BloodPressureLoweringTreatmentTrialists’Collaboration
Abstract ObjectiveTodefinethecardiovasculareffectsofloweringbloodpressure inpeoplewithchronickidneydisease. DesignCollaborativeprospectivemeta-analysisofrandomisedtrials. DatasourcesandeligibilityParticipatingrandomisedtrialsofdrugs tolowerbloodpressurecomparedwithplacebooreachotherorthat comparedifferentbloodpressuretargets,withatleast1000patientyears offollow-upperarm. MainoutcomemeasuresMajorcardiovascularevents(stroke, myocardialinfarction,heartfailure,orcardiovasculardeath)incomposite andindividuallyandallcausedeath. Participants26trials(152290participants),including30295individuals withreducedestimatedglomerularfiltrationrate(eGFR),whichwas definedaseGFR<60mL/min/1.73m2. DataextractionIndividualparticipantdatawereavailablefor23trials, withsummarydatafromanotherthree.Meta-analysisaccordingto baselinekidneyfunctionwasperformed.Pooledhazardratiosper5mm Hglowerbloodpressurewereestimatedwitharandomeffectsmodel. ResultsComparedwithplacebo,bloodpressureloweringregimens reducedtheriskofmajorcardiovasculareventsbyaboutasixthper5 mmHgreductioninsystolicbloodpressureinindividualswith(hazard ratio0.83,95%confidenceinterval0.76to0.90)andwithoutreduced eGFR(0.83,0.79to0.88),withnoevidenceforanydifferenceineffect (P=1.00forhomogeneity).Theresultsweresimilarirrespectiveof whetherbloodpressurewasreducedbyregimensbasedonangiotensin convertingenzymeinhibitors,calciumantagonists,ordiuretics/βblockers. Therewasnoevidencethattheeffectsofdifferentdrugclassesonmajor cardiovasculareventsvariedbetweenpatientswithdifferenteGFR(all P>0.60forhomogeneity). ConclusionsBloodpressureloweringisaneffectivestrategyfor preventingcardiovasculareventsamongpeoplewithmoderatelyreduced eGFR.Thereislittleevidencefromtheseoverviewstosupportthe preferentialchoiceofparticulardrugclassesforthepreventionof cardiovasculareventsinchronickidneydisease.
Introduction Chronickidneydisease,mostcommonlydefinedbyareduced glomerularfiltrationrate(GFR)orabnormalconcentrationsof proteinuria,orboth,isanimportantpublichealthproblem, affecting10-15%oftheadultgeneralpopulation.1-3 Itis associatedwithanincreasedriskofkidneyfailureand cardiovasculardisease.4-6Individualswithearlychronickidney diseasearemorelikelytoexperienceacardiovascularevent thankidneyfailure,7andpreciseandreliableevidenceaboutthe effectsofstrategiestopreventcardiovasculardiseaseinthis largepopulationofpatientsisofgreatimportance. Bloodpressureisanimportantdeterminantoftheriskof cardiovasculardiseaseinthegeneralpopulation.8 Itiswell establishedthatinterventionsthatlowerbloodpressureprevent cardiovascularevents.9 10 Bloodpressureiscommonlyraised inindividualswithchronickidneydisease,5 6 andguidelines recommendlowerbloodpressuretargetsinthispopulationthan inpeoplewithoutchronickidneydisease.11 12 Severalstudies havealsosuggestedparticularbenefitsofdrugclassesacting throughtherenin-angiotensinsystemforthepreventionofrenal complications.13 Theevidencethatloweringbloodpressureis beneficialforpatientswithchronickidneydiseaseaswellas thosewithout,however,remainslimited,andthecomparative efficacyofdifferentregimenstolowerbloodpressureonthe riskofcardiovasculareventsinpatientswithandwithoutchronic kidneydiseaseremainsuncertain.14-19 TheBloodPressureLoweringTreatmentTrialists’ Collaboration20 wasestablishedtoperformaseriesof prospectivelydefinedoverviewsofrandomisedtrialsto investigatetheeffectsofbloodpressureloweringdrugson cardiovascularmorbidityandmortality,includingassessments ofthecomparativeeffectsofregimensbetweenmajorpatient subgroups.Thisanalysis,prespecifiedintheoriginal collaborationprotocol,20quantifiestheproportionalbenefitsof bloodpressurelowering,andthecomparativeeffectsofdifferent
