estatina en ckd bmj 2009

8/6/2019 Estatina en CKD BMJ 2009

1/14


2/14

large trial in patients with diabetes whoneeded dialysisdid not show a significant reduction in cardiovascularend points with statin treatment.13

International guidelines reflect these uncertainties;some recommend that cholesterol concentrationsshould be lowered in chronic kidney disease,1415 but

others suggest that more data are needed.1617

In tworecent meta-analyses, claims of improved renal out-comeshavebeenmade,encouragingbroaderadoptionof statins in patients with pre-dialysis (stages I-IV)chronic kidney disease.1819 Douglas et al reported thebeneficial effects of statins in patients with overtproteinuria (protein excretion >300 mg/day) but notin thosewith microalbuminuriaor normoalbuminuria.Similarly,Sandhuetalreportedaslowingofthedeclinein glomerular filtration rate in patients with chronickidney diseaseand concomitant cardiovasculardisease(0.93 ml/min/year slower than in control patients),along with improvement in proteinuria. Neither of

thesereviews intendedto analyse theeffect of statins onlipid concentrations or mortality in patients withkidney disease specifically on the basis of the stage ofk i d n ey d i s ea s e ( p r e- d i al y s is , d i a ly s i s, a n dtransplantation).1819 The aim of our study was toevaluate the efficacy and safety of statins for renal andcardiovascular outcomesin all stages of chronic kidneydisease (pre-dialysis, dialysis, and transplantation).

METHODS

Inclusion criteria

We included all randomised controlled trials andquasi-randomised controlled trials (method of allocat-

ing participants to different forms of care that is nottruly random) of any statin against placebo, notreatment, or another statin in adult patients withchronickidneydisease,includingsubstudiesofpatientswithchronickidneydiseaserecruitedwithinlargescalestatin trials. We defined patients with chronic kidneydisease as those whowere having maintenancedialysistreatment, had had renal transplantation, had anelevated baseline mean serum creatinine (>1.4 mg/dl(0.121 mmol/l) or as defined by authors), or had animpairment of the glomerular filtration rate as definedby the kidney disease outcome quality initiativeguidelines with values of glomerular filtration rate

60 ml/min/1.73m2 along with other markers of kidney damagesuch asproteinuria.20 We excluded studies with patientsdefined as havingrenal impairment but where wecould not ascertain the baseline glomerular filtrationrate or creatinine concentration. We included onlyrandomised controlled trials of at least eight weeksduration, because studies with a shorter follow-upperiod would not allow detectionof significant changesin lipid concentrations or other major patient leveloutcomes. We excluded studies in participants withimpaired liver function tests or elevated creatininephosphokinase concentrations and those that exam-

ined the efficacy of combination treatments with otherlipid lowering agents.

Data sources and searches

We searched Medline, Embase, the Cochrane CentralRegister, and the Renal Health Library of theCochrane Renal Group in July 2006 with thefollowingterms: chronic kidney disease, dialysis, haemodialysis,peritoneal dialysis, renal transplantation, kidney trans-plantation, acute renal allograft rejection, renal allo-g r a f t r e j e c ti o n , h y p e r ch o l e st e r o la e m ia ,hyperlipidaemia, dyslipidaemia, HMG-CoA reduc-tase inhibitors, and statins. We also used text words forallstatins.Thesearchstrategyusedisavailablefromtheauthors. We considered randomised controlled trialswithout language restriction. We searched the refer-ence lists of identified trials and review articles foradditional trials. We sought information about unpub-lished and ongoing randomised controlled trials fromauthors of the included randomised controlled trials,drug companies, and experts in the field.

Data collection and analysis

Two authors independently reviewed literaturesearchestoidentifyrelevanttrialsthatmettheinclusioncriteria. We extracted data on study sample, popula-tion characteristics (age, race, sex, and baselinecomorbidities), interventions (type and dose of statin,duration of treatment), co-interventions (administeredin a non-randomised fashion), and the methodologicalquality of the trials. We extracted data on the followingoutcomes: all cause mortality; fatal cardiovascular andcerebrovascular events; non-fatal cardiovascular andcerebrovascular events (defined as myocardial infarc-tion, stroke, sudden death, or composites of these endpoints) separately; end stage renal disease; doubling of

serum creatinine concentration;lipid concentrationsatthe end of treatment (total cholesterol, low densitylipoprotein cholesterol, high density lipoprotein cho-lesterol, triglycerides); creatinine clearance (ml/min orml/min/1.73 m2) and 24 hour urinary protein excre-tion (g/24 h) (in trials of patients with pre-dialysischronic kidney disease who did not need renalreplacement treatment); acute allograft rejection rates(in trials of transplant recipients); adverse events(including elevated liver function tests and elevatedcreatinine kinase concentrations); and missing patientdata owing to withdrawals, non-intention to treatanalysis, and loss to follow-up. Some trials included

different populations (pre-dialysis, dialysis, transplant)and thus contributed to more than one analysis, so wehave separated the individual comparisons for eachpopulation of interest and included them in therelevant meta-analysis by comparisons or subsetsrather than for trials.

We used relative risk with 95% confidence intervalsto analyse dichotomous data (all cause mortality, fataland non-fatal cardiovascular or cerebrovascularevents, acute allograft rejection, and adverse events).We pooled risk estimates from individual trials byusing the Der Simonian-Laird random effects model.21

Where continuous measurements of outcomes were

used (24 hour urinary protein excretion, creatinineclearance, lipid concentrations), we calculated the

R E S E A R C H

page 2 of 14 BMJ | ONLINE FIRST | bmj.com


3/14

weighted mean difference with 95% confidence inter-vals by using the values of the outcome at the end oftreatment. We used end of treatment data because fewtrials reported changes in the value of these outcomesfrom beginning to end of treatment with a varianceestimator. We contacted authors for clarification about

end points (such as mortality, cardiovascular events,renal function, lipid parameters, toxicity) up to threetimes.

