Transcript
Page 1: Impact of Pre-Procedural Planning with 3D Printed Cardiac ...consistency of myocardium. In addition, a hard plastic model will be printed to provide a rigid replica of the cardiac

1

ImpactofPre-ProceduralPlanningwith3DPrintedCardiacModelsforVentricularAssistDevicePlacementinPatientswithCongenitalHeartDisease

Author: HannahTredway,MDFacultyMentor/PI: KanwalFarooqi,MDDate: August9,2019A.StudyPurposeandRationale:Congenitalheartdisease(CHD)isthemostcommontypeofmajorcongenitalmalformationandtheleadingcauseofmortality frombirthdefects.Up to twentyfivepercentofpatientswithCHDwillprogress to heart failure (HF) by age 30. Patients with complex CHD include thosewith lesionssuchasdoubleoutlet rightventricle (DORV), transpositionof thegreatarteries (TGA),and thosewithsingleventriclepalliation,amongothers.Thispopulationisat the highest risk for developingHF. Ventricular assist devices (VADs) are mechanical pumps, w h i c h are surgically placed inthe chest to augment cardiac output in patientswith heart failure.UtilizationofVADsinpatientswithCHDandHFremains raredueinparttothehighlyvariableanatomyandcomplexphysiologyinthispopulation.Anatomicfactors,whichplayaroleinpotentially complicating VAD placement,are numerousinpatientswithCHD.Specifically, ventricular dilationandpresenceoftrabeculationsand abnormal vascular anatomy make the delineation of inflow and outflow cannula placementp a r t i c u l a r l y challenging. While advanced imaging techniques such as cardiac magnetic resonance imaging (CMR) orcomputed tomography (CT) aid cardiothoracic surgeons in pre-surgical planning for VADplacementinthispopulation,itremainsachallengetoadequatelydepictallofthecomponentsofacomplexpatient’scardiacanatomyinatwodimensional(2D)imagingdataset.Three- dimensional(3D) printing,atechnologythatisquicklygainingutilityinthemedicalfield, enablesthe creationof patient- specific physical anatomic models from a patient’s 3D imaging data. 3D printedmodelsprovideaphysicalguide tounderstandingtheanatomic features that can make VAD andcannula placement challenginginpatientswithCHD.Thegoalofthisstudyistoassesstheutilityofusing3DmodelsintheplanningofVADplacementinpatientswithcomplexCHD.Wewillsurveycardiacsurgeonsregardingtheutilityofthe3Dmodelsinvisualizing theanatomyandplanninganapproach todevice implantationbeforeandafter theprocedure.Theseresultswillbecomparedtocontrols,inwhichtraditionaltwo-dimensionalimagesare used for pre-surgical planning. It is expected that the use of 3D models will enhanceunderstandingofcomplexanatomyandultimatelycutdownonproceduretimes.Wehypothesizethatpatient specific3Dprintedmodelswillallowmoreinformedpreoperativeplanningwithcleardemonstrationof thebest site for inflow cannula, outflow cannula and VAD placement leadingto better surgical preparedness, less operatingroomtimeandimprovedpatientoutcomes.B.StudyAims:AIM1:Toassess ifa3Dprintedcardiacmodel improvesvisualizationofVADandcannulaplacementsites inCHD-HFpatientsascompared to2D imaging.WewillprospectivelyenrollCHD-HFpatientsatmultiplecentersandrandomizetogroupA(3Dprintedmodelswillbeusedforpre-VAD planning) or Group B (controls). For patients in Group A, the surgeon performing theprocedurewillcompleteaquestionnaire1)afterreviewing2Dimagingdataand2)afterreviewingapatientspecific3Dmodel.Ourprimaryoutcomemeasurewillbeanimprovementintheclarityofcannula and VAD site demonstration. We hypothesize that the 3D models will more clearlydemonstratethesitesofcannulaandVADplacementascomparedto2Dimaging.

Page 2: Impact of Pre-Procedural Planning with 3D Printed Cardiac ...consistency of myocardium. In addition, a hard plastic model will be printed to provide a rigid replica of the cardiac

2

AIM 2: To determine if perioperative factors and patient outcomes improve in CHD-HFpatients with use of 3D printed model versus traditional imaging in VAD placementplanning. Clinical characteristics will be collected at the time of enrollment including primarydiagnosis and indication forVAD.AfterVADplacement, information regarding the intraoperativeandpostoperativecoursewillbecollectedincludingsurgicalcardiopulmonarybypasstime(CPB),need for cannula repositioningand lengthof stay. LongerCPB increasesmorbidity andmortalityand isassociatedwith intensivecare readmission inpatientsafterLVADplacement.Ourprimarymeasures of improvement will be CPB and length of stay. We hypothesize that the improvedpreoperativeplanningbasedonthe3Dmodelswill leadtoadecrease inCPBtimeanddecreasedlengthofstay.C.LocationoftheStudyThe primary center for this study is New York Presbyterian, Columbia University MedicalCenter. Nine other centers to date have agreed to serve as collaboratingcenters.