Correspondence to:V Perkovic, George Institute for Global Health, University of Sydney, Level 13, 321 Kent St, Sydney NSW 2000, Australia VPerkovic@george.org.au
Extra material supplied by the author (see http://www.bmj.com/content/347/bmj.f5680?tab=related#webextra)
Appendix: Supplementary tables and figures [posted as supplied by author]
No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe
BMJ2013;347:f5680doi:10.1136/bmj.f5680(Published3October2013) Page1of15
Research
RESEARCH
classesofbloodpressureloweringdrugsinindividualswith andwithoutchronickidneydisease. Methods Datasourcesandstudyselection Trialswereeligibleforinclusioninthisprospectivecollaborative meta-analysisiftheymetoneofthefollowingcriteria:patients wererandomisedtoabloodpressureloweringdrug/regimenor acontrolgroup(placeboorlessintensivebloodpressure loweringregimen)orpatientswererandomisedbetween regimensbasedondifferentclassesofdrugstolowerblood pressure.Trialswerealsorequiredtohaveaminimumof1000 patientyearsofplannedfollow-upineachrandomisedarmand nottohavepresentedorpublishedtheirmainresultsbefore finalisationoftheoverviewprotocolinJuly1995.20 The collaborationwasjointlyestablishedbytheprincipal investigatorsin1995,andtheinclusioncriteriaforthe overviews20 specifiedthatresultsoftrialsreportedonlyafter thistimecouldbeincluded.Intheearlyyearsofthe collaboration,participantscontributedaggregatetrialdatabut, overtime,agreedtoprovideindividualpatientdata.Newtrials wereidentifiedbyarangeofmethodsincludingcomputeraided literaturesearches,scrutinyofabstractsandproceedingsof meeting,andbyinquiringamongcolleagues,collaborators,and themanufacturersofantihypertensivedrugs.Determinationof eligibilitywasbasedonareviewofdetailsofthestudydesign andqualityofthestudy,regardlessofmainresultsofeachtrial (appendixtable1).Theprincipalinvestigatorsforeligibletrials wereinvitedtojointhecollaborationastheywereidentified. Trialsforwhichdatabykidneyfunctionhadbeenobtainedby April2012wereincludedintheseanalyses.Forthisoverview, datawereavailablefrom26trialsprovidingeitherindividual participantdataincludingserumcreatinineconcentrations(23 trials21-44)orsummarydatastratifiedbyestimatedglomerular filtrationrate(tabulardatafromtwotrials15 45 andpublished hazardratiosfromonetrial14).Amongthem,25trials14 15 21-44 wereincludedinthemainanalyses,whileonetrial45contributed participantsonlytothereducedeGFRstratumandwasincluded insensitivityanalyses.Dataregardingurinaryproteinexcretion atbaselinewereavailablefor11trials.21 22 24 27-32 37 40 Two trials15 34 39didnotprovideinformationregardingthecombined endpointofmajorcardiovasculareventsaccordingtochronic kidneydiseasestatusbutprovideddataforotherendpoints.The individualparticipantdatarequestedincludedcharacteristicsof participantsrecordedatscreeningorrandomisation,selected measurementsduringfollow-up,anddetailsofalloutcomes duringthescheduledfollow-upperiod.WeusedtheCochrane