PrevalenceandSignificanceofIncidentallyNotedDilationof - PDF document

PrevalenceandSignificanceofIncidentallyNotedDilationof theAscendingAortaonRoutineChestComputed TomographyinOlderPatients NancyBenedetti,MDandMichaelD.Hope,MD Objective: Theaimofthisstudywastodeterminetheprevalenceofin- cidentalascendingaorticdilationanditssignificanceovertimein55-to 80-year-oldsundergoingroutinecomputedtomographicscans. Methods: Chestcomputedtomographyreportsfor64,092patientswho mettheinclusioncriteriawereusedtodeterminetheprevalenceofinciden- talascendingaorticdilation(4-5cm)and,whenpossible,aorticgrowth rates.Achartreviewwasperformedtoidentifyanyaorticcomplication orintervention. Results: Theprevalenceofincidentalaorticdilationwas2.7%(671/ 24,992patients).Ofthe327patientswithaorticdilationandfollow-up studies(mean,3.4years),only3.7%(n=12)demonstratedintervalgrowth (meanof0.9mm/y).Nopatientunderwentprophylacticsurgeryorinter- ventiononthebasisofaorticsizeorgrowthrate.Onepatientdeveloped atypeAdissection. Conclusions: Currentguidelinesforyearlysurveillanceimagingofaor- ticdilationcouldberevisedtoincreasethefollow-upintervaland/orim- proveriskstratificationtobetteridentifythesmallsubsetofpatients mostlikelytohavediseaseprogression. KeyWords: aorticdilation,surveillanceimaging,incidentalfindings ( JComputAssistTomogr 2014;00:00 – 00) T hediscoveryofincidentalthoracicabnormalitieswillbecome increasinglymorecommonwiththecommencementoflung cancerscreeningintheUnitedStates. 1 – 4 Althoughincidentalim- agingfindingsoccasionallyleadtotheearlydetectionofpoten- tiallydeadlydiseaseandimmediatetreatment,manyareof limitedclinicalsignificancebutresultinfollow-upstudiesthat substantiallyaddtohealthcarecostandpatientanxiety. 4,5 Inci- dentaldilationoftheascendingaortafallsintothiscategory,with increasingprevalenceinrecentyears,likelybecauseofincreased detectionbymorewidespreaduseofimaging. 6 Potentiallydeadly complicationssuchasaorticdissectionorrupturearerare,but guidelinesrecommendyearlyimagingforsurveillance. 7 – 9 Aorticdilationinanolderscreeningpopulationistypically anindolentdiseasewithveryfewresultantdeaths. 10,11 Thereare importantexceptionswherediseaseprogressionismorecommon, notablyinpatientswithconnectivetissuedisease,aorticvalvedis- ease,orfamilyhistoryofaorticdissection.Intheabsenceofsuch conditions,yearlyimagingsurveillanceisrecommendedforas- cendingaorticdilationinthe4-to5-cmrange. 7,11 Theseguide- linesarebasedonpreviousstudiesshowingthefollowing:(1)the upperlimitofnormalforaorticdiameterinadultsis4.0cm 12 ; (2)aorticdilationrarelyleadstodeathwhendiametersareless than6.0cm,withmorefrequentcomplicationsbeyondthispoint; and(3)ascendingaorticgrowthratesaresmall,averagingapprox- imately0.1cm/y. 10,11,13 Weconductedastudytodeterminetheprevalenceofinci- dentallynoteddilationoftheascendingaortaanditssignificance overtimein55-to80-year-olds.Theobjectivewastoevaluatethe impactofcurrentguidelinesformanagementandfollow-upimag- ingofthisconditionwhenidentifiedonroutinechestcomputed tomography(CT)inolderpatients. MATERIALSANDMETHODS PatientSelection Waiverofinformedconsentwasobtainedforthisretrospec- tive,HealthInsurancePortabilityandAccountabilityAct- compliantstudy.Atotalof88,171chestCTscansperformedon 31,963patientsaged55to80yearsatourinstitutionduringa 14-yearperiod(January1,2000,toDecember31,2013)were identifiedthroughanelectronicsearchofourradiologyinforma- tionsystemusingIlluminate(SoftkSolutionsInc,PrairieVillage, Kan)(Table1).Patientdemographicinformation,examination type(eg,non – contrastenhancedvscontrastenhanced),andthe radiologyreportwereobtainedaspartofthesearch.Patientswere excludediftheyhadanaorticaneurysmmeasuringgreaterthan 5cm,known/repairedaorticaneurysmordissection,aorticvalve