TractionalAbnormalitiesoftheCentralFovealBouquetinEpiretinalMembranes - PDF document

TractionalAbnormalitiesoftheCentralFovealBouquetinEpiretinalMembranes:ClinicalSpectrumandPathophysiologicalPerspectivesANDREAGOVETTO,KAVITAV.BHAVSAR,GIANNIVIRGILI,MATTHEWJ.GERBER,K.BAILEYFREUND,CHRISTINEA.CURCIO,CLAUDEF.BURGOYNE,JEAN-PIERREHUBSCHMAN,ANDDAVIDSARRAFToinvestigatethetractionalalterationsofthecentralbouquet(CB)inidiopathicepiretinalmem-branes(ERMs).Retrospective,consecutive,observationalcaseseries.ERMswereclassiÞedaccordingtoa4-stagegradingsystem.TheCBwasdeÞnedasacircularareaofapproximately100mcomposedofdenselypackedcones(andMullercells)inthecentralfovea.Tractionalabnor-malitiesoftheCBwereidentiÞedwithspectral-domainopticalcoherencetomography.Exvivohistopathologicanalysiswasperformed. CONCLUSIONS:Thecottonballsign,foveolardetach-ment,andacquiredvitelliformlesionmaycompriseacon- SupplementalMaterialavailableatAcceptedforpublicationOct18,2017.FromtheRetinaDivision(A.G.,J.-P.H.)andRetinalDisordersandOphthalmicGeneticsDivision(D.S.),SteinEyeInstitute,UniversityofCaliforniaLosAngeles,LosAngeles,California;CaseyEyeInstitute,OregonHealthandScienceUniversity,Portland,Oregon(K.V.B.);PortlandVAHealthcareSystem,Portland,Oregon(K.V.B.);CareggiUniversityHospital,UniversityofFlorence,Florence,Italy(G.V.);DepartmentofMechanicalandAerospaceEngineering,UniversityofCaliforniaLosAngeles,LosAngeles,California(M.J.G.);VitreousRetinaMaculaConsultantsofNewYork,NewYork,NewYork(K.B.F.);DepartmentofOphthalmology,UniversityofAlabamaSchoolofMedicine,Birmingham,Alabama(C.A.C.);OregonHealthandScienceUniversity,Portland,Oregon(C.F.B.);andGreaterLosAngelesVAHealthcareCenter,LosAngeles,California(D.S.).InquiriestoDavidSarraf,RetinalDisordersandOphthalmicGeneticsDivision,SteinEyeInstitute,UniversityofCaliforniaLosAngeles,100SteinPlaza,LosAngeles,CA90095-7002;e-mail: 0002-9394/$36.00https://doi.org/10.1016/j.ajo.2017.10.011UBLISHEDBYLSEVIER ndingasthecottonballsignMoreover,PisonandassociatesreportedthepresenceoffoveolardetachmentinasubsetofeyeswithidiopathicERM,whileotherauthorshavenotedtheformationofacentralacquiredvitelliformlesioninpa-tientswithvitreomacularinterfacedisorder.13–15WeproposethatsuchlesionsmaycompriseacontinuousclinicalspectrumandmayrepresentsubsequentstagesinthedevelopmentoftractionalCBabnormalities.ThemorphologyandlocationoftheseabnormalitiessuggestanintegralroleoffovealMu¨llercellsinthisprocess.Me-chanicalstressmaybeexertedonthestructuralelementsofthefoveaowingtotheintimateinterconnectednetworkoffovealMu¨llercellsthatinterleavewiththeCB.ThepurposeofthisstudyistoevaluatethishypothesisbydescribingandcharacterizingtheSD-OCTndingsofcen-tralfovealhyperreectivity(ie,thecottonballsign),foveo-lardetachment,andacquiredvitelliformlesioninalargeconsecutivecohortofpatientswithidiopathicepiretinalmembraneformation.STUDYDESIGN:ThiswasaninstitutionalstudycarriedoutattheSteinEyeInstituteinadherencetothetenetsoftheDeclarationofHelsinkiandwiththeapprovalof theinstitutionalreviewboardoftheUniversityofCalifor-nia,LosAngeles.Aretrospective,observational,andconsecutivereviewofpaperandelectronicrecordsofpatientsdiag

nosedwithidio-pathicERMsandevaluatedby2retinaspecialists(J.P.H.andD.S.)betweenJanuary1,2010andJanuary31,2017attheSteinEyeInstitutewasperformed.Caseswereidentiedbyamedicalbillingrecordsearch,usingtheInternationalSta-tisticalClassicationofDiseasesandRelatedHealthProb-lems,NinthRevision(ICD-9)diagnosiscode362.56formacularpucker.Inclusioncriteriaincludedthepresenceofunilateralorbilat-eralidiopathicERM;exclusioncriteriaarelistedinTable1CLINICALASSESSMENTANDSPECTRAL-DOMAINOPTI-CALCOHERENCETOMOGRAPHYIMAGING:Allpatientsunderwentacompleteophthalmologicassessment,includingslit-lampbiomicroscopyanddilatedfundusex-amination.Snellenvisualacuitywasobtainedandconvertedintothelogarithmoftheminimumangleofres-olution(logMAR)forstatisticalanalysis.AlleyeswereimagedwiththeHeidelbergSpectralisSD-OCTdevice(HeidelbergEngineering,Heidelberg,Germany)withactiveeye-trackingtechnology,andwereanalyzedwiththeHeidelbergEyeExplorer(version1.8.6.0)usingtheHRA/SpectralisViewingModule(version5.8.3.0).SD-OCTimagesweregeneratedusingthemacularrasterwith19B-scansadministeredoveranareaof20degreeswitheachB-scanspaced242mapartandasinglehigh-denitionhorizontalB-scanat30degrees.Inaddi-tion,someoftheincludedeyesunderwenthigh-densityscanningoveramacularareaofeither1510degreeswith97B-scanseachspaced31mapartor155degreeswith131B-scansspaced13mapart.Theretinalthicknessmapanalysisprotocolwasselectedtoanalyzenumericaveragesofthemeasurementsforthecentralfovealthickness(CFT).Furthermore,themorphologyofthefoveawasqualitativelyevaluatedineachpatient.TractionalCBabnormalitiesweredividedinto3cate-gories:presenceofacottonballsign,presenceofafoveolardetachment,orpresenceofanacquiredvitelliformlesion.ThecottonballsignwasdenedaccordingtoTsunodaandassociatesastheSD-OCTappearanceofathickenedorroundishfuzzyhyperreectiveareabetweentheinnersegmentellipsoidzone(EZ)andtheinterdigitationzone(IZ)inthecentralfovea(Figure1,Top).FoveolardetachmentwasdenedwithSD-OCTasthepresenceofacentralpocketofsubneurosensoryhyporeec-tivityorsubretinaluid(Figure1,Middle).AcquiredvitelliformlesionwasdenedwithSD-OCTasdome-shapedsubretinalhyperreectivemateriallocatedexternaltotheEZandinternaltopreservedretinalpigmentepithelium(RPE),asillustratedinFigure1(Bottom).TheevaluationofallCBabnormalitieswasperformedby2independentgraders(A.G.andD.S.),andpossibledis-agreementswereresolvedwithopenadjudication. TABLE1.ExclusionCriteriaforStudyofTractionalAbnormalitiesoftheCentralBouquetExclusionCriteriaAnypreviousintraocularsurgerywiththeexclusionofuncomplicatedphacoemulsiÞcationHistoryofretinaldetachmentIntermediateoradvancedage-relatedmacularHistoryofchoroidalneovascularizationofanyetiologyCentralserouschorioretinopathyProliferativediabeticretinopathyNonproliferativediabeticretinopathywithhistoryofclinicallysigniÞcantdiabeticmacularedemaMaculartelangiectasiasTractionalanddegenerativelamellarmacularholesHistoryofcentralorbranchretinalveinocclusionandcentralorbranchretinalarteryocclusionAdvancedglaucoma,oropticneuropathyofanykindHistoryofinßammatoryeyedisordersHistoryofIrvine-Gasssyndrom

eVisuallysigniÞcantcataractHistoryofendophthalmitisoranyotherintraocularRetinaldystrophiesFovealhypoplasia/foveaplanaHistoryofoculartraumaAnyotherpotentialcauseofvisionlossotherthanepiretinalmembranesECEMBER2017MERICANOURNALOFPHTHALMOLOGY MaximalwidthandheightoftheseCBabnormalitiesweremeasuredmanuallywiththe‘‘caliper’’functionoftheHeidelbergsoftware,asillustratedinFigure2.ThewidthwasmeasuredbytracingahorizontallineparalleltotheRPE,attheleveloftheIZband,whiletheheight wasmeasuredbytracingaverticallineperpendiculartotheRPE,fromtheIZbandtotheinnerlimitoftheEZband.ERMsweredenedwithSD-OCTasastraightorirreg-ularhyperreectivelineabovetheILM,oftenassociatedwithunderlyingretinalsurfacewrinklingandwiththe FIGURE1.Tractionalabnormalitiesofthecentralbouquet.(Top)Cottonballsign.Astage2epiretinalmembraneisassociatedwithasmall,fuzzyhyperreßectivearea(ÔÔcottonballsign,ÕÕwhitearrows)observedbetweentheellipsoidzone,whichappearsthickened,andtheinterdigitationzone,whichappearslessdistinct.Boththeellipsoidzoneandtheexternallimitingmembranearepreserved.(Middle)Foveolardetachment.Astage2epiretinalmembraneisassociatedwithacentralhyporeßectivepocketofsubretinalßuidundertheinterdigitationzone(whitearrows).Theellipsoidzoneandtheexternallimitingmembraneareintact,butlesswell-deÞned(vstheirmorphologyassociatedwiththecottonballsign).(Bottom)Acquiredvitelliformlesion.Astage2epiretinalmem-braneisassociatedwithathickdome-shapedhyperreßectiveacquiredvitelliformlesionbetweentheretinalpigmentepitheliumandtheellipsoidzone(whitearrows).Theellipsoidzonemaybedisrupted,buttheexternallimitingmembraneispreserved. FIGURE2.Quantitativemeasurementsofthetractionalabnormalitiesofthecentralbouquetusingspectral-domainopticalcoher-encetomography.(Left)Cottonballsign.Themaximalwidthofthehyperreßectivelesionismeasuredbytracingahorizontallineparalleltotheretinalpigmentepitheliumattheleveloftheinterdigitationzone.Themaximalheightismeasuredbytracingaverticallineperpendiculartotheretinalpigmentepitheliumfromtheanteriorsurfaceoftheretinalpigmentepitheliumtotheposteriorborderoftheexternallimitingmembrane.(Middle)Foveolardetachment.Themaximalwidthandheightaremeasuredasdescribedforleftimage.(Right)Acquiredvitelliformlesion.Themaximalwidthofthehyperreßectivelesionismeasuredasdescribedfortheleftim-age.Themaximalheightismeasuredbytracingaverticallineperpendiculartotheretinalpigmentepitheliumfromtheinterdigitationzoneband.Incontrasttotheleftandmiddleimages,theupperlimitoftheheightofthelesionmaybeeithertheellipsoidzoneortheexternallimitingmembrane,dependingontheseverityofthelesion.RACTIONALBNORMALITIESOFTHE presenceofhyporeectivespacesbetweentheERMandILM.AllERMswereclassiedaccordingtothe4-gradeSD-OCTstagingsystemrecentlypublishedbyGovettoandassociates.Briey,stage1ERMsweremild,withpre-servedfovealdepression.Stage2ERMswereassociatedwithwideningoftheouternuclearlayerandlossofthefovealdepression.Stage3ERMswereassociatedwiththepresenceofcontinuousectopicinnerfoveallayers(EIFL)acrosstheentirefovea.Instages1,2,and3allretinallayerswereclearlydenedonSD-OCT.Stage

4ERMswere associatedwithcontinuousectopicinnerfoveallayersandthickeneddisorganizedretinallayers.HISTOPATHOLOGICASSESSMENT:Histopathologicanal-ysiswasperformedin1normaladultmacaquemonkeyeyeandin2normaladulthumaneyesdonatedforresearchtotheAlabamaEyeBank.Themonkeyeyewaspreservedbyintracardiacperfusion,postxedinosmiumtannicacidpara-phenylenediamine,embeddedinepoxyresin,sectionedat0.8m,andstainedwithtoluidineblue.Onehumandonor TABLE2.BaselineCharacteristicsoftheIncludedEyesERMStageBCVAEllipsoidDisruptionHorizontalDiameterVerticalDiameterCottonballsign(n34)Stage1:nStage2:nStage3:nStage4:n0.11logMAR20/27Snellenequivalent3/34229m59Foveolardetachment(n10)Stage1:nStage2:nStage3:nStage4:n0.19logMAR20/36Snellenequivalent7/10383.1m94.8Vitelliformlesion(n12)Stage1:nStage2:nStage3:nStage4:n0.20logMAR20/47Snellenequivalent9/12411.3m90.1value.007best-correctedvisualacuity;ERMepiretinalmembrane;logMARlogarithmoftheminimumangleofresolution.Kruskal-Wallistest. FIGURE3.Anatomicandfunctionalevaluationoftractionalabnormalitiesofthecentralbouquet.(Left)Correlationbetweenhor-izontalandverticaldiameters.Maximalheightandwidthofthetractionalabnormalitiesofthecentralbouquetarestronglycorrelatedcorrelated0.71).ThesubtypesofcentralbouquetabnormalitiesaresigniÞcantlydifferentinsize(.001).ThecottonballsubtypeissigniÞcantlysmaller,whereasthefoveolardetachmentandacquiredvitelliformlesionsubtypesaresigniÞcantlylarger.