We used the heterogeneity 2 (Cochran Q) statisticandtheI2 test to formally analyse heterogeneity acrossincluded trials.22 We did subgroup analysis andrandom effects univariate and multivariate meta-regression where possible to explore the role ofpotential sources of heterogeneity related to theparticipants (age, race, sex, stage of chronic kidneydisease (pre-dialysis, dialysis, or transplant), and base-line risk covariates), the agent used (different types anddoses of statins, different duration of treatment and

modes of administration) and trial quality (allocationconcealment, blinding, use of intention to treatanalysis, and proportions of patients lost to follow-up)on the effect of the interventions. We always analysedtrials done in pre-dialysis, dialysis, and transplantpopulations separately and pooled the results whenformal tests of interaction indicated no significant

difference between the estimates from the separategroups.WeusedRevManversion4.2.10andSAS9.1.2for analyses.

Quality assessment

At least two authors independently assessed the

methodological quality of included randomised con-trolled trials by using standard domains: allocationconcealment (adequate if sequentially labelled, sealedand opaque envelopes or central or pharmacy rando-misation was used; inadequate when pseudo-randomisation was used; unclear in all other cases);blinding of investigators, participants, and outcomeassessors; use of intention to treat analysis; complete-nessof follow-up. Discrepanciesin dataextractionwereresolved by discussion among authors. Where morethan one publication of a trial existed, we includedonlythe publication with the most complete data.

RESULTS

Search results

The combined search of Medline, Embase, and theCochrane Central Register of randomised controlledtrials identified 869 articles, of which we excluded 801because they were not randomised controlled trials orthey evaluated interventions that were not relevant tothis review (such as fibrates or other lipid loweringagents). Full text assessment of the remaining 68potentially relevant articles resulted in identificationof 50 eligible randomised controlled trials with 54comparisons of statins versus placebo or no treatment(comparisons: pre-dialysis, n=26; dialysis, n=11;transplant, n=17), which included 30 144 patients

(fig 1). 13w1-w49

Trial characteristics

Twenty six randomised controlled trials or subsets ofrandomised controlled trials in pre-dialysis chronickidney disease populations, 11 randomised controlledtrials or subsets of randomised controlled trials inhaemodialysis and peritoneal dialysis patients, and 17randomised controlled trials or subsets of randomisedcontrolledtrialsinrenaltransplantrecipientsevaluatedthe effects of statins versus placebo. All trials studiedthe effect of statins on lipid concentrations and safety.Of the 44 trials that reported mortality, the pravastatin

pooling project (PPP, a subgroup analysis of patientswithrenalinsufficiencyinthecholesterolandrecurrentevents (CARE) trial, long term intervention withpravastatin in ischemic disease (LIPID) study, andwest of Scotland coronary prevention project study(WOSCOPS)),w10 the atorvastatin in patients with type2 diabetes mellitus undergoing hemodialysis (4D)trial,13 and the assessment of Lescol in renal transplan-tation (ALERT) trialw42 provided most event data. Wetreated the pravastatin pooling project as one trialbecause we could not get separate trial level data.

The 26 randomised controlled trials or subsets ofrandomised controlled trials in pre-dialysis chronic

kidney disease enrolled participants with diabeticnephropathy (n=6), hypertensive nephropathy (n=2),

Overall result of database searches (n=869)

No of trials by outcome:All cause mortality (44 trials, 23 665 patients)Cardiovascular mortality (43 trials, 23 266 patients)Non-fatal cardiovascular events (8 trials, 22 863 patients)Lipid concentration (42 trials, 6390 patients)Acute allograft rejection (5 trials, 639 patients)24 hour urinary protein excretion (6 trials, 311 patients)Creatinine clearance (11 trials, 548 patients)Elevated liver function tests (26 trials, 6726 patients)Rhabdomyolysis (29 trials, 6829 patients)Withdrawal owing to adverse events (20 trials, 4887 patients)

Renal Health Library: Pre-dialysis (n=38)

Dialysis (n=14)Transplant (n=19)

Embase: Pre-dialysis (n=367)

Dialysis (n=67) Transplant (n=118)

Medline: Pre-dialysis (n=146)

Dialysis (n=32)Transplant (n=68)

Full text analysis (n=68):Pre-dialysis (n=22)Dialysis (n=16)Transplant (n=30)

Excluded (n=801):Search overlap (n=104)Non-randomised trials, or

reviews (n=693)Duration


4/14

or various forms of nephrotic and non-nephroticglomerulonephritis. One trial (n=20 patients) wasdone in patients with polycystic kidney disease.w12

Fifteen of these randomised controlled trials evaluatedthe potential renoprotective effect of statins, particu-larly their effects on 24 hour urinary protein excretion.In essence, most of these studies enrolled patients withearly or late stagesof chronic kidneydisease andwith ahistory of coronary heart disease. These studies didnotinclude patients with moderate chronic kidney diseasebut without cardiac disease.

Five randomised controlled trials in renal transplantrecipients reported biopsy proved acute allograftrejection rates, but no trial reported the effects onchronic allograft nephropathy.w35-w39 Liver functiontests and creatinine phosphokinase concentrationswere the only safety parameters reported consistentlyin all trials. Follow-up ranged from two months to60 months in alltrials,and several differentstatins wereused. The UK-HARP study included pre-dialysis,dialysis, and renal transplant patients,w13 and thestudy by Stegmayr et al included both pre-dialysisand dialysis patients.w8 Table 1 details additionalcharacteristics of the populations and interventions inthe included randomised controlled trials.