1.MontefioreMedicalCenter,AlbertEinsteinCollegeofMedicine,NewYork2.UniversityofIowa,Iowa3.MayoClinic,Rochester,Minnesota4.UniversityofMichigan,AnnArbor,Michigan5.UTSouthwestern,Texas6.WashingtonUniversitySchoolofMedicine,St.Louis,MO7.HarvardUniversity,Boston,MA8.HeartInstitute,MexicoCity,Mexico9.JohnsHopkins,Baltimore,MD10.CaseWesternReserveUniversity,Cleveland,OH11.NationwideChildren’sHospital,Columbus,OH12.MontrealHeartInstitute,Montreal,Canada

AllparticipatingcenterswillberequiredtocompletetheirownIRBinordertobepartofthisstudy.D.StudyDesignRegistryInclusionPatientsoverage13withcongenitalheartdiseaseandclinicalheartfailurewhoarecandidatesforMCS will be prospectively identified at the participating centers. After identification, informedconsent or assent will be obtained and preoperative clinical data will be collected. Exclusioncriteria:AnyCHD-HFpatientunabletotolerateaCMRorCTwillbeexcluded.TransferofImagesThe cardiac MRI or CT images will be uploaded to an online secure website, Lifeimage, by theparticipating institution. Lifeimage is a cloud based,medical image sharing and viewingwebsite,whichisHIPAAcompliantandsecure.Theimageswillthenbeaccessedforpost-processing.Imagepost-processingWewillbeable tousesoftware,which is licensedbytheDivisionofPediatricCardiology,Mimics(Materialise©, Belgium) to post process the cardiac imaging datasets. Post-processing involvesisolatingtheareaofinterest(AOI)bysegmentation.Bloodpoolsegmentationwillbeusedtocreatethe3Dmodels inallpatients.Aftersegmentation,apreliminary3Dvirtualmodelwillbecreated,which will be edited to exclude extra cardiac structures. The 3D object will be cropped to bestdemonstrate the relevant anatomy of interest for each patient and subsequently stored as astereolithographyfile(STL).

Page 3: Impact of Pre-Procedural Planning with 3D Printed Cardiac ...consistency of myocardium. In addition, a hard plastic model will be printed to provide a rigid replica of the cardiac

3

Rapidprototyping/3DprintingAStratasysJ7503Dprinterwillbeusedtocreatecardiacmodelswithfineresolutionandexcellentanatomic detail. This printer is able to print in a soft tissue like material, which will mimic theconsistency of myocardium. In addition, a hard plastic model will be printed to provide a rigidreplicaofthecardiacanatomyforeachpatient.Anadditionalhardplasticmodelwillbecreatedforour center (CUMC) to keep a library of themodels created for this study aswell as to ease anyfurther communication between our center and other centers regarding anatomic details of aspecificpatient.Aim1.Preoperative,intraoperativeandpostoperativeassessmentThesurgeonwill completeaquestionnaireonce the initialpatientdata including the2D imagingdatahasbeenpresentedtohimorher,priortothe3Dmodelbeingexamined.Oncethe3Dprintedmodel is received, the surgeon will fill a second questionnaire. The final questionnaire will becompletedinthepostoperativeperiodintoassesstheaccuracyandutilityofthemodel.Aim2.ClinicalvariablecollectionThe initial clinical variableswill be collected at the time of enrollment. After the surgical repair,additionalrelevant intraoperativevariablessuchasneed forcannularepositioning,CPBtimeandestimated blood loss will be collected. Our primary outcome variable will be cardiopulmonarybypasstime.OursecondaryanalysiswillassessforimprovementinVADflowsanddecreaseintheneedforrevisionofVADorcannulaplacement.30-daypostoperativeassessmentFurther datawill be collected at the 30-day time point,mainly relating to patient postoperativecourse, recoveryandclinicaloutcome.Wewillbe focusingon identifying improvement in30-daymortalityanddecreaseintheneedforVADrevisionwith3Dmodelusepreoperatively.Informationregardingpostoperativecourse includingtheneedfor inotropicorrespiratorysupport,renalandliverfunction,arrhythmia,bleedingcomplicationsorstrokewillalsobecollected.