(Right)Dif-ferencesinbest-correctedvisualacuity(BCVA)betweendifferenttractionalcentralbouquetabnormalitystages.Functionaldiffer-encesbetweenthecottonball,foveolardetachment,andacquiredvitelliformlesionsubtypesarestatisticallysigniÞcant(EyeswithcottonballsignareassociatedwiththestrongestBCVA,whereaseyeswithacquiredvitelliformlesionhavethepoorest.ECEMBER2017MERICANOURNALOFPHTHALMOLOGY eye(86yearsold;death-to-preservationinterval:4.15hours)waspreservedbyimmersioninbuffered1%paraformaldehydeand2.5%glutaraldehydeafteranteriorsegmentremoval,thenpostxedinosmiumtannicacidparaphenylenediamineandprocessedasdescribedwiththemonkeyretina.Theotherhu-mandonoreye(87yearsold;death-to-preservationinterval:1.88hours)waspostxedin1%osmium,andwasotherwisehandledlikethemonkeyretina.STATISTICALANALYSIS:Descriptivestatisticswerecalculatedforallvariablesofinterest.Meanandstandarddeviationvalueswerecalculatedforcontinuousvariables,whilefrequencyandpercentagewerecalculatedforcate-goricalvariables.Kruskal-Wallistestwasusedtocomparethestatisticallysignicantdifferenceincontinuousmea-surementsamongallsubgroups;testwasusedtocompareproportionsamongthestudypopulation.TheassociationoftractionalCBabnormalitiesandotherSD-OCTparam-eterswithBCVAwasassessedbymeansofunivariateandmultivariatelinearorlogisticregression,asappropriate.AllanalyseswereconductedusingStata14.2software(StataCorp,CollegeStation,Texas,USA). RESULTSAFTERACOMPREHENSIVECHARTREVIEW,263EYESFROM255patients(131,51.4%male;and124,48.6%female)mettheinclusioncriteriaandwereenrolledinthestudy.BilateralERMswerediagnosedin8outof255patients(3.1%).Inthestudypopulation,143outof263eyes(54.4%)hadame

anfollowupof21.216.7months(range3–84months).Longitudinaldatawerenotavailablefortheremaining120(45.6%)eyes.Atbaseline,stage1ERMswerediagnosedin39outof263eyes(14.8%),stage2ERMswerediagnosedin121outof264eyes(46.0%),stage3ERMsin83outof264eyes(31.5%),andstage4ERMsin20outof264eyesAtbaseline,tractionalabnormalitiesoftheCBwerediagnosedin58outof263eyes(22.0%).Themostfrequentsubtypewasthecottonballsign(36outof58eyesor62.1%),followedbytheacquiredvitelliformlesion(12outof58eyesor20.7%),andfoveolardetachment(10outof58eyesor17.2%).BaselinecharacteristicsofeyeswiththeseCBabnormalitiesaresummarizedinTable2 FIGURE4.Sequentialevolutionoftractionalabnormalitiesofthecentralbouquetinthesameeye.(Top)Stage2epiretinalmem-braneisnoted,buttherearenosignsoftractionalabnormalitiesofthecentralbouquet.Theexternallimitingmembrane,ellipsoidzone,andinterdigitationzonearepreserved.Visualacuityis20/20.(Middle-top)Twoyearslater,asmallfuzzyhyperreßectivearea,consistentwithacottonballsign,isnotedbetweentheinterdigitationzoneandtheellipsoidzone.Theexternallimitingmem-brane,ellipsoidzone,andinterdigitationzonebandsarepreserved,althoughtheyareslightlydisplacedupwardcomparedtobaseline.Visualacuityisstable.(Middle-bottom)Threeyearsafterthedevelopmentofthecottonballsign,acentralacquiredvitelliformlesionovertheretinalpigmentepitheliumisnoted,andthehyperreßectivelesionislarger.Theellipsoidzoneandexternallimitingmem-braneareintact,butvisualacuityhasdeclinedto20/40.(Bottom)Oneyearlater,theacquiredvitelliformlesionhasincreasedinsize,buttheellipsoidzoneandexternallimitingmembraneremainintact.Visualacuityisstableat20/40.RACTIONALBNORMALITIESOFTHE ThemeanheightandwidthoftheseCBabnormalitiesatbaselinewere66m(range41–114m),and282.6111m(range149–519m),respectively,andbothvalueswerestronglycorrelated(r0.71),asillustratedinFigure3(Left).Differencesinbothheightandwidthbetweenthe3differentsubtypeswerestatisticallysignicant(.001).BASELINEASSOCIATIONSOFTRACTIONALCENTRALBOUQUETABNORMALITIESWITHBEST-CORRECTEDVI-SUALACUITYANDOTHERSPECTRAL-DOMAINOPTICALCOHERENCETOMOGRAPHYPARAMETERS:Inthestudypopulation,meanbaselineBCVAwas0.220.2logMAR(20/33Snellenequivalent),rangingbetween0and1.3logMAR(20/20to20/400Snellenequivalent).VisualacuitydifferencesbetweenERMstageswerestatisticallysignicant;thatis,BCVAwasbetterinearlierstage1and2ERMsandpoorerinadvancedstage3and4ERMs(.001).BCVAwasalsoassociatedwiththevariousformsofCBabnormalities;thatis,visualacuitywasbetterineyeswithacottonballsignandworseineyeswithanacquiredvitelli-formlesion(.001),asillustratedinFigure3Tosupportthehypothesisthatcottonballsign,fovealdetachment,andacquiredvitelliformlesionmayrepresentsequentialstagesofCBabnormalities,weappliedthesecat-egoriesasalineartrendtoinvestigatetheirassociationwithBCVA.AdjustingbyERMtype,therewasasignicant associationwithBCVA(.001)andthetrendforworsevisioncorrelatedsignicantlywiththesequentialstagesofCBabnormalities(Atbaseline,asignicantassociationwasidentiedwiththepresenceofanyformofCBabnormalityandERMstage.007),withhighestfrequencynotedinstage2ERM(32.2%)andlowestfrequ

