Anaesthetic and peri-operative management for thrombectomy procedures in stroke patients

(1)

Guidelines

Anaesthetic and peri-operative management for thrombectomy procedures in stroke patients

^§

Herve´ Quintard

^a,

*, Vincent Degos

^b

, Mikael Mazighi

^c

, Je´roˆme Berge

^d

, Pierre Boussemart

^e

, Russel Chabanne

^f

, Samy Figueiredo

^g

, Thomas Geeraerts

^h

, Yoann Launey

ⁱ

,

Ludovic Meuret

ⁱ

, Jean-Marc Olivot

^j

, Julien Pottecher

^k

, Francesca Rapido

^l

,

Se´bastien Richard

^m

, Suzana Saleme

ⁿ

, Virginie Siguret-Depasse

^o

, Olivier Naggara

^p

, Hugues De Courson

^q

, Marc Garnier

^r

aDivisionofIntensiveCare,UniversityHospitalofGeneva,Geneva,Switzerland

bPitie´-Salpetrie`reHospital,DepartmentofAnesthesia,CriticalCareandPerioperativeCare,47-83,Boulevarddel’Hoˆpital,SorbonneUniversity,Groupe RechercheCliniqueBIOSFAST,47-83,Boulevarddel’Hopital,75013Paris,France

cDepartmentofNeurologyandStrokeCentre,Lariboisie`reHospital,AP-HP,ParisUniversity,SorbonneParisCite´,De´partementHospitalo-Universistaire Neurovasc,Paris,France

dDepartmentofNeuroradiology,CHUdeBordeaux,Bordeaux,France

eNeurosurgicalIntensiveCareDepartment,LilleUniversityHospital,France

fDepartmentofPerioperativeMedicine,UniversityHospitalofClermont-Ferrand,Clermont-FerrandCedex,France

gUniversite´ Paris-Saclay,AP-HP,Serviced’Anesthe´sie-Re´animationetMe´decinePe´riope´ratoire,HoˆpitaldeBiceˆtre,DMU12Anesthe´sie,Re´animation,Douleur, LeKremlin-Biceˆtre,France

hPoˆleAnesthe´sie-Re´animation,Inserm,UMR1214,ToulouseNeuroimagingCenter,ToNIC,Universite´ Toulouse,CHUdeToulouse,Toulouse,France

iCriticalCareUnit,DepartmentofAnaesthesia,CriticalCareandPerioperativeMedicine,UniversityHospitalofRennes,Rennes,France

jAcuteStrokeUnit,ClinicalInvestigationCenterandToulouseNeuroImagingCenter,ToulouseUniversityMedicalCenter,Toulouse,France

kAnaesthesiology,CriticalCareandPerioperativeMedicine,StrasbourgUniversityHospital-EA3072,FMTS,Strasbourg,France

lDepartmentofAnesthesiologyandCriticalCareMedicine,GuideChauliacHospital,MontpellierUniversityHospitalCenter,Montpellier,France

mUniversite´ deLorraine,CHRU-Nancy,ServicedeNeurologie,Unite´ Neurovasculaire,CIC-P1433,INSERMU1116,F-54000Nancy,France

nServicedeNeuroradiologie,CHUDupuytren,87000,Limoges,France

oServiced’He´matologieBiologique,HoˆpitalLariboisie`re,Universite´ ParisCite´,75010Paris,France,UMR-S-1140InnovationsThe´rapeutiquesenHe´mostase, Universite´ ParisCite´,75006Paris,France

pPediatricRadiologyDepartment,NeckerEnfantsMalades&GHUParis,Sainte-AnneHospital,InstitutdePsychiatrieetNeurosciencesdeParis(IPNP),UMR S1266,INSERM,Universite´ deParis,75015Paris,France

qDepartmentofAnaesthesiologyandCriticalCarePellegrin,BordeauxUniversityHospital,Universite´ deBordeaux,InstitutNationaldelaSante´ etdela RechercheMe´dicale,UMR1219,BordeauxPopulationHealthResearchCenter,CHUBordeaux,Bordeaux,France

rSorbonneUniversite´,GRC29,DMUDREAM,Serviced’Anesthe´sie-Re´animationetMe´decinePe´riope´ratoireRiveDroite,Paris,France

ARTICLE INFO

Articlehistory:

Availableonline1January2023

Keywords:

Thrombectomy Guidelines Stroke Management

ABSTRACT

Purpose: To provide recommendations for the anaesthetic and peri-operative management for thrombectomyprocedureinstrokepatients

Design:Aconsensuscommitteeof15expertsissuedfromtheFrenchSocietyofAnaesthesiaandIntensive CareMedicine(Socie´te´ Françaised’Anesthe´sieetReánimation,SFAR),theAssociationofFrench-language Neuro-Anaesthetists(AssociationdesNeuro-Anesthe´sistesReánimateursdeLangueFrancaise,ANARLF), theFrenchNeuro-VascularSociety(Socie´te´ FrancaisedeNeuro-Vasculaire,SFNV),theFrenchNeuro- Radiology Society (Socie´te´ Francaise de Neuro-Radiologie, SFNR) and the French StudyGroup on HaemostasisandThrombosis(GroupeFrançaisd’E´tudessurl’He´mostaseetlaThrombose,GFHT)was convened,underthesupervisionoftwoexpertcoordinatorsfromtheSFARandtheANARLF.Aformal conflict-of-interestpolicywasdevelopedattheoutsetoftheprocessandenforcedthroughout.The entireguidelineelaborationprocesswasconductedindependentlyofanyindustryfunding.Theauthors

§TextvalidatedbytheparleComite´ desRe´fe´rentielsCliniques(clinicalreferencecommittee)oftheSFARon16/05/2022,bytheConseild’Administration(boardofdirectors)of theSFARle29/06/2022,bythebureauoftheANARLFon29/06/2022,bytheConseild’Administration(boardofdirectors)oftheSFNVon07/10/2022andbytheConseil Scientiﬁque(scientiﬁccouncil)oftheSFNRon16/08/2022.

* Correspondingauthor.

E-mailaddress:Herve.quintard@hcuge.ch(H.Quintard).

ContentslistsavailableatScienceDirect

Anaesthesia Critical Care & Pain Medicine

j ou rna l hom e pa ge : w w w. e l s e v i e r. c om

https://doi.org/10.1016/j.accpm.2022.101188

2352-5568/^C2022TheAuthor(s).PublishedbyElsevierMassonSASonbehalfofSocie´te´ franc¸aised’anesthe´sieetdere´animation(Sfar).Thisisanopenaccessarticleunder theCCBYlicense(http://creativecommons.org/licenses/by/4.0/).

(2)

Introduction

Managementofischemiccerebrovascularaccident(ICVA)must be undertakenwith maximumurgency, and represents a major issueinpublichealth.Upuntil2015,ICVAtreatmentwasbasedon rapidrecanalizationoftheoccludedarterybymeansofintravenous thrombolysis. The advent of mechanical thrombectomy (MT) enlarged the therapeutic arsenal and modiﬁed management policies. MT is indicatedeither in association with intravenous thrombolysisorasasecondorthirdoption(followingfailureofIV thrombolysisoralone,incaseofcontraindicationtoIVthromboly- sis),withinsixhoursaftersymptomonsetinacuteICVApatients withocclusionofalargearterialtrunkintheanteriorcirculation, and possibly in the posterior circulation, which is visible on imagery.Consideringtheresultsofrecentrandomizedcontrolled studies,some indicationscanbeprolonged forasmanyas24h.

Vascularocclusionmustbediagnosedbymeansofanon-invasive ﬁrst-linemethod(angio-tomodensitometryormagneticresonance angiography)beforeconsideringatherapeuticphaseinvolvingMT.

AnydecisiontocarryoutMTmustbemadebyamultidisciplinary team includingatleastoneneurologist and/or aphysician with competenceinneurovascularpathologiesfromtheestablishment’s neurovascular care unit, and a doctor qualiﬁed to perform mechanical thrombectomy. The latest French health authority (HAS) guidelines (https://www.has-sante.fr/jcms/c_2757616/fr/

organisation-de-la-prise-en-charge-precoce-de-l-accident- vasculaire-cerebral-ischemique-aigu-par-thrombectomie- mecanique)alsoinsiston theneedforananaesthesiateamand, morespeciﬁcally,ananaesthetistwithexperienceinthemanage- mentofpatientstreatedbyinterventionalneuroradiology,aswell as a registered nurse anaesthetist (the French IADE). Patient eligibilityforMTmustbediscussedcollegially ina consultation involving the vascular neurologist, the interventional neuro- radiologist,andtheanaesthetistincharge.

