. Mozaffarian, E. Minami, R.A. Letterer, et al., The effects of atorvastatin (10mg) on systemic inflammation in heart failure, Am. J. Cardiol. 96 (2005)1699e1704.

[43] S. Tehrani, F. Mobarrez, P.-E. Lins, et al., Impaired endothelium-dependentskin microvascular function during high-dose atorvastatin treatment in pa-tients with type 1 diabetes, Diabetes Vasc. Dis. Res. 10 (2013) 483e488.

[44] P. Usharani, A.A. Mateen, M.U.R. Naidu, et al., Effect of NCB-02, atorvastatinand placebo on endothelial function, oxidative stress and inflammatorymarkers in patients with type 2 diabetes mellitus, Drugs R. D. 9 (2008)243e250.

[45] P.A. Economides, A. Caselli, E. Tiani, et al., The effects of atorvastatin onendothelial function in diabetic patients and subjects at risk for type 2 dia-betes, J. Clin. Endocrinol. Metabol. 89 (2004) 740e747.

[46] P. Usharani, A. Mateen, M. Naidu, et al., Effect of NCB-02, atorvastatin andplacebo on endothelial function, oxidative stress and inflammatory markers inpatients with type 2 diabetes mellitus, Drugs R. D. 9 (2008) 243e250.

[47] F.C. Tanner, C.M. Boulanger, T.F. Lüscher, Endothelium-derived nitric oxide,endothelin, and platelet vessel wall interaction: alterations in hypercholes-terolemia and atherosclerosis, Seminars Thromb. Hemost. (1992) 167e175.

[48] A. Ludman, V. Venugopal, D.M. Yellon, et al., Statins and car-dioprotectiondmore than just lipid lowering? Pharmacol. Ther. 122 (2009)30e43.

[49] C.G. Mihos, A.M. Pineda, O. Santana, Cardiovascular effects of statins, beyondlipid-lowering properties, Pharmacol. Res. 88 (2014) 12e19.

[50] D.E. Kohan, Endothelins in the normal and diseased kidney, Am. J. Kidney Dis.29 (1997) 2e26.

[51] B. Battistini, P. D'Orleans-Juste, P. Sirois, Endothelins: circulating plasma levelsand presence in other biologic fluids, Lab. Investig. J. Tech. Methods Pathol. 68(1993) 600.

[52] H. Krum, E. Ashton, C. Reid, et al., Double-blind, randomized, placebo-controlled study of high-dose HMG CoA reductase inhibitor therapy on ven-tricular remodeling, pro-inflammatory cytokines and neurohormonal pa-rameters in patients with chronic systolic heart failure, J. Card. Fail 13 (2007)1e7.

[53] H.Y. Qu, Y.W. Xiao, G.H. Jiang, et al., Effect of atorvastatin versus rosuvastatinon levels of serum lipids, inflammatory markers and adiponectin in patientswith hypercholesterolemia, Pharm. Res. 26 (2009) 958e964.

[54] T. Sakamoto, S. Kojima, H. Ogawa, et al., Usefulness of hydrophilic vs lipophilic

statins after acute myocardial infarction: subanalysis of MUSASHI-AMI, Circ. J.Off. J. Jpn. Circ. Soc. 71 (2007) 1348e1353.

[55] M. Fujita, T. Yamazaki, D. Hayashi, et al., Pleiotropic effects of statins on car-diovascular events in the Japanese coronary artery disease study, Int. J. Car-diol. 129 (2008) 294e296.

[56] T. Kai, S. Arima, Y. Taniyama, et al., Comparison of the effect of lipophilic andhydrophilic statins on serum adiponectin levels in patients with mild hy-pertension and dyslipidemia: Kinki adiponectin interventional (KAI) Study,Clin. Exp. Hypertens. 30 (2008) 530e540.

[57] T. Tsutamoto, M. Yamaji, C. Kawahara, et al., Effect of simvastatin vs. rosu-vastatin on adiponectin and haemoglobin A1c levels in patients with non-ischaemic chronic heart failure, Eur. J. heart Fail. 11 (2009) 1195e1201.

[58] M.C. Kim, Y. Ahn, S.Y. Jang, et al., Comparison of clinical outcomes of hydro-philic and lipophilic statins in patients with acute myocardial infarction,Korean J. Intern. Med. 26 (2011) 294e303.

[59] J.R. Murrow, S. Sher, S. Ali, et al., The differential effect of statins on oxidativestress and endothelial function: atorvastatin versus pravastatin, J. Clin. Lipidol.6 (2012) 42e49.

[60] K.M. Bailey, S.P. Romaine, B.M. Jackson, et al., Hepatic metabolism andtransporter gene variants enhance response to rosuvastatin in patients withacute myocardial infarction: the GEOSTAT-1 study, Circ. Cardiovasc Genet. 3(2010) 276e285.

[61] R.P. Mason, M.F. Walter, C.A. Day, et al., Intermolecular differences of 3-hydroxy-3-methylglutaryl coenzyme a reductase inhibitors contribute todistinct pharmacologic and pleiotropic actions, Am. J. Cardiol. 96 (2005)11e23.

[62] A. Tomizawa, Y. Hattori, K. Suzuki, et al., Effects of statins on vascular endo-thelial function in hypercholesterolemic patients with type 2 diabetes melli-tus: fluvastatin vs. rosuvastatin, Int. J. Cardiol. 144 (2010) 108e109.

[63] K. Nezasa, K. Higaki, T. Matsumura, et al., Liver-specific distribution of rosu-vastatin in rats: comparison with pravastatin and simvastatin, Drug Metab.Dispos. 30 (2002) 1158e1163.

[64] G. Liu, X.-X. Zheng, Y.-L. Xu, et al., Effects of lipophilic statins for heart failure:a meta-analysis of 13 randomised controlled trials, Heart, Lung Circulation 23(2014) 970e977.

[65] A. Sahebkar, G.F. Watts, New LDL-cholesterol lowering therapies: pharma-cology, clinical trials, and relevance to acute coronary syndromes, Clin. Ther.35 (2013) 1082e1098.

[66] S. Dragan, M.-C. Serban, M. Banach, Proprotein convertase Subtilisin/Kexin 9inhibitors an emerging lipid-lowering therapy? J. Cardiovasc. Pharmacol. Ther.20 (2015) 157e168.