repairorcongenitalabnormality,mycoticaneurysm,orhistoryof connectivetissuediseasesuchasMarfanandEhlers-Danlossyn- dromes.Examinationswerealsoexcludediftheywereobtained atanoutsidehospitalanduploadedintooursystemorifthey wereobtainedsolelyforanimage-guidedprocedure. DataAnalysis Aftertheseexclusions,theprevalenceofreportedascending aortadilation(4-5cmindiameter)wasdetermined.Forpatients withatleast2CTscansofthesametype(non – contrastenhanced orcontrastenhanced)6monthsapartorlonger,chestCTreports werereviewedtodeterminewhethertherewasintervalaortic growthreported.Anyreportedincreaseinaorticsizewasthen confirmedby2independentreviewersusingorthogonalmeasure- mentoftheascendingaorta. 12,14 Acomprehensivechartreview wasalsoperformedtoassessfordiseaseprogressionand/or vascularintervention. RESULTS PrevalenceofAorticDilation Fromtheinitialsearchof85,171relevantCTexaminations on31,963patients,21,097examinationsand6971patientswere excluded,leaving64,092examinationson24,992patientsmeet- ingtheinclusioncriteria.Ofthe64,092examinations,32,594 wereperformedonmen(50.9%)and31,498wereperformedon women(49.1%).Themeanandmedianageswere66and65years old,respectively.Halfofthepatientswerestudiedfordiagnosisor FromtheDepartmentofRadiologyandBiomedicalImaging,Universityof California,SanFrancisco,SanFrancisco,CA. Receivedf

orpublicationJune27,2014;acceptedSeptember8,2014. Reprints:MichaelD.Hope,MD,DepartmentofRadiologyandBiomedical Imaging,UniversityofCalifornia,SanFrancisco,505ParnassusAve, Box0628,SanFrancisco,CA94143-0628(e
mail:michael.hope@ucsf.edu). Theauthorsdeclarenoconflictofinterest. Copyright©2014byLippincottWilliams&Wilkins O RIGINAL A RTICLE JComputAssistTomogr  Volume00,Number00,Month2014www.jcat.org 1 follow-upofmalignancy,withtheremaindersplitbetweenthe followinggroups:evaluationofpulmonaryembolism,infection, nodules,andinterstitiallungdisease.Twothirdsofthescans werecontrastenhanced.Ofthe24,992patientsincludedinthe study,671patients(2.7%)hadreporteddilationoftheascending aortainthe4-to5-cmrange. Follow-up Ofthe671patientswithadilatedascendingaorta,327 (48.7%)hadserialCTexaminationsofthesametype(ie,non – contrastenhancedorcontrastenhanced)withinthe14-yearexam- inationperiod,withameanlengthoffollow-upof3.4years (Fig.1).Ofthese327patients,12(3.7%)hadconfirmedgrowth oftheascendingaorta,withameanlengthoffollow-upof 5.0years.Themeangrowthratewas0.9mm/y,witharangeof 0.2to2.2mm/y.Ofthese327patients,222weremen(67.9%) and105werewomen(32.1%);182were65yearsoryounger (55.7%)and145wereolderthan65years(44.3%).Overall,the likelihoodofhavingincidentaldilationofthethoracicaortaand documentedintervalgrowthwas0.05%(12/24,992).Noelective surgeryorinterventionwasperformedonthebasisofaorticsize orgrowth.Asingleaorticcomplicationoccurredinthecohort:1 patientdevelopedatypeAdissection1yearafteranincidental notewasmadeofanascendingaorticdiameterof4.0cm. DISCUSSION Incidentallynoteddilationoftheascendingaortawasfound in2.7%ofthepatients55to80yearsoldundergoingroutinechest CTexaminations.Ofthepatientswithaorticdilationandfollow- upimaging,thevastmajorityhadstableaorticdimensions,with asmallgroup(3.7%)demonstratingmodestaorticgrowthrates of0.9mm/y.Nopatientinthecohortstudiedunderwentprophy- lacticsurgeryorinterventiononthebasisofaorticsizeor growthrate. Yearlysurveillanceimagingisrecommendedforpatients withdilatedascendingaortas( � 4cm),withsurgeryrecom- mendedatthresholdsof4.5to5.5cmdependingonconcomitant factorsincludingaorticvalvedisease,connectivetissuedisorder, orhistoryoffamilialaorticdissection. 7 Intheabsenceofthoserisk factors,aorticdilationisarelativelyindolentdiseaseintheel- derly. 10,11 Thepurposeofyearlysurveillanceimagingisto(1) identifypatientswithrapidaorticgrowthrates(ie, � 5mm/y) and/or(2)identifypatientsatorapproachingthe5.5-cmthreshold forsurgicalintervention. 8,9 OurstudysuggeststhatyearlyCTsur- veillance,however,istoofrequentandunlikelytoidentifypro- gressiveaorticdiseaseinacost-effectiveorefficientmanner. Giventheabsenceofgrowthinthemajorityofpatientsinourco- hort,andslowgrowthratesintheremainderofthegroup,which areequaltoorslowerthanthoseofsmall(