Trial quality

The methodological quality of many trials wassuboptimal. Concealment of allocation was adequatein 11 (22%) randomised controlled trials, clearlyinadequate in 9 (18%), and unclear in the remainder.Participants, investigators, and outcomeassessors wereblinded in only 10 (20%) randomised controlled trials,

and only 10 (20%) randomised controlled trials wereanalysedonanintentiontotreatbasis.Thedropoutratewas less than 10% in 43 (86%) randomised controlledtrials, between 10% and 19% in 4 (8%), 20-39% in 2(4%), and over 40% in only 1 (2%) trial.

Effect of statins on surrogate end points in chronic kidney

disease

Total cholesterolTotal cholesterol concentrations were significantlylower with statins than with placebo (42 comparisons,6390 patients; weighted mean difference42.28 mg/dl(1.10 mmol/l), 95% confidence interval 47.25 to

37.32) (table 2). We found significant heterogeneity

for this outcome (heterogeneity 2=804.09, I2=94.9%),which was largely explained by the type of statin(38.42% of heterogeneity) and the baseline cholesterolconcentration (60.73% of heterogeneity; greater differ-ences in patients with higher baseline values). Stage ofchronic kidney disease was not an effect modifier. Thecholesterol lowering effect of atorvastatin and ceriva-statin was significantly higher than that of other statins(such as pravastatin or fluvastatin) and increasedproportionally with increasing baseline cholesterolconcentrations (model 2, table 3).

Low density lipoprotein cholesterolLow density lipoprotein cholesterol concentrations

were significantly lower with statins than with placebo(39 comparisons, 6216 patients; weighted meandifference 43.12 mg/dl (1.12 mmol/l), 47.85 to38.40) (table 2). We found significant evidence ofheterogeneity between agents (for example, ceriva-statin v pravastatin or fluvastatin). The type of statinexplained38.42%ofexistingheterogeneityintheeffectof statins on total cholesterol concentrations at the endof treatment and 10.99% of that on low densitylipoprotein cholesterol (model 2, table 3).

High density lipoprotein cholesterolStatins had no significant effect comparedwith placeboon the concentration of high density lipoproteincholesterol in chronic kidney disease (40 comparisons,5621 patients; weighted mean difference 0.41 mg/dl(0.01 mmol/l), 0.78 to 1.60) (table 2). We foundconsiderable heterogeneity (heterogeneity 2=693.78,I2=94.4%), but this was not explained by stage ofchronic kidney disease.

TriglyceridesWe found a significant reduction in triglycerideconcentrations withstatins in comparison withplacebo(39 comparisons, 5569 patients; weighted meandifference 23.71 mg/dl (0.23 mmol/l), 33.52 to13.90) overall (table 2) and separately in pre-dialysispatients (15 comparisons, 836 patients; 28.71 mg/dl(0.28 mmol/l), 48.55 to 8.87) and transplantpatients (11 comparisons, 2955 patients; 25.24mg/dl (0.25 mmol/l), 33.49 to 16.99) but not indialysis patients (13 comparisons, 1778 patients;22.67 mg/dl (0.22 mmol/l), 46.80 to 1.46).

Proteinuria and creatinine clearanceWe found a significant reduction in 24 hour urinaryprotein excretion (g/24 h) in chronic kidney disease(pre-dialysis) patients receiving statins compared withplacebo (6 randomised controlled trials, 311 patients;weighted mean difference 0.73 g/24 hour, 0.95 to0.52), with significant heterogeneity in this analysis(heterogeneity 2=12.07, I2=58.6%) (table 2). Creati-nine clearance (either in ml/min or ml/min/1.73 m2)didnot changewith statins in comparison with placebo(11 randomised controlled trials, 548 patients;weighted mean difference 1.48 ml/min (0.024 ml/s),

2.32 to 5.28).

Effect of statins on patient level end points in chronic

kidney disease

All cause and cardiovascular mortality and non-fatalcardiovascular eventsWe found no significant reduction in the risk of allcause mortality with statins in chronic kidney diseaseoverall (44 trials, 23 665 patients; relative risk 0.92,95% confidence interval 0.82 to 1.03). A significant(approximately 20%) reductionin risk occurred in pre-dialysis patients (21 trials, 18 781 patients; relative risk0.81, 0.74 to 0.89), largely driven by the pravastatin

poolingproject,butwefoundnostatisticallysignificantinteraction betweenthe separategroupsof pre-dialysis,

R E S E A R C H



5/14

Table 1 | Characteristicsof populations,interventions, andoutcomes in includedtrials

Study IDIntervention(statin)

Dose(mg/day)

Studyduration(months)

No ofpatients

Mean(SD) baselinecholesterol (mg/dl)

Mean (SD) baselineglomerular filtration rate

(ml/min/1.73 m2)

Outcomes

Plasmalipids

Mortality, CV events(fatal, non-fatal) Renal

Sideeffects

Pre-dialysis

Yutaka 99w12* Pravastatin 20 6 57 247.5 (5.3) 64.4 (5.4) Yes No Yes No

Zhang 95w15

Pravastatin 20 3 20 NA NA Yes Yes No YesLee 02w22 Pravastatin 10 6 63 210 (23) 85 (16) Yes Yes Yes No

