shorterdurationofcirculatoryarrestandoutcomeisuncertain,butexamination
ofthedatafromtheBostontrialrevealsarangeofoutcomesacrosstherangeof
circulatoryarrestintervals,indicatingamultiplicityofdeterminantsofoutcome,
includingindividualbiologicsusceptibility.303Thesafedurationofhypothermic
circulatoryarrestisrelatedtotherateofuseofoxygenfromavailablestoresand
canbepredictedonthecombinationofhemoglobin,temperature,pH,andtime.
ThetransitiontoanaerobicmetabolismduringDHCAcanbedetectedbya
reductionintherateofcerebraldesaturationasmeasuredbyNIRS.195,199,252The
distributionoftensionsofoxygeninthebrainduringhypothermiccirculatory
arrestishigherandapoptoticregulatorsarelowerwithapH-statstrategyfor
cooling,predictingalongersafetimeforhypothermiccirculatoryarrest.200,304
Inanindividual,theuseofNIRScanguidethesafedurationofhypothermic
circulatoryarrestbylimitingthetimeoflowcerebraloxygenation.203,204With
optimalcoolingto18°Candacirculatoryarresttimeoflessthan30minutes,
cerebralinjuryisunlikely.302Moderateandprofoundhypothermiamayinitiate
protectiveproapoptoticmechanismsthatoff-setthedeleteriouseffectsof
sublethalischemictimes.305Duringlongerperiodsofcirculatoryarrest,
intermittentreperfusionatintervalsof15to30minuteshasbeendemonstrated
tomaintaincytoarchitecture,cerebraldistributionofoxygen,andindicatorsof
excitotoxicityinanimals.306–308Inhumans,avoidanceofhypoxicischemia
usingastrategyofintermittentreperfusionduringDHCAcanbearational
strategyforprolonged,complexrepairs.202
SelectiveAntegradeCerebralPerfusion
Becauseofthevariabilityintimenecessarytocompleterepairandthelimited
durationofDHCA,strategieshavebeenemployedtomaintaincontinuous
deliveryofoxygentothebrain.Selectiveantitradecerebralperfusionviathe
brachiocephalicarteryhasbecomewidelyused.309–314Theoptimalstrategyfor
thistechniqueremainspoorlycharacterizedbecausemeasurementsofcerebral
bloodflowarenotreadilyavailableandautoregulationmaybealteredduring
coldCPBandselectiveperfusion.315pH-statcoolingtoatargetof20°Cto26°C,
asinanticipatedcirculatoryarrest,followedbydirectperfusionofthe
brachiocephalicartery,iscommonpractice.Atourinstitutionwecurrentlycool
to18°Cwithalpha-statbloodgasmanagementandperfuseviaacannulatedgraft
totheinnominateartery.Flowratesof10to80mL/kgperminutehavebeen
described.203Flowratesoflessthan30mL/kgperminute,however,arenot
likelytoprovideadequatecerebralbloodflowtoopenallcapillarybeds.316,317
Moreover,theincreasedaffinityofhemoglobinathypothermiamaylimit
availabilityofoxygenduringperfusion,partiallyoff-settingtheanticipated
reductioninmetabolism.Limiteddatafromadultsundergoingelectivearch
reconstructionsuggestthatmoderatehypothermiamaybeanalternativestrategy
withacceptableneurologicoutcome.318,319However,theoptimaltemperatureis
notknown,andishighlyrelatedtoflowstrategiesandthelikelihoodof
interruption.Noncerebralbedsremainpoorlyperfusedduringselectivecerebral
perfusiontechniques,andthusarestillsusceptibletoischemicinjury.Forthese
reasons,atargetofdeephypothermiashouldbemaintainedinanyoperationin
whichprolongedselectiveperfusionoraperiodofcirculatoryarrestmightbe
necessary.320,321Techniquestomonitoradequacyofcerebralflowduringthis
techniqueincludetranscranialDopplerandNIRS.203
Experimentalmodelsofcontinuouscerebralperfusioncomparedto
hypothermiccirculatoryarrestshowimprovedcerebraloxygenation,198,322better
postperfusionhemodynamics,323reducedapoptosis288withlessischemicinjury,
andimprovedoutcomes.321,324,325Arecentcomparisonofcerebralperfusion
versushypothermiccirculatoryarrestinneonatesundergoingreconstructionof
theaorticarchshowednodifferenceinneurodevelopmentaloutcomeat1year.