4cm)abdominalaor- ticaneurysms,the2 – 3yearscreeningintervalthathasbeenpro- posedforsmallabdominalaorticaneurysmsseemsreasonable forascendingaorticdilationwithoutotherriskfactors. 15,16 Themanagementofincidentalimagingfindingssuchasas- cendingaorticdilationwillbecomeincreasinglyimportantas screeningforandsurveillanceofmalignancybecomemorecom- mon. 4 Theprevalenceof2.7%thatwefoundforincidentaldila- tionoftheascendingaorta,whichissimilartootherreports (2.3%), 17 couldsubstantiallyimpacthealthcarespendingwhen appliedtolargescreeningpopulations(eg, � 7millionpatients forlungcancerscreening). 2 Yearlyandindefinitesurveillanceim- agingcouldresultinhundredsofthousandsoffollow-upstudies. TABLE1. StudyPopulation No.PatientsPercentage* Initialsearch31,963 „ Metinclusioncriteria24,99278.2% Dilatedascendingaorta6712.7% Serialstudies32748.7% Intervalgrowth  123.7% Aorticcomplication1 ‚ 0.15% *Referstothepercentageofpatientsincludedfromthepreviousrow.  Meangrowthrateof0.9mm/y. ‚ TypeAdissection1yearafterincidentalnotewasmadeofa4-cmas- cendingaorta.Thenumber671isusedasthedenominatorforpercentage calculation. FIGURE1. Percentageofpatientsstudiedwithincidentallynoteddilationoftheascendingaorta,follow-upstudies,andintervalgrowth forameanof3.4years. BenedettiandHope JComputAssistTomogr  Volume00,Number00,Month2014 2 www.jcat.org ©2014LippincottWilliams&Wilkins Onthebasisofoursingle-centeranalysis,however,surveillance imaginginitscurrentformisunlikelytoimpactaorticdisease – relatedmorbidityormortality. The1patientinourcohortwithcleardiseaseprogression (typeAdissection)hadanascendingaorticdiameterof4.0cm 1yearearlier.Thishighlightsthechallengesofevaluatingrisk foraorticdiseaseprogression „ aorticdissectionisaveryrareevent, intherangeof30casesper1millionindividualsperyear 18 „ and thelimitationsofusingaorticdimensionsalone.Thesinglecase reflectsthefindingsofpriorworkthatshowthatthemajorityof casesofaorticdissectionoccursatdiameterslowerthanthecom- monsurgicalthresholdof5.5cm. 19 Itsuggeststhatabetterap- proachisneededforidentifyingthes mallsubsetofpatientswith dilatedaortaswhoareatimminentriskforaorticcatastrophe.Ad- vancedimagingmaybeofuse,withne

wfunctionalandmolecular imagingtechniquesrevealingpreviouslyunseendriversofdisease suchasabnormalhemodynamicsandvascularinflammation. 20 Ourstudyislimitedtoasingleacademiccenter.Radiology reportswererelieduponformostoftheimagingdata,andthus, theprevalenceofaorticdilationmayhavebeenunderreported. However,webelievethatthereportingofaorticdilationatourin- stitutionreflectsthenormamongradiologists.Notallpatientshad serialimaging,buttherewasnorelevantselectionbiasthatweare awareofforthosewhodidhavefollow-upstudies.Wefocusedon ascendingaorticdimensions,ratherthanontheentirethoracic aorta,butdidsobecauseascendingaorticaneurysmsarethemost commonandmeasurementsherearebeststudiedandcommonly usedforscreening. 8,12 Thepatientsincludedwerenotnecessarily smokers,andforpopulationsofsmokerssuchasthoseundergo- inglungcancerscreening,vasculardiseasemaybemorecommon thaninourcohort. Inconclusion,incidentalnoteofascendingaorticdilationis relativelycommoninolderpatientsundergoingroutineCTscans. Thevastmajorityofpatientsdemonstratesstableaorticdimen- sionsovertime.Nopatientunderwentprophylacticsurgeryorin- terventiononthebasisofaorticsizeorgrowthrate.Current guidelinesforyearlysurveillanceimagingofascendingaorticdi- lationcouldberevisedtoincreasethefollow-upintervaland/or improveriskstratificationtobetteridentifythesmallsubsetofpa- tientsmostlikelytohavediseaseprogression. REFERENCES 1.MoyerVA.Screeningforlungcancer:U.S.PreventiveServicesTaskForce recommendationstatement. AnnInternMed .2014;160:330 – 338. 