ThecliniciansresponsibleforMTroutinelyuseseveralscores corresponding tothe differentstepsof management and treatment:

-TheNIHSSscore[1]:TheclinicalNIHSS(NIHStrokeScale)isa diagnosticscaleusedfollowingischemicorhaemorrhagiccerebro- vascularaccidentstomeasuretheintensityofneurologicalsignsin view ofappraisingtheirseverityandmonitoringtheirevolution.

TheNIHSS[1]isbasedonthecollectionof15neurologicalitems;it permitspreciseandrapidassessmentofthedeﬁcitsobservedandis closely associated with patient outcomes. Correlated with the

volumeofthecerebralinfarction,ithasaquantitativeaswellasa prognostic function. It is associated with high inter-observer reliability. An NIHSS score between 1 and 4 signiﬁes a minor CVA;between5and15,a moderateCVA;between15and20,a moderate/severeCVA;andover20points,asevereCVA.

-TheASPECTSscore[2]:This10-pointradiologicalscale(Alberta StrokeProgramEarlyCTScore)assessesischemicstrokesinthearea ofthemiddlecerebralartery(MCA)bymeansofCTwithoutinjection.

Inclinicalpractice,thescoredividestheMCAinto10sectors:3deep orsubcorticalregions,and7superﬁcialorcorticalregions.Foreach sector, the absence ofhypodensity yields1 point. A score 7 is predictive of a pejorative prognosis in terms of both residual handicapandriskofhaemorrhagictransformation.AnASPECTSscore of0correspondstohypodensityoftheentireMCAarea.

-TheTICIscore[3]:TheangiographicTICIscore(TreatmentIn Cerebral Ischemia Scale revisited) quantifies the degree of revascularization. TICI 3 corresponds to complete radiological success;TICI2ctoalmostcompletefillingofthevascularterritory, butwhichinsomepartsisslowerthannormal;TICI2btofilling coveringhalfoftherevascularizedterritory;TICI2atolessthan halfoftheterritory;TICI1topenetrationofthecontrastproduct withminimal perfusion(absence of recanalization);and, lastly, TICI0correspondstono perfusion/absenceofrevascularization.

TICI2c/3isconsideredasagoodthrombectomyoutcome.

- The modified Rankin score [4]: As a global and clinical assessmentofhandicap,themodifiedRankinscoreisdetermined in five minutes. As a six-level scale, it ranges from 0 for no symptomatall;(1)fornosignificantdisabilitydespitesymptoms (abletocarry out allusual duties and activities);(2) for slight disability(unabletocarryoutallpreviousactivities,butableto look after own affairs without assistance); (3) for moderate disability (requiring some help, but able to walk without assistance);(4)formoderatelyseveredisability(unabletowalk without assistance, and unable to attend to own bodily needs withoutassistance);(5)forseveredisability(bedridden,inconti- nentandrequiringconstantnursingcareandattention);to(6)for death.

Objectivesoftherecommendations

The objective of these recommendations is to produce a frameworktofacilitatedecision-makinginasituationofextreme urgency,facetoapatientrequiringmanagementforathrombec- tomy procedure. Thegroup strove to puttogethera minimum wererequiredtofollowtheprinciplesoftheGradingofRecommendationsAssessment,Development andEvaluation(GRADE)systemtoguidetheirassessmentofqualityofevidence.

Methods:Fourﬁeldsweredeﬁnedpriortotheliteraturesearch:(1)Peri-proceduralmanagement,(2) Prevention and management of secondarybrain injuries,(3) Management of antiplateletand anticoagulanttreatments,(4)Post-proceduralmanagementandorientationofthepatient.Questions wereformulatedusingthePICOformat(Population,Intervention,Comparison,andOutcomes)and updatedasneeded.Analysisoftheliteraturewasthenconductedandtherecommendationswere formulatedaccordingtotheGRADEmethodology.

Results: The SFAR/ANARLF/SFNV/SFNR/GFHT guideline panel drew up 18 recommendations regardinganaestheticmanagementofmechanicalthrombectomyprocedures.Duetoalackofdata intheliteratureallowingtoconcludewithhighcertaintyonrelevantclinicaloutcomes,theexperts decidedtoformulatetheseguidelinesas‘‘ProfessionalPracticeRecommendations’’(PPR)ratherthan

‘‘FormalizedExpertRecommendations’’.Aftertworoundsofratingandseveralamendments,astrong agreementwasreachedon100%oftherecommendations.Norecommendationcouldbeformulated fortwoquestions.

Conclusions: Strong agreement among experts was reached to provide a sizable number of recommendationsaimed at optimisinganaesthetic management forthrombectomy in patients sufferingfromstroke.

C2022TheAuthor(s).PublishedbyElsevierMassonSASonbehalfofSocie´te´ franc¸aised’anesthe´sieet dere´animation(Sfar).ThisisanopenaccessarticleundertheCCBYlicense(http://

creativecommons.org/licenses/by/4.0/).

(3)

numberofrecommendationstohighlightthemainpoints,which havebeengroupedintofourpredeﬁnedﬁelds:(1)peri-procedural management,(2)preventionandmanagementofsecondarybrain injuries, (3) management of antiplatelet and anticoagulant treatments,(4)post-proceduralmanagement,andorientationof the patient. The golden rules for good medical practice are considered well-known, and consequently excluded from the recommendations; pre-hospital management is likewise not considered. The targetedpublic islarge-scale, correspondingto all ofthehealthprofessionals (neurologists,radiologists, anaesthetists-intensivists,etc.)involvedinmanagement.

Generalorganization

Theserecommendationsresultfromworkbyagroupofexperts broughttogetherbytheSFARand theANARLF,in collaboration withtheSFNVandtheSFNR.Priortotheanalysis,eachexpertfilled out a declaration concerningpossiblecompeting interests.Asa first step, theorganizingcommittee definedtheobjectives, the methodology,thefield(s)ofapplication,andthequestionstobe addressedintherecommendations.Thesedifferentelementswere subsequentlymodifiedandvalidatedbytheexperts.

Tothegreatestpossibleextent,thequestionswereformulated in accordance with the PICO (Population – Intervention – Comparison–Outcome)format.Thepopulationforwhomthese recommendations are addressed is composed of ‘‘patients presenting with a cerebral artery occlusion and eligible for endovascular treatment’’ (this is not repeated in each of the recommendations).

Therecommendationﬁelds

For thepresent recommendations, the experts unanimously decidedtofocusonthefollowingﬁelds:

FIELD1–Peri-proceduralmanagement

FIELD 2 – Prevention and management of secondary brain injuriesFIELD3–Managementofantiplateletandanticoagulant treatments

FIELD4–Post-proceduralmanagementandorientationofthe patient.

Methodologyofthebibliographicresearchandformulationof therecommendations

Up until March 2022, extensive bibliographic research was carried out on the MEDLINE PubMed^TM and clinicaltrials.gov databases byatleasttwo expertsineachﬁeld ofapplicationin accordance with the Preferred Reporting Items for Systematic ReviewsandMeta-Analyses(PRISMA)methodologyforsystematic reviews.

Were included in the analysis: meta-analyses, randomized controlledtrials,non-randomizedprospectivetrials,retrospective cohorts, case series, and case reports conducted in patients presenting with cerebral artery occlusion and eligible for endovasculartreatment,publishedinEnglishorFrench.

Analysisoftheliteraturewasthencarriedoutinaccordance with the GRADE¹ (Grade of Recommendation Assessment, Development,andEvaluation)methodology.Theendpointswere preliminarilydeﬁnedasfollows:

- Primaryendpoint:neurologicalprognosisat3months(assessed bytheRankinscore)(importance8).

- Secondaryendpoints: short-termneurologicalmorbidity (im- portance7),successofthethrombectomy(assessedbytheTICI score)(importance6).

Given the very low number of available studies presenting sufﬁcient power with regard to the most substantive primary endpoint, it was decided before the recommendations were draftedtoadoptaProfessionalPracticeRecommendations(PPR) ratherthanaFormalizedExpertRecommendations(FER)format.

The GRADE¹ methodology was applied in the analysis of the literatureandinthedraftingofsummarytablesrecapitulatingthe dataintheliterature.Alevelofevidencewasdeterminedforeach ofthecitedbibliographicreferencesaccordingtothetypeofstudy.