Lee 05w16 Pravastatin 10 6 82 208 (23) 90 (19) Yes Yes Yes Yes

PPP 04w10 Pravastatin 40 60 16 824 NA 30.0-89.9 Yes Yes No No

PREVEND IT 04w25 Pravastatin 40 48 864 193.3 (33.3) NA Yes Yes Yes Yes

Gheith 02w2* Fluvastatin 20 12 43 NA 98.8 (37) Yes Yes Yes Yes

Lemos 05w26 Fluvastatin 80 36-48 310 200.0 (33.0) 47.0 (7.0) Yes Yes No No

Buemi 00w20 Fluvastatin 40 6 21 NA NA Yes Yes Yes No

Yasuda 04w19 Fluvastatin 20 12 80 265.3 (5.7) 59 (5) Yes Yes Yes No

Lam 95w4 Lovastatin 20 24 34 210.0 (3.3) 83.1 (9.5) Yes Yes Yes No

Hommel 92w3 Simvastatin 10 3 26 226.6 (33.3) 64.0 (30.0) Yes Yes Yes Yes

Thomas 93w9 Simvastatin 10 6 30 315.3 (79.0) 76.5 (36.5) Yes Yes No Yes

Nielsen 93w6 Simvastatin 10-20 9 18 223.3 (10.0) 96.6 (8.0) Yes Yes Yes No

Rayner 95w7 Simvastatin 10-40 24 17 NA 84.3 (10.5) Yes Yes Yes Yes

Tonolo 97w11* Simvastatin 20 12 19 223.3 (10.0) 97.0 (7.0) Yes Yes Yes No

HARP 05w13 Simvastatin 20 12 241 NA 62.0 (55.4) Yes No No Yes

Van dijk 01w14* Simvastatin 40 2 20 NA >50.0 Yes Yes Yes Yes

Scanferla 91w18 Simvastatin 10 12 24 263 (22) 41 Yes Yes Yes No

Fried 01w17 Simvastatin 10-20 24 36 191.8 (21.5) NA Yes Yes Yes No

HPS 03w24 Simvastatin 40 57 5963 NA NA Yes No No

Stegmayr05w8 Atorvastatin 10 24 14 NA


6/14

dialysis, and transplant patients (heterogeneity

2=16.05, I2=25.2%, P=0.12 for interaction), suggesting

that stage of chronic kidney disease is not a provedeffect modifier, which makes the overall estimate ofeffect the most robust (fig 2).

We found a significant (approximately 20%) reduc-

tion in the risk of cardiovascular mortality (43 trials,23 266patients; relative risk 0.81, 0.73 to 0.90), with nostatistically significant heterogeneity (heterogeneity

2=8.45, I2=0%, P=0.23 for interaction) (fig 3) and no

apparent difference in treatment effect across pre-dialysis, dialysis, and transplant populations. Com-pared with placebo, statins also significantly decreasedthe risk of non-fatal cardiovascular events by 20% (8trials, 22 863 patients, relative risk 0.78, 0.73 to 0.84),with no significant heterogeneity among the studies(heterogeneity2=7.68, I2=8.9%) (fig 4). This effect wasconsistentacross pre-dialysis and dialysispatients,withno significant interaction (P=0.18 for interaction).

End stage renal disease and allograft rejectionWe found no trials reporting end stage renal disease ordoubling of creatinine as an outcome. No significantreduction occurred in the risk of acute allograftrejection with statins used for three months in theimmediate post-transplant period compared withplacebo (5 trials, 639 patients; relative risk 0.73, 0.49to 1.10) (fig 5). We found a significant heterogeneityamong these trials (heterogeneity 2=8.97, I2=55.4%),which could be explained by the different sample sizesin the included studies, the different subsets of patientsincluded (cadaveric and living related transplantrecipients), and the different immunosuppressive regi-mens used. These studies did not reporton the effect ofstatins on chronic allograft nephropathy.

Adverse effectsWe found no significant increase in the risk ofabnormalities in liver function tests (26 trials, 6726patients) or raised creatinine phosphokinase concen-trations (risk of rhabdomyolysis) (29 trials, 6829patients; relative risk 1.50, 0.86 to 2.59; heterogeneity

2=7.73, I2=0%) for statins compared with placebo. We

also found no significant difference in the risk ofwithdrawal from the study owing to adverse events forstatins comparedwith placebo (20 trials, 4887 patients;

relative risk 1.03, 0.84 to1.25; heterogeneity 2=17.79,I2=10%) (fig 6).

Analysis of heterogeneity

Meta-regression of heterogeneitywas possible only forthe outcome of total cholesterol and low density

lipoprotein cholesterol at end of treatment, as allother outcomes were reported in too few trials foranalysis to be robust. On univariate meta-regression,the type of statin (38.42%), baseline cholesterolconcentrations (60.73%), and allocation concealment(13.80%) seemed to be responsible for most of theheterogeneity in the effect of statins on totalcholesterolconcentrations; statin type (10.99%), age (19.39%),baseline cholesterol concentrations (34.43%), andallocationconcealment(20.05%)werethemajorcausesof heterogeneity in the analysis of effect of statins onlow density lipoprotein cholesterol concentrations(table 3). The observed significant heterogeneity bytype of statin may be largely explained by the singletrial of cerivastatin.w5 When we excludedthis trial fromthe analyses, heterogeneity became non-significant.Trials with adequate concealment of allocation hadmore precise estimates (total cholesterol: weightedmean difference 38.16 mg/dl, 49.59 to 26.73; lowdensity lipoprotein cholesterol: 36.67 mg/dl, 48.23to 25.10), whereas those with inadequate allocationconcealment (total cholesterol: 55.01 mg/dl, 69.19to 40.83; low density lipoprotein cholesterol:59.93 mg/dl, 76.31 to 43.54) or unclear allocationconcealment (total cholesterol: 52.06 mg/dl, 62.93to 41.19; low density lipoprotein cholesterol:49.44 mg/dl, 60.64 to 38.24) tended to suggestgreater benefits for the experimental intervention(statin).