2.AberleDR,AdamsAM,BergCD,etal.Reducedlung-cancermortality withlow-dosecomputedtomographicscreening. NEnglJMed .2011;365: 395 – 409. 3.deKoningHJ,MezaR,PlevritisSK,etal.Benefitsandharmsofcomputed tomographylungcancerscreeningstrategies:acomparativemodeling studyfortheU.S.PreventiveServicesTaskForce. AnnInternMed .2014; 160:311 – 320. 4.JacobsPC,MaliWP,GrobbeeDE,etal.Prevalenceofincidentalfindings incomputedtomographicscreeningofthechest:asystematicreview. JComputAssistTomogr .2008;32:214 – 221. 5.SwensenSJ,JettJR,SloanJA,etal.Screeningforlungcancerwith low-dosespiralcomputedtomography. AmJRespirCritCareMed .2002; 165:508 – 513. 6.OlssonC,ThelinS,StahleE,etal.Thoracicaorticaneurysmand dissection:increasingprevalenceandimprovedoutcomesreportedina nationwidepopulation-basedstudyofmorethan14,000casesfrom1987to 2002. Circulation .2006;114:2611 – 2618. 7.CozijnsenL,BraamRL,WaalewijnRA,etal.Whatisnewindilatationof theascendingaorta?Reviewofcurrentliteratureandpracticaladvicefor thecardiologist. Circulation .2011;123:924 – 928. 8.IsselbacherEM.Thoracicandabdominalaorticaneurysms. Circulation . 2005;111:816 – 828. 9.HiratzkaLF,BakrisGL,BeckmanJA,etal.2010ACCF/AHA/AATS/ ACR/ASA/SCA/SCAI/SIR/STS/SVMguidelinesforthediagnosisand managementofpatientswiththoracicaorticdisease:areportofthe AmericanCollegeofCardiologyFoundation/AmericanHeartAssociation TaskForceonPracticeGuidelines,AmericanAssociationforThoracic Surgery,AmericanCollegeofRadiology,AmericanStrokeAssociation, SocietyofCardiovascularAnesthesiologists,SocietyforCardiovascular AngiographyandInterventions,SocietyofInterventionalRadiology, SocietyofThoracicSurgeons,andSocietyforVascularMedicine. Circulation .2010;121:e266 – e369. 10.DaviesRR,GoldsteinLJ,CoadyMA,etal.Yearlyruptureordissection ratesforthoracicaorticaneurysms:simplepredictionbasedonsize. AnnThoracSurg .2002;73:17 – 27discussion27 – 28. 11.PatelHJ,DeebGM.Ascendingandarchaorta:pathology,naturalhistory, andtreatment. Circulation .2008;118:188 – 195. 12.WolakA,GransarH,ThomsonLE,etal.Aorticsizeassessmentby noncontrastcardiaccomputedtomography:normallimitsbyage,gender, andbodysurfacearea. JACCCardiovascImaging .2008;1:200 – 209. 13.ElefteriadesJA.Thoracicaorticaneurysm:readingtheenemy'splaybook. CurrProblCardiol .2008;33:203 – 277. 14.CayneNS,VeithFJ,LipsitzEC,etal.Variabilityofmaximalaortic aneurysmdiametermeasurementsonCTscan:significanceandmethodsto minimize. JVascSurg .2004;39:811 – 815. 15.BradyAR,ThompsonSG,FowkesFG,etal.Abdominalaorticaneurysm expansion:riskfactorsandtimeintervalsforsurveillance. Circulation . 2004;110:16 – 21. 16.SantilliSM,LittooyFN,CambriaRA,etal.Expansionratesandoutcomes forthe3.0-cmtothe3.9-cminfrarenalabdominalaorticaneurysm. JVasc Surg .2002;35:666 – 671. 17.HunoldP,SchmermundA,SeibelRM,etal.Prevalenceandclinical significanceofaccidentalfindingsinelectron-beamtomographicscansfor coronaryarterycalcification. EurHeartJ .2001;22:1748 – 1758. 18.CriadoFJ.Aorticdissection:a250-yearperspective. TexHeartInstJ .2011; 38:694 – 700. 19.PapeLA,TsaiTT,IsselbacherEM,etal.Aorticdiameter � or=5.5cmis notagoodpredictoroftypeAaorticdissection:observationsfromthe InternationalRegistryofAcuteAorticDissection(IRAD). Circulation . 2007;116:1120 – 1127. 20.HopeMD,HopeTA.Functionalandmolecularimagingtechniquesin aorticaneurysmdisease. CurrOpinCardiol .2013;28:609 – 618. JComputAssistTomogr  Volume00,Number00,Month2014 DilationoftheAscendingAortaonChestCT ©2014LippincottWilliams&Wilkins www.jcat.org