encynotedinstage4ERM(5.0%).Inalogisticregressionmodel,theoddsratio(OR)forthepresenceofCBabnormalitieswas3.2forstage2vsstage1ERMs(.023),whereasalowerORandnosignicantdifferencewasfoundforstage3or4ERMsvsstage1ERMs.Inaddition,thepresenceofEIFL,whichcharacterizesstage3and4ERMs,wassignicantlyassociatedwithalowerincidenceofCBabnormalities(OR:0.33,Therewasalsoasignicantassociationbetweenthesub-typeofCBabnormalityandtheERMstage(.007).Spe-cically,instage2ERMsthemostcommonCBabnormalitysubtypewasthecottonballsign(25outof121eyes,20.7%),whereasinstage3ERMsitwastheacquiredvitelliformlesion(6outof83eyes,7.2%).However,therewasnosignicantassociationofCBabnormalitieswithCFT.ThepresenceofEZdisruptionincreasedsignicantlywithmoreadvancedCBabnormalities(.001)andwaslessfrequentineyeswiththecottonballsign(3outof34eyes,8.8%)andmorefrequentineyeswithfoveolar FIGURE5.Anatomicprogressionoftractionalabnormalitiesofthecentralbouquet.(Top)Case1.Stage2epiretinalmembrane.Afuzzyhyperreßectiveareabetweentheinterdigitationzoneandtheellipsoidzoneisnoted,andcorrespondstothecottonballsign(whitearrow).Visionis20/20.Oneyearlater,acentralhyporeßectivepocketofsubretinalßuidisnotedundertheinterdigitationzone(blackarrow),consistentwithafoveolardetachment.Visionisstable.(Bottom)Case2.Stage2epiretinalmembrane.Afoveolardetachmentispresent(whitearrow).Visionis20/20.Twoyearslater,acentralacquiredvitelliformlesionisnotedovertheretinalpigmentepithelium(blackarrow).Visualacuitydeclinedto20/30.ECEMBER2017MERICANOURNALOFPHTHALMOLOGY detachment(7outof10eyes,70.0%)andacquiredvitelli-formlesion(9outof12eyes,75.0%).FUNCTIONALANDANATOMICPROGRESSIONOFTRAC-TIONALCENTRALBOUQUETABNORMALITIESINEYESWITHLONGITUDINALFOLLOW-UP:Longitudinalfollow-upwasavailablefor143eyes,ofwhich41(28.7%)werediagnosedwithaCBabnormalityatbaseline.Althoughtheheightdidnotchangesignicantlyoverthefollow-upperiod(.360),asignicantincreaseinthewidthoftheCBabnormalitywasrecorded(IneyeswithoutanyCBabnormalityatbaseline,newCBabnormalitiesdevelopedin6outof102eyes(5.9%)overthefollow-upperiod.Inthissubgroup,4eyesdevelopedacottonballsign,1eyedevelopedanacquiredvitelliformlesion,andin1eyeitwaspossibletoidentifythesequentialevolutionfromcottonballsigntovitelliformlesionFigure4IneyeswithanyCBabnormalityatbaseline,anatomicprogressionwasnotedin6outof41eyes(14.6%)duringthefollow-upperiod.Inthissubset,3eyesprogressedfromcottonballsigntofoveolardetachment(Figure5Top),while2eyesevolvedfromfoveolardetachmenttoac-quiredvitelliformlesion(Figure5,Bottom).In1case,thesequentialevolutionfromcottonballsigntoacquiredvitel-liformlesionwasdescribed.SpontaneousresolutionofaCBabnormalitywasdescribedin2outof41eyes(4.9%).Intherstcase,acot-tonballsignregressedowingtothespontaneousreleaseof ERMtractionovertheCB,whichwasassociatedwithimprovementinBCVA(Figure6).Inthesecondcase,anacquiredvitelliformlesionspontaneouslyresolvedwiththemigrationofhyperreectivefociintotheouterretina,asillustratedinFigure7Attheendofthefollow-upperiod,aCBabnormalitywaspresentin45outof143eyes(31.5%).Sevenoutofthese45eyes(15.5%)illustrated

signsofanatomicprogres-siontolaterstagesandtheBCVAdeclinedin5outofthese7eyes(71.4%).HISTOPATHOLOGY:Histopathologicsectionsoftheretinafromamonkeyeye(Top)and2humaneyes(MiddleandBottom)areillustratedinFigure8.Eacheyedisplaysdistinctivestainingwithtoluidineblue,whichiscommonlyusedforcomprehensivetissuestainingforlightmicroscopicexaminationofepoxy-embeddedmaterialpre-paredforelectronmicroscopy.ThemonkeyretinaFigure8,Top)and1humanretina(Figure8,Bottom)illustratecysticchangeinthecentralfovealoorconsistentwithswellingofMu¨llercellprocesses.Itshouldbenotedthatthecontrastisdifferentinthetopandmiddlepanels(Mu¨llercellspresumedtobedarkstain-ing)vsthebottompanels(Mu¨llercellspresumedtobepalestaining).ToluidineblueisnotspecicforMu¨llercellsinthewayimmunohistochemistryforglutaminesynthase,vimentin,orglialbrillaryacidicproteincanbespecicforthesecells,andthebasisofthedifferentialtoluidinebluestainingisunknown.Nevertheless,thespecimens FIGURE6.Spontaneousresolutionofthecottonballsign.(Top)Stage2epiretinalmembranewithtractionoverthecentralbouquetisnoted,causingthedevelopmentofacottonballsign(whitearrows).Visualacuityis20/30.(Middle)Twoyearslater,thefreeedgeoftheepiretinalmembranespontaneouslydetachesfromthefovealcenter,releasingthetractionoverthecentralbouquet.Thisleadstoanimprovementofthecottonballsign,whichisstillvisible(whitearrows).Visualacuityis20/40.(Bottom)Oneyearlater,thefreeedgeoftheepiretinalmembraneisstilldetached,butthethicknessoftheretinaisremarkablyreducedandthecottonballsignisresolved.Visualacuityisimprovedto20/25.RACTIONALBNORMALITIESOFTHE analyzedshowpeculiarstainingpatternsintheCBconsis-tentwithmoleculardifferencesbetweentheMu¨llercellsthataccompanythecentralconesandthoseelsewhereintheretina.ThesestainingpatternsofthefovealMu¨llercellsremarkablysupportthepotentialmechanismof¨llercelltractionuponthecentralbouquet.SIGNIFICANTIMPROVEMENTSINMACULARIMAGINGTECH-nologyhaveenabledinvivovisualizationofretinalmicro-structures(eg,outerretinaltubulations)thatwerepreviouslydescribedonlyhistologically.WithSD-OCT,TsunodaandassociatesdescribedthepresenceofafuzzyhyperreectiveareabetweentheEZandtheIZinthecentralfoveaineyeswithvitreomaculartractionorERMformation,andattributedthissubtlendingtoin-wardtractionoverthefovealregion.Inlaterpublications,otherauthorsdescribedmoreseveremodicationsoftheouterretinalmorphologyinvitreomacularinterfacedisor-ders,includingthedevelopmentoffoveolardetachmentandcentralacquiredvitelliformlesion.Inthepresentstudy,weperformedanin-depthanalysisofSD-OCTimagesinalargecohortofpatientswithidio-pathicERM,andproposedthatthecottonballsign,foveo-lardetachment,andacquiredvitelliformlesionmaycompriseacontinuuminthesameclinicalspectrumandmayrepresentprogressivestagesoftractionalCBabnor-malities.Inourseries,theseabnormalitieswereconsis-tentlylocatedinthesamecentralregionofthefoveacorrespondingtotheCB,conrmingthatthecentralfoveamaybeparticularlysensitivetomechanicalstresslikelyoriginatingfromMu¨llercellsthatareinterleavedwithcones. ¨llercellsextendfromtheinternaltotheexternallimitingmembran