Thetechniqueforcerebralperfusioninthisstudy,however,usedarateofflow
ofonly5to20mL/minandmeasuresofcerebraloxygenationwerenot
reported.326Becausetheratesofflowinthosereceivingcerebralperfusionwere
notlikelytoresultinadequatedeliveryofoxygen,theresultsarenotsurprising,
showingnodifferencebetweencompleteischemiaandinadequateperfusion.
Outcomesutilizinganalternativeapproachofhigh-flowperfusionof50to70
mL/kgperminuteshownoevidenceofischemicinjuryonpostoperative
magneticresonanceimagingandgoodneurodevelopmentaloutcomes.327–329
Becausebypassexposesthebodytoasignificantinflammatorystimulus,there
maybearelativedisadvantageofprolongedcerebralperfusioncomparedto
circulatoryarrestofshorterduration.330–332Atpresent,thereexistsnodirect
comparisonbetweencerebralperfusionatratesofflowwithmeasuredadequate
cerebraloxygenationandDHCA.TheoptimalpHstrategyforcontinuous
selectiveperfusionisalsodebatableandconflicting,withsomeevidence
favoringanalpha-statapproach,333andotherevidencefavoringapH-stat
approach.200,304Becauseoftheinherentriskofprolongedischemiaandthe
unpredictabledelayedeffectsofhypothermiccerebralperfusionandcirculatory
arrestonpostoperativeflowofbloodtothebrain,wefavorstrategiesthatrely
bothonmeasurementandmaintenanceofcerebraloxygenationthroughoutthe
operativeperiod.334
PharmacologicandMechanicalAdjuncts
Corticosteroids
Pretreatmentwithcorticosteroidsiswidelyusedwithbroadbutconflicting
evidenceforalterationofoutcome.Pretreatmentinadultsreducespostoperative
levelsoftumornecrosisfactor-α,interleukin-6,theincidenceofatrial
fibrillation,andmarkersofmyocardialischemia.335–338Evidenceexistsforboth
exacerbationandameliorationofhypoxicandischemiccerebraldamage.339–342
Itwasshownthattwodosesof30mg/kgmethylprednisolonemayamelioratethe
inflammation-relateddelayedreflowandcerebralmetabolismafterhypothermic
circulatoryarrest.343Themembrane-stabilizingeffectmayreduceexcitatory
neurotoxicity341andperivascularedemamaybereduced,344butnecroticcell
deathappearedtobeunaffectedandapoptoticcelldeathmaybeincreased.345,346
CorticosteroidsarecommonlyusedinCPBpumpprimes;however,arecent
studyoutofChildren'sHospitalofPhiladelphiarevealedalackofclinical
benefitaswellaspotentialcontributionstoincidentsofwoundinfectionsandas
aresult,theyhavediscontinuedtheuseofcorticosteroidsintheirpumpprime.347
Arecentanalysisoflinkeddatabasesalsofoundnoevidenceforimproved
outcomeswithcorticosteroidadministration.348
α-AdrenergicBlockade
Thedistributionofcardiacoutputisstronglyinfluencedbythesympathetic
nervoussystem,mainlythroughα-adrenergicmechanisms.Althoughdeep
anestheticstrategiescanaltertheneurohumoralstressresponsetosurgery,274
evidenceexistsforhighlevelsofsympatheticresponseduringcardiacsurgery
regardlessoftheanestheticregimen.349Wehaveshownanimprovementin
whole-bodyeconomyofoxygenusingphenoxybenzaminethatpermitsand
necessitatesastrategyofbypassathighratesofflow.350,351Inthepresenceof
milrinone,α-adrenergicblockadeismoreeffectivethannitrovasodilatorsin
improvingflowbothon352andoffbypass.316,353,354Althoughoutcomesseem