AIM:Toassesstheeffectivenessofimmunosuppressantsintheprophylaxisofcornealallograftrejectionafterhigh-riskkeratoplastyandnormal-riskkeratoplasty.METHODS:WesearchedtheCochraneCentralRegisterofControlledTrials(CENTRAL),MEDLINE,EMBASE,CNKI,VIPandreferencelistsofarticles.Dateofmostrecentsearch:18June,2011.Allrandomisedcontrolledtrials(RCTs)assessingtheuseofimmunosupressantsinthepreventionofgraftrejection,irrespectiveofpublicationlanguage.Twoauthorsassessedtrialqualityandextracteddataindependently.Onlydichotomousoutcomes(cleargraftsurvival,ratioofimmunereactionsandsideeffects)wereavailableandwereexpressedasrelativerisk(RR)and95%confidenceintervals(CI).RESULTS:Sevenstudieswereincludedinthisreview.Inthecomparingofmycophenolatemofetil(MMF)withplacebo,theresultsshowedMMFcouldsignificantlyreduceimmunereactionscomparedwithplacebo(RR1.0895%Cl0.95to1.21),butnoeffectoncleargraftsurvival(RR1.1195%Cl0.90to1.35).Incleargraftsurvivalandimmunereactions,MMFandcyclosporineA(CsA)showedsimilareffect(RR1.1195%Cl0.90to1.35,andRR1.48,95%Cl0.56to3.93,respectively).Tacrolimus(FK506)andsteroidshowedsimilareffectsoncleargraftsurvivalandimmunereactions(RR0.32,95%CI0.02to6.21,andRR1.00,95%CI0.88to1.14,respectively).Nodrugrelativesideeffecthasbeenfound.CONCLUSION:MMFmayreduceimmunereactionsinbothnormal-riskandhigh-riskrejectionofpenetratingkeratoplasty.CsAandFK506showedsimilareffectsasMMF.However,duetothelackoflargeclinicaltrials,theevidenceremainweak,thequalityofevidenceswereratedasverylowtomoderate.Large,properlyrandomised,placebo-controlled,doublemaskedtrialsareneededtoevaluatetheeffectofimmunosuppressants.