Itcouldsubsequentlybere-evaluatedaccordingtothemethodo- logicalqualityofthestudy,consistencyoftheresultsbetweenthe differentstudies,thedirectorindirectnatureoftheevidence,and analysisofthecostandextentofbeneﬁt.Therecommendations werethenwrittenout,usingtheSFARterminologyforPPRs:‘‘the experts suggest todo’’ or ‘‘theexperts suggest not todo’’. The proposedrecommendationswerethenpresentedtotheexperts anddiscussed,onebyone.Thegoalwasnotnecessarilytoarriveat asingleandconvergentopiniononalltheproposals,butratherto distinguishpoints ofconvergence frompoints ofdivergence or indecision.

Results

Therecommendationﬁelds

DuringtheﬁrstPPRorganizingmeeting,theexpertsconsensu- allydecidedtoaddress15questionsdistributedinfourﬁelds.The followingquestionswerechosenforthecollectionandanalysisof theliterature:

FIELD1–Peri-proceduralmanagement Questions:

Inapatientpresentingwithcerebralarteryocclusionandwhois eligible for endovascular treatment, does local anaesthesia alone,comparedtogeneralanaesthesiaorsedation,leadtoan improvedneurologicalprognosisat3months?

Inapatientpresentingwithcerebralarteryocclusionandwhois eligibleforendovasculartreatment,doessedation,comparedto generalanaesthesia,leadtoanimprovedneurologicalprognosis at3months?

FIELD2–Preventionandmanagementofsecondarybrain injuries

Questions:

Inapatienthavingpresentedwithcerebralarteryocclusionand received endovascular treatment, is the use of target blood pressure(BP)levelsduringpost-recanalizationassociatedwith improvedneurologicalprognosisat3months?

Inapatienthavingpresentedwithcerebralarteryocclusionand received endovascular treatment, does the setting of target saturation as an objective lead to improved neurological prognosisat3months?

Inapatienthavingpresentedwithcerebralarteryocclusionand received endovascular treatment under general anaesthesia, does per-procedure monitoring of CO2 lead to improved neurologicalprognosisat3months?

Inapatienthavingpresentedwithcerebralarteryocclusionand received endovascular treatment under sedation, does per-

(4)

procedure monitoring of CO2 lead to improved neurological prognosisat3months?

Inapatienthavingpresentedwithcerebralarteryocclusionand receivedendovasculartreatment,doesper-proceduremonitor- ing of glycemia lead to improved neurological prognosis at 3months?

FIELD3–Managementofantiplateletandanticoagulant treatments

Questions:

Inapatienthavingpresentedwithcerebralarteryocclusionand receivedendovasculartreatmentafterhavingbeenpreliminari- ly treated by intravenous thrombolysis, does systemic per- procedureheparinizationleadtoimprovedneurologicalprog- nosisat3months?

Inapatienthavingpresentedwithcerebralarteryocclusionand receivedendovasculartreatmentwithouthavingbeenprelim- inarilytreatedbyintravenousthrombolysis,doessystemicper- procedureheparinizationleadtoimprovedneurologicalprog- nosisat3months?

Inapatienthavingpresentedwithcerebralarteryocclusionand received endovascular treatment, does per-procedure anti- aggregation lead to improved neurological prognosis at 3months?

Inapatienthavingpresentedwithcerebralarteryocclusionand receiving endovascular treatment necessitating emergency stenting, does anti-aggregation of blood platelets lead to improvedneurologicalprognosisat3months?

FIELD4–Post-proceduralmanagementandorientationofthe patient

Questions:

In a patienthavingundergonecerebral thrombectomy under generalanaesthesia,doesimplementationofanearlyneurolog- icalassessmentstrategy(endofsedation,earlyextubation)lead toimprovedmorbi-mortality?

In a patienthavingundergonecerebral thrombectomy under general anaesthesia, does implementation of an early scale- guided extubation strategy (VISAGE score, etc.), lead to improvedmorbi-mortality?

In a patient having undergone cerebral thrombectomy, does orientationtowardanadaptedcarestructure(criticalcarevs.

stroke unit) according to severity criteria lead to improved morbi-mortality?

In a patient having undergone cerebral thrombectomy, does extended stay in an expert center, usually according to orientationcriteria,leadtoimprovedmorbi-mortality?

Synthesisoftheresults

Followingasynthesisoftheexperts’workandapplicationof the GRADE¹ method, 18 recommendations were formalized, whereas,for2questions,theexpertsrefrainedfromissuanceofthe latter.Alloftherecommendationsweresubmittedtotheexpert group for assessment, using the GRADE¹ Grid method. After 2roundsofrating,astrongagreementwasreachedon100%.

ThepresentPPRsreplacetheprecedingSFARguidelinesforthe sameﬁeldsofapplication.TheSFARandtheANARLFstronglyurge

all the professionals involved in the management of patients treated by endovascular thrombectomy to comply with the recommendations putforward. However, when they are being effectivelyapplied,eachpractitioneriscalledupontoexercisehis ownjudgment,takingintofullaccounthisareaofexpertiseandthe speciﬁcitiesofhisestablishment,soastodecideonthemeansof interventionbestsuitedtothestateofthepatientofwhomheisin charge.

FIELD1.Peri-proceduralmanagement

Experts:Y. Launey (ANARLF) – S.Saleme (SFNR) – O.Naggara (SFNR)-L.Meuret(SFAR)

Question: In a patient presenting with cerebral artery occlusionandwhoiseligibleforendovasculartreatment,does localanaesthesiaalone, compared to generalanaesthesiaor sedation,leadtoimprovedneurologicalprognosisat3months?

R1.1.1 - The experts suggest that general anaesthesia with orotrachealintubation,carriedoutbyananaesthesiateam,be preferred to local anaesthesia alone, the objective being to improvetheneurologicalprognosisat3months,whenatleast oneofthefollowingsituationsispresent:

-intheeventofimpairedposteriorcirculation

-intheeventofdifﬁcultplannedradiologicalneuro-navigation -intheeventofNIHSSscore15

-intheeventofdiminishedmentalalertness -intheeventofrespiratoryfailure

-intheeventofpatientagitation -intheeventofvomiting

Expertopinion(Strongagreement)

R1.1.2 – With the exception of situations necessitating intubation (see above), the experts suggest not to prefer general anaesthesia to local anaesthesia monitored by an anaesthesiateam,theobjectivebeingtoachieveanimproved neurologicalprognosisat3months.

Expertopinions(Strongagreement)

Argumentation: The distinction between local anaesthesia (LA)andproceduralsedation(PS)wasrelativelyrecentlydrawnin studiesaimingtoassesstheimpactofanaesthesiastrategyonthe prognosis of patients undergoing treatment for acute cerebral arteryocclusion.Asanexample,post-hocanalysisoftheIMS-III trial reported higher morbi-mortality following endovascular treatment under general anaesthesia (GA), as compared to treatmentunder‘‘localanaesthesia’’,whichinthisstudywasan entitybringingtogetherallformsofanaesthesiawithoutintuba- tion(withorwithoutsedation)[5].

Forocclusionsoftheanteriorcirculation

Mostofthestudiesdealing withthedifferencesbetweenLA aloneandPSareobservationalorretrospective,andconsequently notmethodologicallyrobust[6,7].Eventhoughameta-analysisof randomized trials [8] in which GA and LA with and without sedation are compared, as well as two analyses of the Dutch prospectiveandobservationalthrombectomyregistry[9,10]have beenpublished,theresultsareattimescontradictoryandriddled with several potential biases (methodologies, deﬁnition of LA, missingdata). Forexample,thestudybyBenvegnu` etal.seems unfavourable to LA compared to PS insofar as the former is associatedwithlesssatisfactoryfunctionalneurologicalevolution, alowerrateofreperfusion,andhigher3-monthsmortality[11];

whereastheanalysesoftheMRCLEANregistry[9,10]foundthatLA

(5)

wasassociatedwitha betterfunctionalprognosisthanPS.Inan Italianprospectivecohortstudyinvolving4429patientstreatedby endovascularthrombectomy[12],GAwasassociatedwithaworse 3-monthsfunctionalprognosisthanLAorconscioussedation(CS).

Itbearsmentioningthatinthis study,theinitialmedianNHISS scorewaslowerintheLAgroupthaninthePSorGAgroup,and thatthereasonsforconversiontoGAwerenotreported(amajor sourceofbias).Infact,itisnotpossibletoprioritizeonetechnique for endovascular thrombectomy following an ischemic anterior circulation stroke. While awaiting the results of more robust randomizedcontrolledstudiesinwhichLAwouldbemoreclearly deﬁned with regard to the other anaesthesia techniques (NCT02677415), it seems reasonable to privilege optimized anaesthesiamanagementthroughanapproachadoptedtopatient characteristicsandreliantonlocalexpertise.