e,wheretheyconnectwiththephotore-ceptorsthroughzonulaeadherens.Althoughourunder-standingoffovealMu¨llercellsisincomplete,recentevidencesuggesttheexistenceofanintimaterelationshipbetweenfovealMu¨llercellsandconesynapticterminals.Inthisregard,someauthorshavesuggestedthatthenumberof¨llercelltrunksinthecentralfoveamayequal(orevenexceed)thenumberoffovealcones.Wepresentedhisto-pathologicimages(Figure8)thatillustratedifferentialstainingoftheMu¨llercellsandvalidatetheintimaterela-tionshipofMu¨llercellswiththecentralconesandremark-ablydemonstratethepotentialforMu¨llercellstocausetractionupontheCB.Spaidehasproposedthattractionalforcesaretrans-mittedthroughretinaltissuetothephotoreceptorsbythe¨llercells.However,heconsideredthecentralfoveatobedevoidofMu¨llercellsandsuggestedamoreeccentric,parafoveallocationofmechanicalstress.¨llercellsthatparticipateintheHenleberradiationemanatingfromparafoveaandperifoveadisplayaZ-shapedmorphology,withtheouterprocessesintimatelycoupledwiththecentrifugallyrunningconeaxons(Figure9,Top).thefovealcenter,however,differentcongurationsof¨llercellswithintheCBhavebeenproposed.TheremaybeaninvertedconicalplugofspecializedMu¨llercells,namedMu¨llercellcones,comprisingtheoorofthefoveacentralis(Figure9,Bottom).InthedepictionbyGass,supportedbyYamada’searlyoriginalwork,theseMu¨llercellsappearmoreverticallyorientedinthecentral-mostre-gionofthefoveathanelsewhere.Inthismodel,Mu¨llercellshavethecapacitytotransmitmechanicalforcestothephotoreceptors.Therefore,wespeculatethatthedevelopmentofdifferentsubtypesofCBabnormalitymaydependonthestrengthandchro-nicityoftheinwardtractionexertedbyfovealMu¨llercellsovertheCB.Theproposedsequenceofeventsinthe FIGURE7.Spontaneousresolutionofanacquiredvitelliformlesion.(Top)Stage2epiretinalmembraneisassociatedwithacentralacquiredvitelliformlesion.Boththeellipsoidzoneandtheexternallimitingmembranearedisrupted,andanupwardmigrationofpigmentedhyperreßectivefociisidentiÞed(whitearrows).Thehyperreßectivedotsextendtotheborderoftheinnernuclearlayer.Visualacuityis20/60.(Bottom)Oneyearlater,theacquiredvitelliformlesioniscollapsedbuttheexternallimitingmembraneandellipsoidzonebandsarestilldisruptedanddiscontinuous.ECEMBER2017MERICANOURNALOFPHTHALMOLOGY developmentandevolutionoftractionalCBabnormalitiesisillustratedinFigure10andhasbeenvalidatedinthisstudywithqualitativeandquantitativelineardata.ThecottonballsignmayrepresenttheearliestandmostfrequentstageoftractionalCBabnormality(Figure10Middle-left).Inwardtractionmaycausetheupwarddisplacementoftheconephotoreceptorswithoutsigni-cantcompromiseoftheEZandELM,asreectedbythegoodvisualfunctionnotedinthiscohort.Infact,eyeswiththecottonballsignmaintainedoverallvisualandmorphologicstabilityoverthefollow-upperiod,althoughanatomicprogressionwithBCVAlosswasdescribedin5ofthestudyeyes. Thepathophysiologicmechanismcausingthecharacter-isticSD-OCThyperreectivityofthecottonballsignisstillpoorlydened.WithSD-OCT,contrastisproducedbydifferencesintherefractiveindexandthescatteringpropertiesofdifferentretinallayers.Intheouterretina,waveguidi

ngbyphotoreceptorsmayleadtodifferentialreectivity.Changesinthenormalreectivityoftheouterretinaasnotedwiththecottonballsignmaythushavedifferentexplanations.Misalignmentofphotorecep-torsowingtoverticaltractionmightbeexpectedtoresultindecreased,notincreased,reectivity,asoccursoversubretinaldrusenoiddeposits.Ontheotherhand,increasedreectivityhasbeencorrelatedtossioningof FIGURE8.Histologicsectionsofthecentralbouquet,includingMullercells,innormalmonkeyandhumanretina.Eachretinaisshowninpanoramic(leftcolumn)anddetailed(rightcolumn)views.NoneoftheseexamplesresemblespreciselytheeyewithaÔÔMullercellconeÕÕillustratedbyYamadaandreprintedbyGass.Inthetopandbottompanels,theretinaisartifactuallydetachedfromtheretinalpigmentepitheliumÐchoroidcomplex(notshown).(Top)NormalMacaquemonkey.Inthecentralfoveaandextendingoutwardalongthefovealslope(yellowarrows),tissuesarestainedmoredarklythanelsewhere.Cysticchangeinthecentralfovealßoor,consistentwithswellingofMullercellprocesses,arepresent(yellowasterisk,magniÞedinsets).ThepatternofdarkerstainingintheinnerandouterretinaisconsistentwithstainingofMullercells.Becausethesestainingdifferencesarevisibleinarapidlypreservedeye,theycannotbeattributedtopostmortemdelaytopreservation.(Middle)Humaneye(86yearsold).Darkerstaining(yellowarrows)isvisibleinthecentralretina,extendingalongthefovealslopeandintotheinnerandouterplexiformlayers.Histologicsectionavailableinitsentiretyat.(Bottom)Humaneye(87yearsold).Theapparentabsenceofdarkerstaininginthiseyecomparedtotopandmiddlepanels(yellowarrows)maybeattributabletodifferencesinpostÞxationmethods.However,theverycentralfoveahaspaleMullercellcytoplasmandtheconephotoreceptors(whitearrowsatinnersegments)arestainedmoredarklythantheinterveningMullercells,sospecialpropertiesarestilldetectable.Similartothemonkeyretina,cysticchangeinthecentralfovealßoor,consistentwithswellingofMullercellprocesses,isagainpresent(yellowasterisk,magniÞedinsets).AsmallepiretinalmembranecanalsobeidentiÞed(blackarrow),aswellasadark-stainingFriedmanlipidglobule(greenarrow).Noteinall3panelsthedifferentialstainingoftheMullercellsinthefoveathatdisplaysapatternsupportingthemechanismfortractionuponthecentralbouquet.GCLGCLganglioncelllayer;HFLHFLHenleÞberlayer;INLINLinnernuclearlayer;ONLONLouternuclearlayer;RPERPEretinalpigmentepithelium.RACTIONALBNORMALITIESOFTHE mitochondriawithinconeinnersegmentsinouterretinaltubulationofage-relatedmaculardegeneration.Further,increasedreectivitythatisverticallyalignedandlocatedbetweentheEZandtheRPE(Figure5)maybeattributabletodisorganizeddiscmembranesassociatedwithdegeneratingoutersegments,asrecentlyillustratedinamousemodel. Thepresenceofashallowfoveolardetachmentwaslesscommonlyencounteredinthestudypopulation,andmayrepresentamoreadvancedstageofCBabnormalityFigure10,Middle).Iftheinwardtractionexertedbythe¨llercellsovertheconesexceedstheadhesiveforcebe-tweenthephotoreceptorsandtheRPE,theconesmayseparateordetachcompletelyfromtheRPE,withthe FIGURE9.CartoondiagramsillustratingthearchitectureofMullercellsintheparafoveaandcentralfovea.(Top)Z-sha