Table 4 shows the results of a multivariate meta-regression analysis in which we analysed potentialsources of heterogeneity in the effects of statinscompared with placebo on total cholesterol and lowdensity lipoprotein cholesterol at the end of treatmentafter accounting for any covariate that was clinicallyrelevant or statistically significant on univariate analy-sis. For both total cholesterol and low densitylipoprotein cholesterol concentrations at the end oftreatment, the model including the type of statin, doseof statin, age of participants, allocation concealment,

Table 2 | Effectsof statinson lipidconcentrationsand renal functionin patientswith chronickidneydisease

Outcome No of trials (No of patients)Weighted mean difference (95%

CI) Heterogeneity (I2; %)

Total cholesterol (mg/dl)* 42 (6390) 42.28 (47.25 to 37.32) 94.9

LDL cholesterol (mg/dl)* 39 (6216) 43.12 (47.85 to 38.40) 94.1

HDL cholesterol (mg/dl)* 40 (5621) 0.41 ( 0.78 to 1.60) 94.4

Triglycerides (mg/dl) 39 (5569) 23.71 (33.52 to 13.90) 89.5

Glomerular filtration rate (ml/min orml/min/1.73 m2)

11 (548) 1.48 ( 2.32 to 5.28) 62.0

U rinary protein excre tion (g/24 h) 6 (311) 0.73 (0.95 to 0.52) 58.6

HDL=high density lipoprotein; LDL=low density lipoprotein.*Multiply by 0.0259 to convert to mmol/l.Multiply by 0.01 to convert to mmol/l.

R E S E A R C H



7/14


8/14

and baseline cholesterol concentrations explainedmost of the identified heterogeneity in the effect ofstatins versus placebo. The effect of statins on totalcholesterol and lowdensity lipoprotein cholesterol wasdose dependent.

DISCUSSION

Key findings

Ourmeta-analysisfoundthatstatinsareassociatedwithlipid lowering, cardiovascular, and antiproteinuricbenefits in chronic kidney disease. They seem to besafeinchronickidneydisease,withrespecttotheriskof

rhabdomyolysis and hepatotoxicity and because lim-ited withdrawals occurred in the treatment group. Therisk of cardiovascular events and cardiovascularmortality is reduced by statin treatment in people atdifferent stages of chronic kidney disease (pre-dialysis,dialysis, and transplant), and the magnitude of cardio-

vascular benefit achieved seems broadly similar inthese groups and approximates that of statin treatmentin other populations.23

Although statins have been conclusively shown toreduce cardiovascular mortalityand allcause mortalityin the general population,8-10 considerable uncertainty

Pre-dialysis patients

Rayner 1995

PPP 2004

PREVEND IT 2004

Lemos 2005

Verma 2005

Subtotal (95% CI)

Total events: 708 (statin), 883 (placebo)

Test for heterogeneity: 2=1.79, df=4, P=0.77, I2=0%

Test for overall effect: z=4.40, P


9/14


10/14

That the relative benefits of statins in patients withknown coronary heart disease are similar to those inpeople with mild renal impairment and people withwellpreserved renal function is plausiblethat is,mildrenal impairment should not prevent these patientsfrom receiving a statin. Controversy arises when

patients have established (stages 3-5) chronic kidneydisease and have not yet had a vascular event.In our systematic review, data on major clinical end

points were available from relatively few studies.Although we identified no heterogeneity in the resultsof the analyses of effect on outcome, the small numberof participating studies with substantial numbers ofthese end points meant that statistical power to detectsuch differences was suboptimal. That importantdifferences exist in the effects of statins among peoplewith different degrees of chronic kidney disease

remains possible.Secondly, the magnitude of the effecton the major clinical end points observed in the 4Dtrial, which included patients with the most severeforms of chronic kidney disease (that is, receivingdialysis), was somewhat less than that seen in trials thatincluded participants with better kidney function (all

cause mortality: relative risk 0.95, 0.85 to 1.07), so thatimportant differences in the balance of risks andbenefits may exist for this group.13 We should under-line that although the 4D study is generally interpretedas a negative trial, its results are broadly in line withexpectations from other trials, except for the findingson stroke. Thirdly, a substantial proportion of theparticipantsinthestudiesincludedinthismeta-analysishad established occlusive arterial disease and maytherefore benefit more than people without suchdisease (primary prevention), although a recent review


Fried 2001

HPS 2003

PPP 2004

PREVEND IT 2004

Lemos 2005

Subtotal (95% CI)


Test for heterogeneity: 2=5.78, df=4, P=0.22, I2=30.7%

Test for overall effect: z=4.58, P


11/14


12/14

reported creatinine data and proteinuria, meaning thatthe remaining studies either did not measure theseoutcomes or measured them and did not report them.None of the available trials reported end stage renaldisease or doubling of serum creatinine as study endpoints.

The beneficial effect of statins on cardiovascular endpoints and some evidence of renal benefit seen in oursystematic review may be potentially explained bycholesterol dependent effects, cholesterol independenteffects, or both. In addition to the well documentedassociation between cholesterol lowering and reduc-tion in cardiovascular risk in non-chronic kidneydisease populations,8-10 statins may modulate cardio-vascular risk by decreasing inflammation, enhancingendothelial function, inhibiting smooth muscle pro-liferation, exerting direct antithrombotic effects, andstabilising pre-existing atherosclerotic plaque.35-37 Sta-tins have also recently been shown to ameliorate

vascular calcification, which is an important problemin patients with chronic kidney disease. Similarmechanisms may underpin the beneficial actions ofstatins on progression of chronic kidney disease,although other cholesterol independent renoprotec-tiveactionssuchasinhibitionofrenalcellproliferation,antifibrotic effects, suppression of macrophage recruit-ment, antioxidation, and down regulation of inflam-matory cytokines also contribute.3839

In addition to considering the potentially beneficialactions of statins on clinical end points in chronickidney disease, our review provides some importantdata onthe potentialharms of statins in this population.