Forocclusionsoftheposteriorcirculation

Acute occlusion of posterior arterial circulation, which involvesthevertebralarteryorthebasilarartery,isresponsible formoreseverestrokeswithlessfavourableprognosisthanthose affectinganteriorcirculation.Intheeventofbasilarocclusion,it occasionsbrainstemdysfunction(ascendingreticularactivating system,longmotorandsensorypathways,mixedcranialnerves, autonomic nervoussystem).Endovascularrevascularizationby thrombectomyis realizablein thisindication,notwithstanding theabsenceofformalscientiﬁcvalidation.Thepathophysiology differsfromthatofanteriorcirculation,withamechanismmore atherosclerotic than embolic. The concepts of collaterality, ischemicpenumbra,timelapseandsubsequentpatientselection onradiologicalcriteriaarelikewisedifferentfromthosehaving to do with anterior circulation. Indications are frequently

‘‘compassionate’’, particularly in case of severe forms, with individual beneﬁtsoftenfoundonlyafter anextendedlapseof time. Procedures are technically more challenging than in anterior circulation,and correspondinglylengthier. The litera- tureassessinganaestheticmanagementofposteriorcirculation occlusionsislimitedanduniquelyobservational,withdatabases includingvery fewpatients (atmost298patientsina Spanish series) [13]. Deﬁnition of a ‘‘general anaesthesia’’ group is generally restricted to ‘‘intubated patient’’, and that of a

‘‘sedation/local anaesthesia’’ groupto‘‘non-intubatedpatient’’.

Sometrialsonlyincludebasilarocclusion,whileothersinvolve allocclusionoftheposteriorcirculation.Outofthe7analysed trials, 4pertaintoan Asianpopulationpresentingmuch more atherosclerotic aetiology than others and possibly markedly different healthcare organization, which may be a source of selection bias and external validity bias. An ongoing single- centerrandomizedcontrolledChinesestudy[14]isattemptingto demonstratetheequivalenceofthedifferentformsofanaesthe- sia,andeventhegreaterbeneﬁtsofPSandLAascomparedtoGA in the selected patients, who are among the least severely impaired.Whileawaitingtheresultsofthisstudyandofother randomized studies yet to come, and given the possibility of severeformswithcomaand/orrespiratorydistress,theexperts have positioned themselves in favour of GA with intubation duringendovascularthrombectomyproceduresforocclusionof theposteriorcirculation.

WhenGAiscontemplatedaccordingtotheabove-mentioned criteria, thepresenceto monitor theoperation ofa speciﬁcally experienced and available anaesthesia team is of paramount importance.

Question: In a patient presenting with cerebral artery occlusion, and who is eligible for endovascular treatment, does sedation, compared to general anaesthesia, lead to improvedneurologicalprognosisat3months?

R1.2–Withtheexceptionof situationsnecessitating orotrachealintubation(cf.R1.1.1),theexpertssuggestnottoprefer generalanaesthesiatoproceduralsedationbyananaesthesia team,theobjectivebeingtoimprovetheneurologicalprognosis at3months.

Expertopinions(Strongagreement) Argumentation:

Forocclusionsoftheanteriorcirculation:

The literature concerning anaesthesia management [general anaesthesia(GA)orproceduralsedation(PS)]forcerebralartery occlusionsoftheanteriorcirculationissizablebutofpoorquality.

Numeroussetsofobservationaldatawithmeta-analysescharac- terizedbyselectionbias(non-controlledanaestheticintervention, severe patients managed under GA...) exist. In these series, definitionof theGAand sedationgroupsis imprecise(patients treated with local anaesthesia alone in the sedation group, or patientssecondarilyintubatedduetocomplications,andanalysed intheGAgroup).Since 2015,observationalmeta-analyseshave foundsedationtobebeneficialintermsoffunctionalautonomyat 3 months (modified Rankin score ‘‘mRS’’ 2). However, only 3randomizedcontrolledstudiesdealingspecificallywithanaes- thesia management exist [15–17]. These studies were single- center,European,andincludedfewpatients.Inaddition,allexcept one[16]hadanintermediateprimaryendpoint(NIHSSatD1or infarctvolume in MRI). Whatis more, the patients, who were treatedduringthe8hfollowingsymptomonset,werepeculiarly selected,as some wereexcluded due toNIHSS<10, agitation, vomitingand/orlossofairwayprotective reflexes,or whenthe anaesthesia team was unavailable. About 40% of the patients eligible for a thrombectomy were not randomized. The teams managingthepatientswerespecializedinneuro-anaesthesiaand/

orneuro-intensivecare,andtheyfullyrespectedthetherapeutic objectivesdefinedinthestudyprotocol,particularlyintermsof blood pressure. In these studies, time to reperfusion did not increaseunderGA,notwithstandingaslightlengtheningoftimeto vascular access. While no significant difference was found regarding the primary endpoint, 2 of the studies found signs favouringGAforthefunctionalprognosisatD90.Ameta-analysis of these trials based on individual data [18] found GA to be beneficialintermsofautonomyatD90.Unfortunately,itspower remained limited; only 368 patients were included, and the confidenceintervalwaswide(mRS2:65/185PS(35.1%)vs.90/

183 GA (49.2%), OR=0.46, 95%CI (0.28 0.76), p= 0.003).The inhomogeneityofsedationwasalsoduetothetypesofanaesthetic moleculesusedandtothefactthattheratesofconversiontoGA observed during the studies ranged from 6 to 16%, and were associatedwithalessfavourablefunctionalprognosis[18].Lastly andmorerecently,2multicentredrandomizedcontrolledFrench trials,GASS[19]andAMETIS[20],whichcomparedGAtoPS,found no signiﬁcant difference concerning their respective primary endpoints(GASS:mRS2atD90:40%PSvs.36%GA,OR=0.91 (0.64–1.31),andAMETIS:compositecriteriaassociatingfunctional autonomy(mRS2)atD90andabsenceofmajorcomplicationsat D7:39.1%PSvs.33.3%GA,OR=1.18(0.86–1.61),p=0.80).

Forocclusionsoftheposteriorcirculation

In the absence of well-conducted studies on procedural sedation in thrombectomy for arterial occlusion of posterior cerebralcirculation,andforthesamereasonsasthoseputforward intheR1.1.1argumentation,generalanaesthesiawithorotracheal intubationshouldprobablybeprioritizedduringthrombectomy, particularlyintheeventofbasilaryarteryocclusion.

(6)

FIELD2.Preventionandmanagementofsecondarybrain injuries

Experts:T.Geeraerts(ANARLF)–M.Mazighi(SFNV)-S.Figueiredo (SFAR)–JMOlivot(SFNV)

Question:Inapatienthavingpresentedwithcerebralartery occlusions and received endovascular treatment, is use of targetbloodpressure(BP)levelsinpost-recanalizationassoci- atedwithimprovedneurologicalprognosisat3months?

R2.1.1 - The experts suggest that in the event of TICI <2b recanalization,post-proceduresystolicbloodpressureshould bemaintainedbetween130and180mmHg,theobjectivebeing toachieveanimprovedneurologicalprognosisat3months.

R2.1.2 - The experts suggest that in the event of TICI 2b recanalization,post-proceduresystolicbloodpressureshould bemaintainedbetween130et160mmHg,theobjectivebeing toachieveimprovedneurologicalprognosisat3months.

Argumentation: Blood pressure management has multiple therapeuticobjectives:toensureadequatecerebralperfusion;and toavoidhaemorrhagicphenomenaorthedevelopmentofcerebral oedemaorothermalignantcomplications.Iftargetsforadequate bloodpressureduringtheﬁrst24post-thrombectomyhoursare met,thepatient’sprognosiscouldbeimproved,andpostoperative complicationsreduced.

TheguidelinesputforwardbytheAHA/ASA(AmericanHeart Association/AmericanStrokeAssociation)aremainlybasedonthe literatureconcerningintravenousthrombolysis.Theyrecommend maintaining systolicblood pressure(SBP) at <180mmHg and diastolicbloodpressure(DBP)at<105mmHgafterrevasculari- zationTICI2bor3[21].TheSNACC(SocietyforNeurosciencein Anaesthesiology and Critical Care) recommends per-procedural SBPbetween140and180mmHg,andDBP<105mmHg[22].

Thedegreeofrecanalizationofthearteryisanimportantpoint, especially insofar as spontaneouslyloweredblood pressurehas beenobservedfollowingsuccessfulrecanalizationandnormalized intracranialhaemodynamics[23].