pedparaf-ovealMullercells.ParafovealMullercells(1mmfromthefovealcenterwhereHenleÞbersarethelongest)displayaz-shapedmorphologyandconnectwiththecentralconesinthefovea,formingtightjunctionswithinnersegmentsofthephotoreceptorsthatcomprisetheexternallimitingmembrane.Mullercellouterprocessesfollowaz-shapedcourse,migratingawayfromthefovealTheend-feetofthefovealMullercellsformtheinternallimitingmembrane(ILM)intheparafovealregion,andtheMucellnucleiarelocatedintheinnernuclearlayer.(Bottom)Centralfovea.Theinnerhalfofthefoveolamayormaynotbecomposedofaninvertedcone-shapedzoneofMullercells.Itsbroadbaseislocatedintheperifovealregionandextendstoanapexlocatedattheexternallimitingmembrane.ThelocationoftheMullercellconenucleiisunclear.TheseMullercellsaremoreverticallyorientedinthecentral-mostregionofthefoveavsotherlocations,andtheydonothavesidebranches,astherearenoplexiformlayersinthecentralfovea.Inbothmodels(TopandBottom),Mullercellscantransmitmechanicalstressestothecentralbouquet(arrows).ThedesignoftheMullercellinthisdrawingisinspiredfromSevillaandCurcio.ECEMBER2017MERICANOURNALOFPHTHALMOLOGY subsequentaccumulationofsubretinaluid.Althoughvariouspathophysiologicalmechanismsmaycausetheaccumulationofuidinthesubretinalspace(eg,reduceduveal-scleraloutow,hemodynamicchangesinchoroidalow,breakdownoftheRPE,anddisruptionandleakagefromtheretinalcapillarysystem),inourstudypopulationthetractionaloriginofthefoveolardetachmentisthemostlikelyscenario.ThissubsetofeyeswasfrequentlynotedtodisplayadisruptedEZ,althoughtheELMwasgenerallypreserved.Thesemoreadvancedouterretinalal-terationswereassociatedwithalowerBCVAnotedinthiscohort.ThedevelopmentofanacquiredvitelliformlesionmayrepresentthenalstageofCBprogression(Figure10 Middle-right).Thechronicseparationoftheconephotore-ceptorsfromtheunderlyingRPEmaymoresignicantlydisruptthemorphologyandphysiologyoftheRPE-photoreceptorcomplex,leadingtoprogressiveaccumula-tionofphotoreceptordebrisinthesubretinalspace.IftheaccumulationofmetabolicconedebrisovercomesthephysiologiccapacityoftheRPE,anacquiredvitelliformlesionmaydevelop.RPEcellsandextrudedRPEorganellesmayalsocontributetoacquiredvitelliformlesions,asreportedbyChenandassociatesandBalaratnasingamandassociates.Theacquiredvitelliformlesionstendedtoincreaseinsizeduringthefollow-upperiodandspontaneouslycollapsedin1case,inwhichSD-OCTimagingillustratedtheupward FIGURE10.ProposedsequentialevolutionoftractionalabnormalitiesofthecentralbouquetasillustratedwithcartoonmodelsandopticalcoherencetomographyB-scans.(Left)Physiologicfovea.Mullercellsareinterconnectedwiththecentralcones.TheexternallimitingmembraneisthelinearaggregationoftightjunctionsbetweenMullerandphotoreceptorcells.Inphysiologicconditions,pho-toreceptorsaremaintainedincorrectalignment.Microvilliextendfromtheapicalsurfaceoftheretinalpigmentepithelium(RPE)cellstoenveloptheconeoutersegments.Withspectral-domainopticalcoherencetomography(SD-OCT),theexternallimitingmem-brane,ellipsoidzone,andinterdigitationzonearecontinuous,anddiscretehyperreßectivelines,runninginparallelwiththeRPE,arecle

arlyidentiÞed.Inthisschematiccartoon,Mullercellsaredrawnwithastraightmorphologybyconventionwithoutsidebranches.ThedesignoftheMullercellinthisdrawingisinspiredfromSevillaandCurcio.(Middle-left)Cottonballsign.Theinwardtractionofanepiretinalmembranedisplacesthecentralbouquet,andthenormalphotoreceptoralignmentislost.WithSD-OCT,thisalter-ationcausesachangeinthereßectivityoftheouterretina,withtheappearanceofafuzzyhyperreßectiveareabetweentheinterdig-itationzoneandellipsoidzone.(Middle)Foveolardetachment.MoreprogressiveinwardtractionleadstotheseparationofthecentralbouquetfromtheunderlyingRPE,withsubsequentaccumulationofcentralsubretinalßuid.Theexternallimitingmembrane,ellip-soidzone,andinterdigitationzonebandsassumeadome-shapedconÞguration.(Middle-right)Acquiredvitelliformlesion.ChronicseparationoftheconephotoreceptorsfromtheRPEdisruptsthenormalphysiologyoftheRPE-photoreceptorcomplex.Asaresult,thereisprogressiveaccumulationofmetabolicconedebrisinthesubretinalspace,whichmayovercomethephagocyticcapacityoftheRPE.RPEcellsandorganellesmayalsocontributetothevitelliformlesion.WithSD-OCT,athickdome-shapedhyperreßectivelesionisnotedbelowatypicallydisruptedellipsoidzone.