AIM:Tocharacterizetemporalpatternofresolutionandrecurrenceofnaivechoroidalneovascularization(CNV)secondarytowetage-relatedmaculardegeneration(AMD)treatedwithintravitrealbevacizumabonasneededregimen,andtoanalyzebaselineriskfactorsforCNVresolutionorrecurrence.METHODS:Ninety-oneeyesof80patientswithnewlydiagnosedwetAMDwereretrospectivelystudied.Alleyesweretreatedwitharoundofthreemonthlyintravitrealbevacizumabinjections,followedbyoneadditional’bonus’injectionafterresolutionofCNVactivity.Duringfollow-up,eyesweremonitoredwithfluoresceinangiography,opticalcoherencetomography,andbest-correctedvisualacuity(BCVA).IncaseofrecurrencesofCNVactivity,eyeswereretreatedwithotherroundsofbevacizumabinjectionsfollowingthesametreatmentprotocol.RESULTS:Overamedianfollow-upof532d,themedianresolutiontimeofCNVactivityinthefirst,second,andthirdtreatmentroundwas98d,126d,and111d,respectively.Themedianrecurrencetimeforthethreeroundswas154d,126d,and151d,respectively.Nosignificantdifferenceinresolutiontime(P=0.09)orinrecurrencetime(P=0.11)wasdetectedamongtreatmentrounds.Age(P=0.0082)andlensstatus(P=0.035)werefoundtobeassociatedwithCNVresolution;forevery1-yearincreaseinagetherewas4%greaterchanceofCNVresolution;Phakiceyesdemonstrateda33%betterchancetoexperienceCNVresolutionthanpseudophakiceyes.ForCNVrecurrence,lensstatus(P=0.0009)andgender(P=0.0446)werefoundtobepredictive;pseudophakiceyeshada3.69-foldgreaterrisktoexperiencerecurrenceofCNVactivitycomparedtophakiceyes;maleshada2.19-foldgreaterrisktoexperiencerecurrenceofCNVactivitythanfemales.NosignificantBCVAchangesamongthreetreatmentroundswerenoted(P=0.56).CONCLUSION:ResolutiontimeandrecurrencetimeofCNVactivitywerenotsignificantlydifferentamongtreatmentrounds,suggestingabsenceoftachyphylaxistobevacizumab.Acautiousdecisionshouldbemadeupondisco