The one presently available work on the subject is theBP- TARGETstudy(BloodPressureTargetinAcuteStroketoReduce HaemorrhageAfterEndovascularTherapy)[24].Itisamulticenter randomized controlled prospective study includingsuccessfully recanalizedpatients(TICI2band3)inwhichanintensivestrategy (target SBP100-129mmHg)vs. a standard strategy (targetSBP 130 185mmHg)wasproposed.Theprimaryendpointconsisted inappearanceofhaemorrhageonCTat24 36hours.Allinall,the 158 randomized patients in the intensive arm had the same incidence of cerebral haemorrhage as the 160 patients in the standardarm,andacomparablefunctionalprognosisat3months.

That is why, according to the currently available evidence on haemorrhagicriskandfunctionalprognosisat3months,itwould beofnointeresttomaintainSBPtargetsat<130mmHg.

Numerous retrospective studies have reported congruent resultsregardingtheupperlimitofSBP.Theretrospectivestudy by Matusevicius et al., in which 2920 patients were included, concluded that SBP>160mmHg was associated with a poor prognosisforsatisfactorilyreperfusedpatients(TICI>2b)[25].In theretrospectivestudybyAnetal.,maximumSBPof155mmHg was determined to be the threshold value associated with increasedriskofintracranialhaemorrhage[26].Dingetal.reported a similarresult,asthresholdvalues of151mmHg(AUROC0.74 95%CI (0.66 – 0.82), p< 0.001) were associated with a poor

neurologicalprognosisandthresholdvaluesof155mmHg(AUROC 0.64 (0.55 – 0.73), p= 0.006) associated with intracranial haemorrhage[27].Lastly,intheretrospectiveobservationalstudy byGigliottietal.,SBP>180mmHgoccurringatleastonceoverthe courseofthe25hfollowingthrombectomywasassociatedwitha poormRSscoreatdischarge[28].

Bloodpressuretargetsmustbeindividualizedaccordingtothe characteristicsofagivenpatient.Underlyingchronichighblood pressure, for example,may necessitate speciﬁcpost-thrombec- tomytargets.In2studies,onlyinpatientswithahistoryofchronic highbloodpressuredidexcessivebloodpressurevariationsduring the24hafterthrombectomyrepresentariskfactorforintracere- bralhaemorrhage[25,29].

Question:Inapatienthavingpresentedwithcerebralartery occlusion and received endovascular treatment, does the settingoftargetsaturationasanobjectiveleadtoanimproved neurologicalprognosisat3months?

R2.2–TheexpertssuggestmaintainingpatientSpO2at95%

during and after surgery so as to avoid aggravating the neurologicalprognosisat3months.

Argumentation: As of today, no study has dealt with the inﬂuence on patients’ neurologicalprognoses of differentSpO₂ levelsduringmechanicalthrombectomyprocedures.

Whileseveralrandomizedcontrolledstudieshaveassessedthe interestofsystematicoxygenadministrationinpatientspresent- ing with ischemic CVA, none of them have shown this to be effectivewithregardtoneurologicalevolutionat3months[30–

32].Thatmuchsaid,noneofthestudieshavedealtspeciﬁcallywith patients having undergone thrombectomy, and the rate of recanalizationamongthesepatientshasnotbeenreported.

Only1studyhastestedtheeffectofnormobaricoxygentherapy in theframework of a thrombectomy procedure [33]. It wasa single-center study carried out on a population consisting in patientshavingbeenadmittedforischemic CVAoftheanterior circulationandhavingreceivedthrombectomywithrecanalization without general anaesthesia. The intervention consistedin the administrationsubsequenttorecanalizationof15L.min ¹ofO2

suppliedbyaVenturimaskfor6h.Thecontrolgroupreceivedlow- flowoxygentherapybynasalcannula(3L.min ¹)duringthesame 6hperiod.Intheinterventiongroup98patientswereincluded, and87inthecontrolgroup.Abeneficialeffectofoxygentherapyat 15L.min ¹regardingtheneurologicalprognosisat3monthswas reported [adjusted OR=2.20 (1.26–3.87)], as well as benefit concerningmortality[RR=0.35(0.13 0.93)].Takenalone,however,theseoutcomesdonotsufficetojustifyarecommendationfor systematichigh-flowsupplementaloxygenfollowingrecanaliza- tionbythrombectomyasotherstudiesareneededforconfirma- tion.

Question:Inapatienthavingpresentedwithcerebralartery occlusionandreceivingendovasculartreatmentundergeneral anaesthesia, does per-procedure monitoring of CO₂ lead to improvedneurologicalprognosisat3months?

R2.3 - The experts suggest that during thrombectomy pro- cedurescarriedout undergeneralanaesthesia,end-tidal CO2

shouldbemonitoredandmaintainedat35 40mmHg,soasto avoidworseningtheneurologicalprognosisat3months.

Argumentation: Cerebral autoregulation (modification of cerebralbloodflowinresponsetoamodificationofmeanblood pressure)isprofoundlyimpactedbyPaCO2level[34].Inapatient treatedbythrombectomyforischemicCVA,thevasoconstriction

(7)

induced by excessive hypocapnia may transform the ischemic penumbra zoneintoanirreversiblyinfarctedzone. Withthis in mind,anincrementaldiminution (1mmHg)ofPaCO2induces a 3.5% reduction of blood ﬂow in the middle cerebral artery [35]. Conversely,alveolarhypoventilation not onlyexposes the patienttoariskofhypoxemia,butalsoprovokescerebralarteriolar vasodilatation,whichcanleadtointracranialhypertensionandto aphenomenonof‘‘vascularsteal’’tothedetrimentofinsufﬁciently perfusedregions.Observationalstudieshavehighlightedsponta- neous hypocapnia during acute ischemic stroke [36,37], which maybeprovokedbysomaticandpsychologicalstressalongwith the perturbedrespiratory patternsassociated withthe cerebral insult [38]. It has also been demonstrated that cerebral autoregulation,neurovascularcouplingandvascularreactivitytoCO2

are modiﬁedintheischemicCVApatient[36].Theinﬂuenceof capniaontheprognosisofpatientshavingundergonethrombec- tomy hasbeenanalysed inonlytwoprospective cohortstudies [39,40], and a before-after study [41]. In the 2014 study by Takahashietal.[39]onpatientdatacollectedprospectivelyfrom 2007to2010andanalysedretrospectively,patientswithagood neurologicalprognosisat3months(mRS0–3)presentedhigher etCO2 measurementsat60 and 90minthanthose witha poor neurological prognosis (mRS 4–6) (respectively 35.2 vs.

32.2mmHg at 60min; OR=0.7695%CI (0.65 0.92),p= 0.03;

and34.9vs.31.9mmHgat90min;OR=0.7695%CI(0.61—0.93),p

=0.01).Publishedin2018,theworkbyAthiramanetal.[40]isa single-center, retrospective study based on data concerning 88 patients having undergone thrombectomy under general anaesthesia between 2010 and 2014. After adjustment for age and NIHSSscore,theauthorsobservedthatthepatients witha good neurological prognosis (mRS 0–2) presented a higher maximum etCO2 level than those with a poor neurological prognosis(mRS 3–6) (498vs.457mmHg;OR=1.1495%CI (1.02–1.28);p= 0.02).However,nodifferencewasfoundinmedian etCO2 level between the ‘‘good’’ and the ‘‘poor’’ neurological prognosis groups (36 [34–40] vs. 35 [33–37] mmHg; p= 0.09).

Published in2016, thework byMundiyanapurathet al. [41]is a before-afterstudyincludingpatientshavingundergonethrombec- tomy under generalanaesthesia. It was initiallycarriedout with retrospectivecollectionofdatafrom2008to2010on60patients,and then,afterimplementationofaprotocolimposingstrictetCO2control (between40and45mmHg)byretrospectivecollectionofdatafrom 2012on64patients.Inunivariateanalysisonly,theauthorsfounda statisticallysigniﬁcantassociationbetweenaprolongedtimelapse (>105min)ofetCO2between40and45mmHgandapoorprognosis (mRS3–6).Asaresult,whenetCO2ismeasuredinpatientsunder generalanaesthesiawithaninvasiveairwayapproach,valuesoutside of the physiological norms (<35 and >40mmHg) seem to be associatedwithanunfavourableneurologicalprognosis,leadingthe expertstosuggestetCO2targetvaluesbetween35and40mmHg.

Question:Inapatienthavingpresentedwithcerebralartery occlusionandreceivedendovasculartreatmentundersedation, doesper-proceduremonitoringleadtoimprovedneurological prognosisat3months?

R2.4–Theexpertssuggest,duringthrombectomyprocedures carriedoutundersedation,thatend-tidalCO2bemonitored,so as to verify the persistence of the patient’s spontaneous ventilation.