(Right)Migrationofvitelliformmaterialintotheouterretina.Chronictrac-tionmaycausedisruptionoftheexternallimitingmembrane.Hyperreßectivefocimayrepresentstructuralalterationsintheouterretinaormaterialdepositionorpigmentmigration.RACTIONALBNORMALITIESOFTHE intraretinalmigrationofhyperreectivefoci,asdisplayedinFigure10(Right).ThisphenomenonwasaccompaniedbyELMdisruption,andhasbeenpreviouslydescribed.15,28 Histologicevaluationofeyeswithage-relatedmaculardegen-erationhasdemonstratedthatintraretinalhyperreectivefocimayrepresentcellsofRPEorigin,containinglipofuscin FIGURE11.SimpliÞedstaticmodelofmechanicalforcetransmissionthroughthefoveawithandwithoutectopicinnerfoveallayers.(Top-left)Absenceofectopicinnerfoveallayers.Intheabsenceofectopicinnerfoveallayers(ie,stage2epiretinalmembrane),thetractionalforceexertedbythemembraneistransmitteddirectlytotheouterretina,withsubsequentverticaldisplacementofthecentralbouquet.Inthisepiretinalmembranesubtype,thedevelopmentoftractionalabnormalitiesofthecentralbouquetiscommon.(Top-right)Presenceofectopicinnerfoveallayers.Inthepresenceofectopicinnerfoveallayers(ie,stage3and4epiretinalmem-brane),thetractionalforceexertedbythemembraneistransmittedtotheouterretinathoughtheectopicinnerfoveallayers.InasimpliÞedstaticmodel,thetotalforceÔÔfeltÕÕbybothlayersisthesame,butthisforceisdistributedoveralongerverticallengthandthereforethedeformationoftheouterretinaandthedisplacementofthecentralbouquetarelesspronouncedowingtotheectopicinnerfoveallayers.Therefore,tractionalabnormalitiesofthecentralbouquetareuncommoninthepresenceoftheectopicinnerfoveallayers.(Bottom)Proposedstaticmodelofforcetransmissioninthepresenceofectopicinnerfoveallayers.Thesystemismodeledasasetofmassesandspringsinseries.Eachlayer(ie,theouterandinnerretina)isconnectedbyspringswithstiffnesscon-stantsof0and0.ThefundamentalequationisHookeÕslaw,law,Ki$Xi,meaningtheforce

feltbyeachlayerisequaltothespringstiffnesstimesthestretchofthespring().Ifforceisexertedontheectopicinnerfoveallayers,thenviaNew-tonÕssecondlawofmotion,wehave2equationsofmotion:0(1)0(2)Equationsindicatethatisequivalentto),whichisequivalentto.Theequationsalsoindicatethatthedisplacementofthemasseswillbedifferent,with.ThissimpliÞedstaticmodelispurelyillus-trativeanddoesnotaccountforthedynamicsofthesystemorallthephysicalpropertiesoftheretinallayers.ECEMBER2017MERICANOURNALOFPHTHALMOLOGY granules,melanolipofuscingranules,andmelanosomes.ItisunclearwhethertheverticalorientationoftheMu¨llercellconemayfacilitatetheupwardmigrationofthehyperreec-tivefoci.Inthepresentstudy,theanatomicprogressionfromearliertolaterstagesoftractionalCBabnormalitieswasaccompaniedbyvisionlossinmostofthecases.ItisunclearwhyCBabnormalitiesdevelopmorefrequentlyinasubsetofERMs.Tsunodaandassociatesfoundasigni-cantassociationbetweenthecottonballsignandthickerCFT,suggestingthatincreasedretinalthicknessmaypredis-posetoCBabnormalities,andingthatwasnotconrmedinthepresentstudy.Infact,inourseriesthepresenceofCBabnormalitieswasstronglyassociatedwithstage2ERMandtheabsenceofEIFL,ratherthanwithCFT.Thisrelationshipmaybeexplainedwithasimpliedmodelofmechanicalforcetransmissionthroughtheretina,inwhichthemultilayeredstructureoftheretinaismodeledtoasetofmassesandspringsinseries(Figure11).Wespec-ulatethatthepresenceoftheEIFL(ie,stage3and4ERM)maybeprotectiveagainstthedevelopmentofCBabnor-malitiesbydecreasingthedisplacementandstretchingoftheouterretinainthefovealregion.Finally,itshouldbenotedthattractionalabnormalitiesoftheCBmaybeappreciatedinothermacularconditionsotherthanvitreomacularinterfacedisorders.Thepresenceoffoveolardetachmenthasbeendescribedinpatientswithcystoidmacularedemasecondarytodiabeticretinopathy,retinalveinocclusion,andposterioruveitis.cystoidmacularedema,intraretinaluidaccumulationmayleadtoexpansionoftissueandstretchingofthe¨llercells,withsubsequenttransmissionofmechanicalstressovertheCBleadingtotractionalabnormalitiessuchasfoveolar(orevenfoveal)detachment. Thisstudyhadseverallimitations,includingitsretro-spectivenatureandlimitedfollow-up.AlthoughthetotalnumberofstudiedeyeswithERM(263eyes)andCBab-normalities(58eyes)wasrobust,only12eyeswerenotedtohavedocumentedprogression,andthereforethepro-gressiveandsequentialnatureofthesechangesneedsfurtherconrmation.StrengthsofourstudyincludedtheuseofSD-OCTeye-trackingsystemsinallcases,whichallowedpreciseanalysisofCBchangesoverthefollow-upperiod.Inconclusion,thisreportinvestigatedthefunctionalandanatomicsignicanceoftractionalCBabnormalitiesinalargeseriesofeyeswithidiopathicERMs.Thecottonballsign,foveolardetachment,andacquiredvitelliformlesionmaybepartofthesameclinicalspectrumandmayrepresentprogressivestagesofdevelopment.TheanatomicprogressiontolaterstageswasassociatedwithanincreaseinthewidthoftheCBabnormalityandwithadeclineinBCVAinmostcases.ThepresenceoftheEIFLmaybepro-tectiveagainstthedevelopmentoftractionalCBabnor-malities.FovealMu¨llercellsmayplayanintegralroleinthedevelopmentoftheseabn