Page 4: Impact of Pre-Procedural Planning with 3D Printed Cardiac ...consistency of myocardium. In addition, a hard plastic model will be printed to provide a rigid replica of the cardiac

4

E.StatisticalAnalysisForAim1, improvement in survey responses regarding clarity ofVADor cannula siteplacementwill be analyzed using chi-squared tests. Sample size calculations are based on improving thedelineationofanyoffouranatomicalelements (placementoftheinflowcannula,placementoftheoutflow cannula, possible cannula obstruction, and location of theventricularapex)usingthe3Dmodel. Wealsoassumeaone-sidedtestofsignificance,asitisnotexpected thatthe3Dmodelwillworsen the anatomical delineation compared with the usual 2D models. Since there are fourpossibleways to improve thedelineation,wecomputesamplesizeeffectsusinganalphavalueof0.0125, based on the Bonferroni criterion (0.05/4) for independent hypothesis tests. Assumingan effect size of 0.25, and62 subjects, theMcNemar testwouldhave80%power, assuming thatthe proportion of discordant pairs is 0.4. If recruitment is as little as29 subjects, then thestudywill have 80% power to detect an increase of 0.35. Because the Bonferroni criterion is veryconservativeingeneralandevenmoresointhecaseofdependenttests,weexpect theactualpowertoexceedthesenominalpowers.Additionally,forAim2,clinicalvariableswillbecomparedbetweenthecontrolandmodelgroups.Our main analysis of the collected clinical variables will be a comparison of cardiopulmonarybypass time using a t-test. 44 subjects in each arm (3Dmodel and control) will be recruited toachieve80%powerassuminga20-minutedecreaseincardiopulmonarybypasstime.Other outcome continuous variables include VAD placement time, estimated blood loss, days of

Figure1:StudyDesignFlowDiagram

Page 5: Impact of Pre-Procedural Planning with 3D Printed Cardiac ...consistency of myocardium. In addition, a hard plastic model will be printed to provide a rigid replica of the cardiac

5

intubation,andlengthofstay(LOS)willalsobeanalyzedusingt-testsorWilcoxonrank-sumtestsifthe data are not normally distributed. We will also analyze pre- and post- values for liver andkidneyfunctiontestsusingpairedt-testsandanalysisofcovariance.Categoricalvariablessuchaspresence/absenceof arrhythmia, inotropic support andheart failure classwill be analyzedusingchi-squaretestsandMcNemartests.Multipleregressionmodelswillalsobeestimated,e.g.,LOSasafunctionofthestudyarm,controllingforotherLOSpredictors.F.StudyQuestionnaires

Page 6: Impact of Pre-Procedural Planning with 3D Printed Cardiac ...consistency of myocardium. In addition, a hard plastic model will be printed to provide a rigid replica of the cardiac

6

G.ConfidentialityofStudyDataAREDCAP databasewill be created for data entry from each of the centers. Each center will beresponsible for submitting clinical variables as well as answers to the surgeon surveys at threedifferent timepoints.Allstudydatawillbecoded,andacodenumbershouldbeusedforallstudysubjects.Datawillbestoredinasecurelocation,accessibleonlytotheinvestigators.

Page 7: Impact of Pre-Procedural Planning with 3D Printed Cardiac ...consistency of myocardium. In addition, a hard plastic model will be printed to provide a rigid replica of the cardiac

7

Citations:

1. Norozi,K., et al., Incidenceandriskdistributionofheart failure inadolescentsandadultswithcongenitalheartdiseaseaftercardiacsurgery.AmJCardiol,2006.97(8):p.1238-43.

2. Brida,M.,G.P.Diller,andM.A.Gatzoulis,SystemicRightVentricleinAdultsWithCongenitalHeartDisease:AnatomicandPhenotypicSpectrumandCurrentApproachtoManagement.Circulation,2018.137(5):p.508-518.

3. Fontan,F.andE.Baudet,Surgicalrepairoftricuspidatresia.Thorax,1971.26(3):p.240-8.4. Piran,S.,etal.,Heartfailureandventriculardysfunctioninpatientswithsingleorsystemic

rightventricles.Circulation,2002.105(10):p.1189-94.5. Gelow, J.M.,etal.,Organallocation inadultswithcongenitalheartdisease listed forheart

transplant: impactof ventricular assistdevices. JHeart LungTransplant, 2013.32(11):p.1059-64.

6. VanderPluym, C.J., et al., Outcomes following implantation of mechanical circulatorysupportinadultswithcongenitalheartdisease:AnanalysisoftheInteragencyRegistryforMechanically Assisted Circulatory Support (INTERMACS). J Heart Lung Transplant, 2018.37(1):p.89-99.

7. Farooqi, K.M., et al., 3D Printing to Guide Ventricular Assist Device Placement in AdultsWithCongenitalHeartDiseaseandHeartFailure.JACCHeartFail,2016.4(4):p.301-11.

8. Salis,S.,etal.,Cardiopulmonarybypassdurationisan independentpredictorofmorbidityandmortalityaftercardiacsurgery.JCardiothoracVascAnesth,2008.22(6):p.814-22.

9. Poirier, N.C., et al., Long-term results of left ventricular reconditioning and anatomiccorrectionforsystemicrightventriculardysfunctionafteratrialswitchprocedures.JThoracCardiovascSurg,2004.127(4):p.975-81.

10. VanPraagh,R.,Whatiscongenitallycorrectedtransposition?NEnglJMed,1970.282(19):p.1097-8.

11. Graham, T.P., Jr., et al., Long-termoutcome in congenitally corrected transposition of thegreatarteries:amulti-institutionalstudy.JAmCollCardiol,2000.36(1):p.255-61.


Recommended