Argumentation:Asregardstherecommendationofsystematic monitoring ofetCO2 duringthrombectomyundersedation,it is essentially based on the contribution of this technique to the reductionofper-procedureepisodesofdesaturationandhypox- emia;thiswasillustratedin2meta-analyses[42,43].Overrecent

years,etCO2monitoringhasbeensubstantiallyimproved[44–46], anditnowensuresreliablesurveillanceofrespiratoryfrequency andearlydetectionofhypoventilationepisodesinspontaneously breathing patients. In these situations, however, non-invasive evaluation of PaCO2 by etCO2 is neither reliable nor precise;

moreover, it is influenced by supplemental oxygen flow rates, contaminationbyatmosphericair,andpreferentialventilationof deadspaceinpatientssufferingfromchronicrespiratorydisease [44,45,47].Inaprospectiveclinicalpilotstudy,every30minfor 2h,Lemurzeauxetal.comparedetCO2,transcutaneouspressurein CO2 (PtcCO2), and PaCO2 in 25 non-intubated patients in respiratory distress [47]. In this population, the correlation (R=0.62), mean bias (13.9mmHg) and limits of agreement (- 5.6–33.6mmHg) between the measurements provided by the 2 techniques did not suffice to render them interchangeable.

Conversely,thevaluesprovidedbymeasurementofPtcCO2were better correlated with PaCO2 (R=0.97), mean bias (1.73.9mmHg)andlimitsofagreement(-5.8–9mmHg),which were more compact and closer to those recommended by the American Association for Respiratory Care (1.967.5mmHg) [46]. Present-day limitations to widespread utilization of PtcCO2

technologystemfromitshighcost,itsunevenavailability,aneedfor regularcalibrationandanincompressibletimelapse(afewminutes) betweenitsinstallationonapatient’sskin,andcompletecalculation ofavalue[46].

Question:Inapatienthavingpresentedwithcerebralartery occlusion and received endovascular treatment, does per- proceduremonitoringofglycemialeadtoimprovedneurologi- calprognosisat3months?

R2.5 - The experts suggest that hyperglycaemia episodes be monitoredandtreated,whileneverthelessavoidinghypogly- caemia,whichmaybeinducedbythetreatment,soastoavoid worseningtheneurologicalprognosisat3months.

Argumentation: Hyperglycaemia on hospital admission is observed in 40–50% of the patients admitted due to ischemic cerebrovascularaccident(iCVA),whetherornottheyarediabetic [48].Numerousstudieshavedemonstratedthathyperglycaemiais associatedwithunfavourableneurologicalevolution,whateverthe meansofrevascularization (thrombolysisormechanical thrombectomy(TM):moreextendedinfarct, morenumeroushaemor- rhagic complications, less satisfactory functional recovery at 6months,higherdeathrateatD30[49–54].Fromapathophysio- logical standpoint,hyperglycaemia increaseslacticacidosis and cytotoxicoedema,decreasescerebralvaso-reactivityandcollateral circulation, and alters the haemato-encephalic (blood-brain) barrier,therebyincreasingtheriskofhaemorrhagictransforma- tion after revascularization. In light of these data, it has been hypothesizedthatstrictglycaemiccontrolcouldbefavourableto patientoutcomes.Tothisday,however,analysisofthepublished meta-analysesandrandomizedstudiesshowsthatstrictglycaemic control by meansof continuous intravenous infusionimproves neitherfunctional recoverynormortalityinpatientspresenting withiCVA,particularlythoseeligibleforMT.Quiteonthecontrary, it increasestherisk ofhypoglycaemia,which in this contextis probablyjustasdeleteriousashyperglycemia[48,55–57].

FIELD3.Managementofantiplateletandanticoagulant treatments

Experts:V.Seguret(GFHT)–J.Pottecher(SFAR)–JM.Olivot(SFNV) –M.Mazighi(SFNV)–S.Richard(SFNV)

Question:Inapatienthavingpresentedwithcerebralartery occlusion and receivedendovascular treatment after having

(8)

beenpreliminarilytreatedbyintravenousthrombolysis,does systemic per-procedure heparinization lead to improved neurologicalprognosisat3months?

R3.1 – The experts suggest not to carry out systemic per- procedureheparinizationofcerebralendovascularrevasculari- zation in patients having preliminarily received intravenous thrombolysis, the objective being to avoid aggravating the neurologicalprognosisat3months.

Argumentation: Inthecontextofcerebral revascularization, systemicper-procedureheparinizationisaimedatpreventing:(a) non-perfusion indistalmicrocirculation;(b) thrombosisaround cathetersandatthelevelofthegeneratedendotheliallesions;and (c)theformationdeneutrophilextracellulartraps(NETs).These different effects, which can contribute to improved distal revascularization(evaluatedbya TICIscoreequalorsuperiorto 2b) are counterbalanced by a theoretical risk of increasingly frequent intraparenchymal cerebral haemorrhage (ICH). More speciﬁcally,theeffectofsystemicheparinizationontheneurolog- icalprognosisat3months(mRSat90days)istobeevaluatedin termsof:(a)thequalityofcerebralrevascularizationand(b)the consequences of symptomatic intra-parenchymal haemorrhage (sIPH).

The oldest studies [58,59] and/or those with the smallest populationarethosethatreportabeneﬁcialeffect,withoddsratios higher than 5, of systemic heparinization in terms of mRS at 3months.Conversely,theretrospectivestudiesbringingtogethera largerpopulationhavereportedalackofeffect[60–63],orevenan unfavourable effect [64] on the mRS at 3 months. The one randomized multicenter prospective study with a factorial or crossed-treatmentdesign[65]utilizingaspirinaswellasheparin (at low or moderate dose) was prematurely halted due to indication of excess mortality [adjusted odds ratio (aOR) 5.85 95% CI (1.7–20.2)]in the groupof patients receiving moderate dosesofheparin(5000unitsasabolusfollowedby1250UI/hour for6h).

Concerningthequalityofcerebralrevascularizationattheend of the procedure (TICI 2b and 3), the results of systemic heparinization were either non-signiﬁcant [58–61,63,64] or unfavourable[62].

Lastly, as concernsanassociation ofsystemic per-procedure heparinizationwiththeriskofsIPH,theresultsoftheretrospective analysesofthelarge-scaleprospectiveregistriesareeithernon- signiﬁcant[58–62],orunfavourabletoheparin[63,64,66].Inthe 2mostrecentstudies,thereexistedeitherathresholdeffect[66]or aconcentration-dependenteffect[63],markedbyincreasedsIPH incidenceinassociationwiththeheparinconcentrationusedinthe rinsingliquid.Theaforementionedrandomizedprospectivestudy byvanderSteenetal.[65]foundahighlysigniﬁcantincreaseof sIPHinthegroupofheparin-treatedpatients(13%,acrossalldoses) comparedtonon-heparin-treatedpatients[7%;aOR=1.98(1.14–

3.46)].

Toconclude,inpatientshavingreceivedprimaryintravenous thrombolysisfollowedbyendovascularcerebralrevascularization, systemic heparinization seemedneithertoincreasethepropor- tions of favourable neurological prognosis nor to improve the quality ofrevascularization,especially inthemostrecenttrials.

Quiteonthecontrary,systemicintravenousheparinadministra- tion exposed patients to increased risk of apparently dose- dependent symptomatic intra-parenchymal hematoma. As a result, systemic heparinization is not recommended,and when it is necessitated for procedural reasons, small doses should probablybeused.

Question:Inapatienthavingpresentedwithcerebralartery occlusionandreceivedendovasculartreatmentwithouthaving beenpreliminarilytreatedbyintravenousthrombolysis,does systemic per-procedure heparinization lead to improved neurologicalprognosisat3months?

ABSENCEOFRECOMMENDATION– Atthistime,theavailable literaturedoesnotallowtomakearecommendationonthe possible interest of systemic heparinizationin patients pre- sentingwithcerebralarteryocclusiontreatedbyendovascular thrombectomywithouthavingbeen preliminarily treatedby intravenousthrombolysis.

Argumentation:Thedatapertainingtosystemichepariniza- tion inpatients treatedby thrombectomy alone (without prior thrombolysis) are few and far between. They concern either patients contraindicated for intravenous thrombolysis [67], or patients admitted subsequent to thetime frames provided for thrombolysisandincludedinaninterventionalstudycomparing medical treatment alone to medical treatment associated with interventionalrevascularization[68].