Concerns have been expressed about an increased riskof side effects of statins in patients with chronic kidneydisease.40 In this analysis of more than 6500 suchpatients, we did notfind statins to be associated with an

increased incidence of abnormalities in liver functiontests or raised serum creatinine phosphokinase con-centrations compared with placebo. Although rando-mised controlled trials and meta-analyses aregenerallyinadequately powered to precisely determine the risksof infrequent serious adverse effects associated with

drugs, the inclusion of a large number of patients withchronic kidney disease should provide some reassur-ance on the relative safety of statins in this group.

Strengths and weaknesses

A strength of this study is that it is the first to assesscardiovascular, renal, and toxicity outcomes, provid-ing a comprehensive systematic review of the benefitsand harms of statins on the basis of a pre-specifieddetailed published protocol, with rigid inclusioncriteria for randomised controlled trials only and acomprehensivesearch.Two independent investigatorsextracted data, analysed data, and assessed methodo-

logical quality. Furthermore, the possibility of publica-tion bias was minimised by inclusionof both publishedand unpublished trials. We sought information aboutunpublished and ongoing randomised controlled trialsfrom authors of the included randomised controlledtrials, drug companies, and experts in the field.

The main weakness of this study was the relativepaucity of high quality randomised controlled trials.The vast majority of studies evaluated failed to specifywhether randomisation allocation was concealed, out-come assessors were blinded, or data were analysed onan intentionto treat basis.Many studies were small andoften short in duration, and the results of our review

were dominated by the results of three major trials: the4Dtrial,PPP,andtheALERTstudy.13w10 w42 Moreover,we found evidence of trial heterogeneity in someanalyses, including total cholesterol, low density

Table 4 | Multivariatemeta-regressionanalysisof potential sourcesof heterogeneity oneffectof statinsversusplaceboon total

cholesteroland lowdensity lipoprotein cholesterolatendof treatment

Covariate beingadjusted for allothers

Total cholesterol Low density lipoprotein cholesterol

Adjusted weightedmean difference* (95%

CI) P valueExplained

variance (%)

Adjusted weightedmean difference * (95%

CI) P valueExplained

variance (%)

Age 0.07 75.26 0.05 37.33

Baseline cholesterol


13/14

lipoprotein cholesterol, triglycerides, creatinine clear-ance, and acute allograft rejection. As meta-analysisassumes that a mean value estimating the effects of anintervention can be determined by combining similartrials, how to handle summary data when significantheterogeneity exists is notclear. However, we have notbeen able to show any significant interaction in anyanalysis in which cumulative estimates were provided,and strong agreement exists between point estimatesfor individual outcomes in the separate subgroups ofpre-dialysis, dialysis, and transplant patients.

Future research and ongoing trialsOur analysis confirmed the role of statins in secondarypreventionin chronic kidney disease patients. Trials ofprimary prevention with statins are ongoing in peoplewith greater degrees of chronic kidney disease andlower levels of cardiovascular risk (study of heart andrenal protection (SHARP) and a study to evaluate theuse of rosuvastatin in subjects on regular hemodialysis:an assessment of survival and cardiovascular events(AURORA)).2627 The SHARP trial has randomisedmore than 9000 patients (6000 pre-dialysis, 2540haemodialysis, 490 peritoneal dialysis)to the combina-tion of simvastatin and the cholesterol absorption

inhibitor ezetimibe versus placebo and statin versusplacebo. The AURORA trial has randomised 2750

haemodialysis patients to rosuvastatin versus placebo.Their combined results, which will involve largenumbers of vascular and non-vascular deaths, willhelp to resolve the question of whether higherconcentrations of low density lipoprotein cholesterolare a cause of vascular disease in patients with chronickidney disease.

Conclusions

These results suggest that statin treatment is safe andreduces the risk of major cardiovascular events inpatients with chronic kidney disease (with cardio-vasculardisease)in a similar fashion tothatseenin trialsof statins in non-chronic kidney disease populations.These agents may have antiproteinuric effects,although the clinical significance of these benefitsremains uncertain. We did not show an effect on allcause mortality, but this may reflect inadequatestatistical power and the fact that, in general, popula-tions with stages 3-5 chronic kidney disease are largelyunderstudied.Ongoingrandomised controlled trials inchronic kidney disease patients (SHARP and AUR-ORA) should help to clarify the role of statins inprimary prevention and whether the postulated

differences in response to statin treatment in chronickidney disease are real.


Hommel 1992

Arnadottir 1993

Hernandez 1993

Thomas 1993

Rayner 1995

PERFECT study 1997

Lepre 1999

Fried 2001

Holdaas 2001

Santos 2001

Harris 2002

Saltissi 2002

DollaNora 2003

Holdaas 2003

PREVEND IT 2004

Yasuda 2004

HARP 2005

Lee 2005

Stegmayr 2005

Verma 2005

Total (95% CI)


Test for heterogeneity: 2=17.79, df=16, P=0.34, I2=10.0%

Test for overall effect: z=0.25, P=0.80 0.001 0.1 1 100.01 100 1000

Study or subcategory

2/12

1/20

1/11

4/15

5/42

1/54

3/32

0/17

30/182

1/34

13/82

6/38

0/12

155/1050

13/433

2/39

42/224

0/42

16/70

3/48

2457

3.85 (0.21 to 71.48)

0.50 (0.05 to 5.08)

1.00 (0.07 to 14.05)

1.33 (0.36 to 4.97)

2.68 (0.55 to 13.07)

2.95 (0.12 to 70.72)

3.82 (0.21 to 69.88)

Not estimable

0.93 (0.59 to 1.46)

0.49 (0.05 to 5.10)

1.37 (0.65 to 2.89)

0.89 (0.25 to 3.16)