Collaboration’stooltoassesstheriskofbias.46 Glomerularfiltrationratesandproteinuria Glomerularfiltrationrateswereestimatedwiththemodification ofdietinrenaldiseaseformula7: eGFR=186.3×(serumcreatinine/88.4)−1.154×(age)−0.203×1.210(if black)×0.742(iffemale) whereeGFRisinmL/min/1.73m2,ageisinyears,andserum creatinineisinµmol/L.Theparticipantsweredividedintotwo categoriesofbaselineeGFR(≥60mL/min/1.73m2 orand<60 mL/min/1.73m2)withestablishedcutpointsrecommendedin renalguidelines.11Publisheddatafromtwotrialsinwhichkidney functionwasestimatedwiththeCockcroft-Gaultformula14 or 125-iothalamateclearance45 andpatientswerecategorisedas havingeGFR<65mL/minand≥65mL/minweredeemed sufficientlycomparableandincludedinthemainanalysis.The presenceofproteinuriawasdefinedasanyofthefollowing:
urinaryalbuminexcretionrates>200μg/minor>300mg/day, urinaryalbuminconcentration>200mg/L,urinaryalbumin creatinineratio>300μg/mg,oraurinaryproteindipsticktest resultof1+orgreater.7 Stagesofchronickidneydiseasewere definedaccordingtotheguidelinesfromtheKidneyDisease OutcomesQualityInitiative(K/DOQI).11 Outcomes Thesixoutcomesweredefinedaccordingtotheinternational classificationofdiseasesandwereprespecifiedinthe collaboration’sprotocol.20 Outcomeswerestroke(non-fatal strokeordeathfromcerebrovasculardisease),coronaryheart disease(non-fatalmyocardialinfarctionordeathfromcoronary heartdiseaseincludingsuddendeath),heartfailure(causing deathorrequiringadmissiontohospital),cardiovasculardeath, andtotalmortality.Themainoutcomewasmajorcardiovascular eventscomprisingstroke,coronaryheartdisease,heartfailure, andcardiovasculardeath.Onlythefirsteventoftherelevant outcometypewasincludedineachanalysis. Treatmentcomparisons Thetreatmentcomparisonstestedwereprespecifiedinthe originalprotocol.20 Inthebroadgroupoftrialscomparingan activetreatmentandacontrol,wecarriedoutseparateoverviews forangiotensinconvertingenzyme(ACE)inhibitorbased regimensversusplacebo;calciumantagonistbasedregimens versusplacebo;andmoreintensiveversuslessintensiveblood pressureloweringregimens.Inthebroadgroupoftrials comparingdifferentactiveagents,wecarriedoutseparate overviewsforACEinhibitorbasedregimensversusconventional treatment(diureticsorβblockerbasedregimens);calcium antagonistbasedregimensversusconventionaltreatment;and ACEinhibitorbasedregimensversuscalciumantagonistbased regimens. Datasynthesisandanalysis Bloodpressurereductions Thereductioninbloodpressureineachtrialarmwascalculated asthemeanofthedifferencebetweeneachparticipant’smean bloodpressureduringfollow-upandtheirbloodpressureat baseline.Themeandifferenceinbloodpressurebetween randomisedgroupswasthencalculatedbysubtractingthevalues forthetwoarmsaccordingtochronickidneydiseasestatus. Meta-analysesofsubgroupsaccordingtokidney function. Toinvestigatetheeffectsofactivetreatmentcomparedwith placebo,moreintensivecomparedwithlessintensiveregimens, ordifferentdrugclassesontheoutcomesinpatientswith differentbaselineeGFR,weperformedmeta-analysesaccording tokidneyfunction.Allthemeta-analysesusedatwostage approachwherebytheriskestimateswerefirstsummarisedin eachtrialandthencombined. Foreachtrialandeachoutcome,weestimatedtheriskestimates separatelyforeachsubgroup,accordingtotheintentiontotreat principle.WecomputedhazardratiosoroddsratiosusingaCox proportionalhazar
การแปล กรุณารอสักครู่..