Ina retrospectiveanalysisoftheFrenchprospectiveregistry ETIS (6 centers), 751 patients, of whom 27% were receiving heparin,werecontraindicatedforthrombolysis[67].Inthisstudy, heparinadministrationwassigniﬁcantlyassociated withapoor neurological prognosis (aOR 1.58 (1.05–2.40); p=0.03). The qualityofcerebralrevascularization waseitherincreasedinthe heparin group (when TICI grades 2b and 3 are taken into consideration; aOR=1.62 (1.06–2.48); p= 0.03) or decreased (takingintoconsiderationonly completerevascularization:TICI grade 3; aOR = 0.68 (0.49 0.95); p= 0.02), whereas counter- intuitively, the frequency of haemorrhagic complications was reduced(OR=0.48(0.34 0.68);p<0.001).

Inaretrospectiveanalysisofthe107patientsincludedinthe DAWNstudyhavingreceivedcerebralrevascularization,30%were treatedbysystemicheparinization[68].Favourableneurological statusat3months(mRS0–2)wasobservedin37.5%ofthepatients treatedbyparheparinandin52.1%oftheuntreatedpatients,a non-signiﬁcantdifference (p= 0.17). Cerebralrevascularization (TICI 2b-3) wasachievedin 65.3% of thepatients treated with heparinandin75.3%ofthenon-treatedpatients(p= 0.35),and haemorrhagiccomplicationswerenotreported.

Amongthe162patientsintheMRCLEANMEDstudynothaving preliminarilyreceivedintravenousthrombolysisandincludedin the ‘‘systemic heparinization’’ arm, there was no signiﬁcant difference with regard to favourable neurological status at 3 months or in intra-cerebral haemorrhage incidence [65].This resultshowsalackofstatisticalpowerandcannotbeextrapolated tothegeneralpopulation.

Question: In a patient having presented with cerebral arterialocclusionsandreceivedendovasculartreatment,does per-procedureanti-aggregationleadtoimprovedneurological prognosisat3months?

R3.2–Intheabsenceofpreliminaryintravenousthrombolysis, the experts suggest not to systematically administer to all patientsananti-GPIIb/IIIa plateletaggregationinhibitoror a directthrombininhibitor;thistreatmentcannonethelessbe proposed in case of distal embolisms during mechanical thrombectomyor intheevent ofpersistentocclusion atthe end of the procedure, the objective being to improve the neurologicalprognosisat3months.

Argumentation:Utilizationofplateletaggregationinhibitors during mechanical thrombectomy (MT)may have as a goal to

(9)

prevent the migration of distal emboli and to favour optimal reperfusion. Platelet aggregation inhibitors are also used as second-line treatment for occlusions or stenoses that persist followingMT[69,70].

The overwhelming majority of studies in this context are designedtoassesstheuseoftirofiban(aninhibitorofthebinding of fibrinogentotheGPIIb/IIIareceptor,withahalf-lifeapproxi- mating 2h) [69–80], or of argatroban (a direct inhibitor of thrombin,whichalsoinhibitsplateletaggregation,andpossessesa half-life inferiortoone hour)[81].There existsonly1 studyin which none of patients included had undergone preliminary intravenous thrombolysis (IVT) [73]. In the other studies, a proportionrangingfrom24to88%ofthepatientshadundergone IVT prior to MT [72,80]. On this subject, there exists only 1randomizedtrial,whichincluded60patientsbygroup(treated or not treated by tirofiban), with univariate analyses alone as judgment criteria [78]. The other studies were prospective observationalorretrospective.Ameta-analysisgroupedtogether 844patientstreatedwithtirofibanduringtheMTprocedurevs.

1645whowerenot[79].

Asregardsfunctional prognosis,a gain in termsof indepen- denceat3months(modifiedRankinscore0–2)wasobservedin 5studiesforthepatientstreatedwithtirofiban,withthehighest odds ratio at 2.99 (1.01–8.85) [70]. The beneficial effect was confirmedintherandomizedstudybyZhanget al.[78],witha higher proportion of patients treated with tirofiban having achievedindependenceat3monthscomparedtothosewhowere not(61.7%vs.45%,p= 0.024);andlikewiseobservedinthemeta- analysis by Zhang et al.,who reported a significantassociation between tirofiban treatment and functional independence at 3months[OR=1.29(1.05–1.58)][79].

Asregardsqualityofreperfusion,optimalreperfusion(TICI2b- 3) was more frequentlyachieved in thetiroﬁban groupin the retrospectivestudybyKimetal.(86.4%vs.42.4%,p= 0.016)[70]

andintherandomizedstudybyZhangetal.(88.3%vs.66.7%,p= 0.036) [78],that said, inmultivariate analysisin theliterature, particularlyinthemeta-analysisbyZhangetal.[OR=1.32(0.97–

1.79)][79],nosigniﬁcantassociationwashighlighted.

AsregardsantiplateletsafetyduringtheMTprocedure,from 3.3–17.6% of the patients presented with symptomatic intra- parenchymal haemorrhage (sIPH) after utilization of tirofiban [78,80]and1%afterutilizationofargatroban[81].Only1outofthe 12studiesfoundmorefrequentsIPHoccurrenceinpatientstreated withtirofiban(14.6%vs.5.7%,p=0.027andOR=2.8(1.0–7.9),p= 0.049) [72]. This did not translate in any of the studies into significantlyincreasedmortality.

As regardslocal complicationsrelated tothe MTprocedure, only Kim et al.reported 2 casesof arterialdissection(1 inthe argatrobangroupvs.1inthenon-treatmentgroup),and3casesof arterialperforationintheargatrobangroup[81].

All in all, utilization of tiroﬁban and argatrobanduring MT seems favourable to a return to functional independence in patients treatedfor cerebralarteryocclusion.Asregardssafety, very few signiﬁcant increases have been reported in IPH occurrence, local complications inherent to the MT procedure, ormortality.Thatmuchsaid,theseconclusionshavebeendrawn fromstudiesinsomeofwhich,thelevelofevidenceislow;whatis more,thepopulations havebeenheterogeneous,particularlyas regardspreliminaryIVTtreatment.

Sincethepublicationoftheserelativelysmall-scalestudies,a multicenter Chinese randomized placebo-controlled study has dealtwiththeeffectsofTiroﬁbanin948ischemicstrokepatients less than 24hbefore treatment, withocclusion of theinternal carotidarteryortheproximalpartofthemiddlecerebralartery and withan NIHSSscore30,andno beneﬁtwasobservedin terms of functional prognosis at D90 (mRS=3 (1–4) in the

2groups,aOR=1.08(0.86–1.36),p= 0.50)[82].Whiletherewas nodifferencebetweenthe2groupsregardingincidenceofmajor intracranial haemorrhage [9.7% vs 6.4%, aOR 1.56 (0.97–2.56)], overall incidence of radiological intracranial haemorrhage was higherinthetiroﬁbangroup(34.9%vs.28.0%,aOR=1.40(1.06–

1.86),p= 0.02)[82].

R3.3–Theexpertssuggestnottoadministeraspirinduringa cerebralendovascularrevascularizationprocedure,regardless ofwhetherthepatienthasundergonepreliminaryintravenous thrombolysis, the objective being to avoid the risk of symptomaticintraparenchymalhaemorrhage.

Argumentation:Publishedin2022,therandomizedmulticen- ter prospective MR CLEAN MED study by Van der Steenet al.

reportedadoubledriskofsymptomaticintracerebralhaemorrhage inpatientstreatedwithaspirinduringMT(300mgadministered intravenouslyfollowingarterialpuncture),ascomparedtothose whohadnotreceivedaspirin[14%vs.7%;aOR=1.95(1.13–3.35)]

[65]. This result was confirmed in the overall population (628 patients having received or not received preliminary thrombolysis). As regards the population limited to patients preliminarilytreatedbyIVT(n=466),eventhoughtheoddsratio remainedunfavourabletoaspirin(aOR=1.72(0.93–3.18)forrisk ofICH),itdidnotreachstatisticalsignificance.Bythesametoken, inthesub-groupofpatientsnothavingpreliminarilyundergone IVT,aspirinadministrationdidnotleadtoanimprovedneurologi- calprognosisat3monthsandwasassociatedwithasubstantial increase of ICH [OR=3.01(0.88–10.26)] that was nonetheless non-significant, possibly because the sub-group analyses were lackinginpower.

Question:Inapatienthavingpresentedwithcerebralartery occlusionandreceivedendovasculartreatmentnecessitating emergencystenting,doesanti-aggregationofblood platelets leadtoimprovedneurologicalprognosisat3months?

R3.4.1–Theexpertssuggest(singleordouble)anti-aggregation ofbloodplateletsduringstenting,theobjectivesbeingtoavoid thrombosis and to improve the neurological prognosis at 3months.