Not estimable

0.90 (0.74 to 1.10)

0.59 (0.30 to 1.15)

2.10 (0.20 to 22.27)

1.05 (0.71 to 1.55)

Not estimable

4.17 (1.47 to 11.87)

6.29 (0.33 to 118.32)

1.03 (0.84 to 1.25)

0.46

0.73

0.57

2.22

1.55

0.39

0.47

14.80

0.71

6.41

2.40

38.57

7.72

0.71

18.38

3.44

0.46

100.00

Statinn/N

0/9

2/20

1/11

3/15

2/45

0/53

0/17

0/19

32/180

2/33

11/95

3/17

0/12

172/1052

22/431

1/41

40224

0/40

4/73

0/43

2430

Placebon/N

Favoursstatin

Favoursplacebo

Relative risk(random) (95% CI)

Relative risk(random) (95% CI)

Weight(%)

1992

1993

1993

1993

1995

1997

1999

2001

2001

2001

2002

2002

2003

2003

2004

2004

2005

2005

2005

2005

Year

Fig 6 | Withdrawal rates for statins compared with placebo in pre-dialysis, dialysis, and transplant patients. Only studies with at

least one withdrawal are included in the plot

R E S E A R C H

BMJ | ONLINE FIRST | bmj.com page 13 of 14


14/14

We acknowledge the contribution of authors (C Wanner, C Baigent,

HHoldaas,MTonnelli,BGStegmayr,RLLins,METhomas,SHDiepeveen,

A Zhang, M VanDijk, K P Harris, H Hernandez, D Saltissi, C Kosch, and T

Nakamura) who provided data about their trials on request. We thank C

Baigentand A Websterfor theirintellectualcontributionto the manuscript.

Wealso thank CatherineClaseandMartinLandrayforcriticalreviewof thismanuscript. We thank the Cochrane Renal Group staff and search

coordinators for their assistance with the search process. This paper has

been presented in part as an abstract at the Annual American Society of

Nephrology meeting at St Louis, USA, in November 2004 and

Philadelphia, USA, in November 2005.

Contributors: GFMS and SDN were equally involved in the concept anddesign of the review, data extraction, analysis and interpretation of data,

and writing the final manuscript. FP was involved in data analysis. AN was

involved in writing the final manuscript. DWJ was involved in analysis and

interpretation of data and writing the final manuscript . VP was involved in

critical revision for intellectual content and interpretation of data and

assistedwithwritingthe finalmanuscript. JCCwas involvedinconceptand

design, analysis and interpretation of data, and writing the final

manuscript. All authors were involved in final approval of the manuscript

to be submitted for publication. GFMS and SDN are the guarantors.

Funding: This review was supported in part by the Cochrane Renal Group.Competing interests: None declared.Ethical approval: Not needed.Provenance and peer review: Not commissioned; externally peerreviewed.

1 CoreshJ, SelvinE, Stevens LA, Manzi J, Kusek JW, EggersP, etal.Prevalence of chronic kidney disease in the United States.JAMA2007;298:2038-47.

2 FoleyRN, Collins AJ.End-stage renaldiseasein theUnited States: anupdatefrom the UnitedStates renal data system.J AmSoc Nephrol2007;18:2644-8.

3 GaneshS,StackA,LevinN.AssociationofelevatedserumPO(4),CaPO(4), andparathyroid hormonewith cardiacmortality riskin chronichemodialysis patients.J AmSoc Nephrol 2001;12:2131-8.

4 MallamaciF, Zoccali C, TripeiG, Fermo I, Benedetto FA, CataliottiA,et al. Hyperhomocysteinemiapredictscardiovascular outcomesin

hemodialysis patients.Kidney Int2002;61:609-14.5 Jungers P,Massy ZA,KhoaTN,Fumeron C,Labrunie M,LacourB, etal.

Incidenceand risk factorsof atherosclerotic cardiovascular accidentsin predialysis chronic renal failure patients: a prospective study.Nephrol Dial Transplant1997;12:2597-602.

6 Rossouw JE,LewisB, Rifkind BM.The value of lowering cholesterolafter myocardial infarction.N EnglJ Med1990;323:1112-9.

7 Law MR, WaldNJ,Thompson SG. Byhowmuch and how quickly doesreduction in serum cholesterol concentration lowerrisk of ischaemicheart disease?BMJ1994;308:367-72.

8 KongSX,Crawford SY, GandhiSK, SeegerJD, SchumockGT, Lam NP,et al. Efficacy of 3-hydroxy-3-methylglutarylcoenzyme A reductaseinhibitors in the treatment of patients with hypercholesterolemia: ameta-analysis of clinical trials.ClinTher1997;19:778-97.

9 HebertPR, GazianoJM, ChanKS, Hennekens CH.Cholesterolloweringwith statindrugs, risk of stroke, and totalmortality:an overview ofrandomized trials.JAMA 1997;278:313-21.

10 Law MR,Wald NJ,RudnickaAR. Quantifyingeffect of statinson low

densitylipoproteincholesterol, ischaemicheart disease,and stroke:systematic review and meta-analysis.BMJ2003;326:1423.

11 Kalantar-Zadeh K, Block G, HumphreysMH, KoppleJD. Reverseepidemiologyof cardiovascular riskfactors in maintenance dialysispatients. Kidney Int2003;63:793-808.

12 SeligerSL, Weiss NS,Gillen DL,Kestenbaum B, Ball A, Sherrard DJ,et al. HMG-CoA reductase inhibitors are associated with reducedmortality in end stage renal disease patients. Kidney Int2002;61:297-304.