ormalitiesbytransmittingmechanicalforcestothefovealconesintheCB.Wehopethatourresultswillencouragethedesignoffutureprospectiveresearch,necessarytoconrmthesequentialstagesofdevelopmentintractionalCBabnor-malities.Additionalclinicopathologicstudiesarealsoneededtobetterunderstandthemicroanatomyoffoveal¨llercells,whichmaybecriticalinthepathophysiologyofCBabnormalitiesandothervitreomaculartractionaldisorders.Futurestudiestoevaluatetheimpactoftrac-tionalCBabnormalitiesonsurgicalprognosisarerecom- FUNDING/SUPPORT:MACULAFOUNDATIONINC,NEWYORK,USA.FINANCIALDISCLOSURES:K.BAILEYFREUND:CONSULTANTforGenentech,Inc(SanFrancisco,California,USA),HeidelbergEngineering(Heidelberg,Germany),OptovueInc(Fremont,California,USA),Optoplc(Dunfermline,Scotland,UK),andGraybugVisionInc(RedwoodCity,California,USA);researchsupportfromGenentech,Inc(SanFrancisco,Cal-ifornia,USA)andRoche(Basel,Switzerland).ChristineA.Curcio:ConsultantforGenentech,Inc(SanFrancisco,California,USA),Merck(White-houseStation,NewJersey,USA)JanssenCellTherapy(Raritan,NewJersey,USA);researchsupportfromHoffmanLaRoche(Basel,Switzerland).Jean-PierreHubschman:consultantforAlcon(FortWorth,Texas,USA),Pixium-Visium(Paris,France),Allergan(Parsippany-TroyHills,NewJerseUSA),andAvalancheBiotechnologies(MenloPark,California,USA).DavidSarraf:consultantforAmgen,Bayer(Leverkusen,Germany),Genentech(SanFrancisco,California,USA),Novartis,Novelution,andOptovue(Fremont,California,USA).ResearchgrantsfromAllergan(Parsippany-TroyHills,NewJersey,USA),Genentech(SanFrancisco,California,USA),HeidelbergEngineering(Heidelberg,Germany),Optovue(Fremont,CaliforniUSA),andRegeneron(Tarrytown,NewYork,USA).Thefollowingauthorshavenonancialdisclosures:AndreaGovetto,KavitaV.Bhavsar,GianniVirgili,MatthewJ.Gerber,andClaudeF.Burgoyne.AllauthorsattestthattheymeetthecurrentICMJEcriteriaforauthorship.PolyakSL.TheRetina.Chicago:UniversityPress;1941:455–PolyakSL.TheVertebrateVisualSystem.Chicago:Univer-sityPress;1957:259–278Rochon-DuvigneaudA.RecherchessurlafoveadelaretinehumaineetparticulierementsurlebouquetdescoArchAnatMicrosc TsunodaK,WatanabeK,AkiyamaK,UsuiT,NodaT.High-lyreectivefovealregioninopticalcoherencetomographyineyeswithvitreomaculartractionorepiretinalmembrane.GassJD.Mu¨llercellcone,anoverlookedpartoftheanatomyofthefoveacentralis:hypothesesconcerningitsroleinthepathogenesisofmacularholeandfoveomacularretinoschisis.ArchOphthalmolMatetA,SavastanoMC,RispoliM,etal.Enfaceopticalcoherencetomographyoffovealmicrostructurein RACTIONALBNORMALITIESOFTHE full-thicknessmacularhole:amodeltostudyperifoveal¨llercells.AmJOphthalmolWilliamsDS,ArikawaK,PaallysahoT.Cytoskeletalcompo-nentsoftheadherensjunctionsbetweenthephotoreceptorsandthesupportiveMu¨llercells.JCompNeurolGuidryC.TheroleofMu¨llercellsinbrocontractiveretinalProgRetinEyeResKrebsW,KrebsIP.Quantitativemorphologyofthecentralfoveaintheprimateretina.AmJAnatYamadaE.Somestructuralfeaturesofthefoveacentralinthehumanretina.ArchOphthalmolBurrisC,KlugK,NgoIT,SterlingP,ScheinS.HowMu¨llerglialcellsinmacaquefoveacoatandisolatethesynapticterminalsofcone

photoreceptors.JCompNeurol2002;453(1):100–111PisonA,DupasB,CouturierA,RothschildPR,TadayoniR.Evolutionofsubfovealdetachmentssecondarytoidiopathicepiretinalmembranesaftersurgery.DupasB,TadayoniR,ErginayA,etal.Subfovealdepositssecondarytoidiopathicepiretinalmembranes.ChowersI,TiosanoL,AudoI,GruninM,BoonCJ.Adult-onsetfoveomacularvitelliformdystrophy:afreshperspective.ProgRetinEyeResFreundKB,LaudK,LimaLH,SpaideRF,ZweifelS,YannuzziLA.Acquiredvitelliformlesions:correlationofclinicalndingsandmultipleimaginganalyses.Retina2011;31(1):13–25GovettoA,LalaneRA3rd,SarrafD,FigueroaMS,HubschmanJP.Insightsintoepiretinalmembranes:presenceofectopicinnerfoveallayersandanewopticalcoherenceto-mographystagingscheme.AmJOphthalmolDolz-MarcoR,LittsKM,TanACS,FreundKB,CurcioCA.Theevolutionofouterretinaltubulation,aneurodegenera-tionandgliosisprominentinmaculardiseases.SevillaMB,McGwinGJr,LadEM,etal.Relatingretinalmorphologyandfunctioninagingandearlytointermediateage-relatedmaculardegenerationsubjects.AmJOphthalmolSpaideRF.Closureofanouterlamellarmacularholebyvit-rectomy:hypothesisforonemechanismofmacularholefor-20.PretiRC,GovettoA,FilhoRGA,etal.Opticalcoherencetomographyanalysisofouterretinaltubulations:sequential evolutionandpathophysiologicalinsights.CooperRF,DubisAM,PavaskarA,RhaJ,DubraA,CarrollJ.Spatialandtemporalvariationofrodphotoreceptorreectanceinthehumanretina.BiomedOptExpress2011;2(9):2577–2589ZhangY,WangX,BlancoE,etal.Photoreceptorperturba-tionaroundsubretinaldrusenoiddepositsrevealedbyadap-tiveopticsscanninglaserophthalmoscopy.AmJOphthalmolLittsKM,MessingerJD,ZhangY,FreundKB,CurcioCA.In-nersegmentremodelingandmitochondrialtranslocationindegeneratingconesofage-relatedmaculardegeneration,includingouterretinaltabulation.InvestOphthalmolVisSciSchaalKB,FreundKB,LittsKM,ZhangY,MessingerJD,CurcioCA.Outerretinaltubulationinadvancedage-relatedmaculardegeneration:opticalcoherencetomo-graphicndingscorrespondtohistology.BonilhaVL,BellBA,RaybornME,etal.LossofDJ-1elicitsretinalabnormalities,visualdysfunction,andincreasedoxidativestressinmice.ExpEyeResSpaideRF.Autouorescencefromtheouterretinaandsubre-tinalspace:hypothesisandreview.ChungJE,SpaideRF.Fundusautouorescenceandvitelliformmaculardystrophy.ArchOphthalmol2004;122(7):1078–1079ChenKC,JungJJ,CurcioCA,etal.Intraretinalhyperreec-tivefociinacquiredvitelliformlesionsofthemacula:clinicalandhistologicstudy.AmJOphthalmolBalaratnasingamC,MessingerJD,SloanKR,YannuzziLA,FreundKB,CurcioCA.Histologicandopticalcoherencetomographiccorrelatesindrusenoidpigmentepitheliumdetachmentinage-relatedmaculardegeneration.GaucherD,SebahC,ErginayA,etal.Opticalcoherenceto-mographyfeaturesduringtheevolutionofserousretinaldetachmentinpatientswithdiabeticmacularedema.AmJYamaguchiY,OtaniT,KishiS.Serousmaculardetachmentinbranchretinalveinocclusion.2006;26(9):1029–1033IannettiL,SpinucciG,AbboudaA,DeGeronimoD,TortorellaP,AccorintiM.Spectral-domainopticalcoherencetomographyinuveiticmacularedema:morphologicalfeaturesandprognosticfactors.Ophthalmologica2012;228(1):13–18ECEMBER2017MERICANOURNALOFPHTHAL