R3.4.2 – The experts suggest to initiate anti-aggregation of blood platelets postoperatively, only after having ruled out cerebralhaemorrhagebyCT-scanduringthe24hfollowingthe procedure, the objective being to avoid aggravating the neurologicalprognosis.

Argumentation:Thereexistsnorandomizedprospectivestudy evaluatingthepotentialgainofanti-aggregationmedicationwhen anMTprocedureincludesstenting.Allofthepublishedstudiesare retrospective,andtheiranti-aggregationindicationsandstrategies wereheterogeneous(tandemproceduresthroughastenosisofthe carotidartery,etc...)[66,83–86].Theaforementionedstudieshave nevercomparedstentingalonetostentingwithanti-aggregation.

Given the speciﬁcities of thestudy methodologies, it is not possible to draw conclusions on potential gain in terms of functionalindependenceduetotheuseofanti-plateletmedication duringstenting fortandem occlusions.At3 months,functional independence(mRS0–2)hasbeenachievedfor37.4%–50%ofthe patientswithstentingandantiplateletdrugs[85,86],notablywith unassistedwalkingfor53.1–62%(mRS0–3)[83,85].DaRosetal.

foundasigniﬁcantassociationbetweenuseofdualanti-aggrega-

(10)

tionafterstenting,andfunctionalindependence(OR=6.03(1.87–

19.4),p= 0.003)[66].

Theobservedpercentagesforoptimalreperfusion(TICI2b-3)are excellent: 90% and 61.1% of patients with stenting and anti- aggregation in the studies by Lee et al. and Da Ros et al. [66]

respectively,and100%ofthepatientsreportedbyElhoranyetal.[86].

Concerning safety, between 3.2 and 27.7% of the patients treated with anti-aggregation medication presented with ICH duringstentingoverthecourseoftheMTprocedure[66,84].The study by Neubergeret al. foundno associationbetween useof tiroﬁban,orofdualanti-aggregation,andICHoccurrence[85].Da Rosetal.observedanegativeassociationbetweenpost-MTdual anti-aggregationandICHoccurrence(OR=0.21(0.06–0.78),p= 0.02)[66].Bu¨ckeetal.reportedhigherICHfrequencyinapatient populationunderdualanti-aggregationwhenaspirinisassociated with ticagrelor (7.1%), in comparison with an associationwith clopidogrel (no ICH observed) [84]. As was the case with the differentanti-aggregationstrategiesoutlinedabove,ofwhichthe power was limited in some cases by small populations, the observeddifferencesdidnotentailexcessmortality.

Tosummarize,theobservationalstudiesintheliteraturedonot permitassessmentoffunctionalprognosisgainsthatwouldbedue toplateletanti-aggregationduring astentingprocedure carried outintheoverallframeworkofmechanicalthrombectomy(MT).

Thatmuchsaid,whileplateletanti-aggregationstentingdoesnot seem to be associated with ICH occurrence, there exists no comparatoronstentingwithoutanticoagulation.Asaresult,given thestateofcurrentpracticeandwhileawaitingstudieswithhigher levels of evidence,theexperts agree to recommendthe useof platelet aggregation inhibitors in stenting during the MT procedure; however, they are currently unable to indicate an optimaltherapeuticscheme.

And to conclude, recent data reportingan increased risk of intracranial haemorrhage following preliminary IVT [65] have persuaded the experts to recommend the introduction of the antiplatelettreatmentonlyafterpost-MTCT-scan.Thisapproach appearssafe,whileawaitingthepublicationofmorerobuststudies onthesubject.

FIELD4–Post-proceduralmanagementandorientationofthe patient

Experts: R. Chabane(ANARLF) – S.Richard (SFNV)– F.Rapido (SFAR)–T.Geeraerts(ANARLF)

Question: In apatient having undergonecerebral throm- bectomyundergeneralanaesthesia,doesimplementationofan earlyneurologicalassessmentstrategy(endofsedation,early extubation)leadtoimprovedmorbi-mortality?

R4.1.1–Theexpertssuggesttostopadministeringanaesthesia drugsassoonasthethrombectomyprocedureisover,exceptin the event of ventilatory failure or complications suggesting intracranial hypertension orstatus epilepticus,theobjective beingtoavoidworseningmorbi-mortality.

R4.1.2 – The experts suggest extubation of the patient immediatelyafterthethrombectomyprocedure,providedthat theusualprerequisitesarepresentandthatstateofalertnessis satisfactory (visualcomponent of theGlasgow score3) to avoid worsening morbi-mortality. However, responding to verbalcommandsisnotnecessary.Swallowingandcoughing mustalsobeassessedforocclusionsoftheposteriorcircula- tion.

Argumentation:Intheliteraturetheredoesnotexistanystudy havingcompared,followingthrombectomywithgeneralanaesthesia, earlyevaluationtodelayedevaluation,duringwhichsedationand mechanicalventilationwouldbepursuedpostoperatively.Itiswidely recognizedthatone ofthetheoretical drawbacksofmanagement undergeneralanaesthesiaisthatitdoesnotallowclinicalmonitoring ofapatientasregardsnotonlypossibleneurologicalimprovement, but also – and especially – early detection of neurological complications[87].Asitdoesnotappearlogicaltoundulyprolong generalanaesthesiaandtherebyforgoreliableclinicalexamination,it is recommended,when the usualprerequisites to extubationare fulﬁlled,toextubatethepatientasrapidlyaspossibleaftersurgery(cf.

SFAR guidelines ‘‘Difﬁcult intubation and extubation in adult anaesthesia’’[88]).Moreover, prolonged mechanicalventilationis associatedwithcomplicationssuchasnosocomialpneumonia,and timeto extubationis associatedwithincreasedmortality inCVA patientsadmittedtointensivecare[89].Andyet,delayedextubation remains quite frequent. In the SIESTAstudy, delayed extubation (deﬁned as absence of extubation 2 hours after the end of MT) occurredin49%ofthepatientsinthegeneralaneasthesiagroup,and duetomalignantCVA,cerebralhaemorrhageorpneumonia,9.2%of patientswerestillintubatedatH24[15].

In a single-center Austrian observational trial including 441 patients treated by MT of the anterior circulation under general anaesthesia (not including malignant CVA), median mechanical ventilation was 3 [1–530] hours [90]. Fifty-eight percentofpatientscouldbeextubatedwithin6h, 34%between 6 and 24h, and 8% after 24h. Early extubation (< 6h) was associatedwithamorefavourableneurologicalprognosis(mRS2) at3monthsthanextubationcarriedoutbetween6and24hand,a fortiori,morethan24h.Timebeforeextubationof6–24hoursvs.<

6hwasassociatedwithadmissionduringthe‘‘permanenceofcare’’

period,afactorconducivetounduedelaysinextubation.Inaddition tobeingassociatedwiththe prognosisat3 months,mechanical ventilationdurationwassigniﬁcantlyassociatedwithoccurrenceof pneumoniaandincreasedlengthofhospitalstay[90].

We must nonetheless remember that due to frequent comorbidities,thepopulationofpatientsundergoingendovascular thrombectomyisparticularlyatriskoffailedventilatorweaning and/or failed extubation. More specifically, in a large-scale American cohort composed of patients having undergone MT, heart failure, diabetes and chronic respiratory disease were associated with prolonged mechanical ventilation (> 96h) [91]. There also exists a risk of Ear-Nose-Throat bleeding in patients having undergone thrombolysis, as well as possible occurrenceofanoedemarelatedtort-PA(incidencefrom0.9to 7.9%) [92]. Lastly, more failures occur in iCVA of the posterior circulation,representingamajorindependentmarkerofextuba- tion failure [93]. In fact, these CVAs often have respiratory repercussionsduetoalteredstatesofconsciousnessanddysfunc- tionof themixednervesof thecerebral trunk, impairedupper airwayprotectivereflexesand,attimes,disruptionofrespiratory controlmechanismson accountofbulbo-pontinelesion[94].In addition,revascularizationbyMTisquitedifficulttoachievein procedures concerning the posterior circulation rather than anterior circulation, frequently leading to irreversible ischemic lesions and heightened clinical severity. Some patients can nonethelessbeeasilyextubated,providedthatthebasilarartery (vertebralartery,posteriorcerebralartery)isnotdamaged,orif revascularization has been carried out in the absence of fully constitutedcerebralinfarction.

Question:In a patient havingundergonecerebral throm- bectomyundergeneralanaesthesia,doesimplementationofan earlyscale-guidedextubationstrategy(VISAGEscore,etc.),lead toimprovedmorbi-mortality?