13 WannerC, Krane V, MarzW, OlschewskiM, MannJF, Ruf G, etal.Atorvastatinin patients with type 2 diabetes mellitus undergoing

hemodialysis.N EnglJ Med2005;353:238-48.14 K/DOQI Group. Treatingdyslipidemias.AmJKidDis2003;41:S39-58.15 EBPGExpertGroup on Renal Transplantation.European bestpractice

guidelines for renal transplantation. Section IV: long-termmanagement of the transplant recipient. IV.5.3. Cardiovascular risks:hyperlipidaemia. Nephrol DialTransplant2002;17(suppl 4):26-8.

16 Renal Association. Clinical practice guidelines. Module 2:complications of CKD disease. Section 1: cardiovascular disease.2007. www.renal.org/guidelines/module2.html#Anchor-Summary-49575.

17 WalkerR. Lipid loweringin patientswithchronickidneydisease. (CARIguidelinesCaring for Australasians with Renal Impairment). 2005.www.cari.org.au/Lipid%20Lowering%20Therapy.pdf.

18 Sandhu S, Wiebe N, FriedLF, Tonelli M. Statins forimproving renaloutcomes: a meta-analysis.J AmSoc Nephrol 2006;17:2006-16.

19 DouglasK, OMalleyPG,JacksonJL.Meta-analysis:the effectof statinson albuminuria.Ann InternMed2006;145:117-24.

20 National Kidney Foundation. K/DOQI clinical practice guidelines for

chronic kidney disease: evaluation classification, and stratification.2000. www.kidney.org/professionals/kdoqi/guidelines_ckd/toc.htm.

21 DerSimonian R, Laird N. Meta-analysis in clinical trials.ContrClinTrials 1986;7:177-88.

22 Higgins JP, ThompsonSG, Deeks JJ,Altman DG.Measuringinconsistency in meta-analyses. BMJ2003;327:557-60.

23 Randomised trialof cholesterolloweringin 4444 patients withcoronaryheart disease: the Scandinavian simvastatinsurvival study(4S). Lancet1994;344:1383-9.

24 Liu Y, Coresh J, EustaceJA. Associationbetweencholesterollevel andmortality in dialysis patients: role of inflammation and malnutrition.JAMA 2004;291:451-9.

25 CostaJ, BorgesM, DavidC, CarneiroAV.Efficacyof lipidlowering drugtreatment for diabetic and non-diabetic patients: meta-analysis ofrandomised controlled trials.BMJ2006;332:1115-24.

26 UnitedStatesRenalDataSystem. Mortality andcauseof death. 2007.www.usrds.org/2007/ref/H_morte_07.pdf.

27 Baigent C, LandryM. Study of heart andrenal protection (SHARP).Kidney Int2003;84:S207-10.28 FellstromBC, Holdaas H,Jardine AG.Why do we need a statintrial in

hemodialysis patients?Kidney Int2003;84:S204-6.29 Lin J,Hu FB, RimmEB, Rifai N,CurhanGC. The association of serum

lipids and inflammatory biomarkers with renal function in men withtype II diabetes mellitus. Kidney Int2006;69:336-42.

30 Wheeler DC,Bernard DB.Lipidabnormalities in thenephroticsyndrome: causes, consequences, and treatment.AmJ KidneyDis1994;23:331-46.

31 Massy ZA,Ma JZ,LouisTA, Kasiske BL.Lipid-loweringtherapy inpatients with renal disease.Kidney Int1995;48:188-98.

32 Navaneethan SD, Shrivastava R. HMG CoA reductase inhibitors(statins) for dialysispatients. CochraneDatabaseSyst Rev2004;(4):CD004289

33 Stern RH, YangBB, HortonM, Moore S, AbelRB, Olson SC. Renaldysfunction doesnot alter the pharmacokinetics or LDL-cholesterolreduction of atorvastatin.J Clin Pharmacol1997;37:816-9.

34 Halstenson CE,TriscariJ, DeVault A, Shapiro B, Keane W, Pan H.Single-dose pharmacokinetics of pravastatin and metabolites inpatients with renal impairment.J Clin Pharmacol1992;32:124-32.

35 Sotiriou CG,ChengJW. Beneficial effects of statins in coronary arterydiseasebeyond lowering cholesterol.AnnPharmacother2000;34:1432-9.

36 RobinsonJG, SmithB, Maheshwari N,SchrottH. Pleiotropic effects ofstatins: benefit beyond cholesterol reduction? A meta-regressionanalysis. J AmColl Cardiol 2005;46:1855-62.

37 KinlayS, SelwynAP. Effects ofstatins oninflammation inpatientswithacuteand chronic coronarysyndromes.AmJ Cardiol2003;91:9-13B.

38 KeaneWF. Theroleof lipidsin renal disease:future challenges.KidneyInt2000;75:S27-31

39 Campese VM, Nadim MK, Epstein M. Are 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors renoprotective?J AmSoc Nephrol2005;16:S11-7.

40 SicaDA, GehrTW. 3-Hydroxy-3-methylglutaryl coenzymeA reductaseinhibitors and rhabdomyolysis: considerations in the renal failurepatient. CurrOpinNephrol Hypertens2002;11:123-33.

Accepted: 11 January 2008

WHAT IS ALREADY KNOWN ON THIS TOPIC

Patients with chronic kidney disease are at increased risk of cardiovascular disease

Statins reduce cardiovascular mortality and all cause mortality in the general population

The role of statins in chronic kidney disease is controversial

WHAT THIS STUDY ADDS

Statinsreducecardiovascular deaths in patients with chronickidney disease by a similar rateto that seen in the general population

Theefficacyof statinsin reducing allcausemortality inkidney diseasepatientsandtheirroleinprimary prevention need to be established in ongoing trials

Statins are safe as regards major side effects such as hepatotoxicity, rhabdomyolysis, andtreatment withdrawal

R E S E A R C H

estatina en ckd bmj 2009

Documents