Regional anaesthesia and antithrombotic agents: recommendations

17 Pages·2010·0.23 MB·English

GUIDELINES
Regional anaesthesia and antithrombotic agents:
recommendations of the European Society
of Anaesthesiology
WiebkeGogarten,ErikVandermeulen,HugoVanAken,SibylleKozek,JuanV.LlauandCharlesM.Samama
BackgroundandobjectivesPerformingneuraxialanaesthesia catheters canbeplaced thenight before surgery.Regional
inpatientsreceivingantithromboticdrugsiscontroversialdueto anaesthesia inpatients receivingfull anticoagulation treatment
theincreased risk ofspinal epiduralhaematoma. Strict continues to becontraindicated. Catheter manipulation and
adherence tothe recommended time intervals betweenthe removal carrysimilarrisksto insertion andthesame criteria
administration ofanticoagulants, neuraxial blockadeandthe should apply. Appropriate neurological monitoring is essential
removal ofcatheters isthought toimprove patient safety and duringthepostoperativerecoveryperiodandfollowingcatheter
reducetheriskofhaematoma.Appropriateguidelineshavebeen removal. The final decision toperform regional anaesthesia in
prepared bya number ofnational societies of patientsreceivingdrugsthataffecthaemostasishastobetaken
anaesthesiologists,buttheydonothaveuniversalacceptance. after careful assessment ofindividual risksandbenefits.
The introductionof newanticoagulants together with recent Eur JAnaesthesiol 2010;27:999–1015
reportsofstentthrombosis inpatientswith perioperative Publishedonline1October2010
cessationofantiplateletdrugshaveconsiderablybroadenedthe Keywords:apixaban,aspirin,bleeding,cilostazol,clopidogrel,dabigatran,
issueandmaderevisionnecessary.Toovercomedeficienciesin epidural,fondaparinux,haematoma,heparin,hirudin,idrabiotaparinux,
low-molecular-weightheparin,prasugrel,regionalanaesthesia,
content andapplicability, theEuropean Society of
rivaroxaban,spinal,ticagrelor
Anaesthesiologyhastakentheinitiativetoprovidecurrentand
Abbreviations:ACCP,AmericanCollegeofChestPhysicians;ACT,
comprehensive guidelines forthe continent as a whole.
ActivatedClottingTime;ADP,AdenosineDiphosphate;aPTT,Activated
MethodsExtensive reviewofthe literature. PartialThromboplastinTime;ASRA,AmericanSocietyofRegional
Anesthesia;DVT,DeepVenousThrombosis;ECT,EcarinClottingTime;ESA,
Resultsand conclusions Inorder tominimise bleeding
EuropeanSocietyofAnaesthesiology;ESC,EuropeanSocietyof
complications during regionalanaesthetic techniques, care Cardiology;FDA,USFoodandDrugAdministration;HIT,Heparin-Induced
Thrombocytopenia;INR,InternationalNormalisedRatio;LMWH,Low-
shouldbetaken to avoidtraumatic puncture. Ifa bloody tap
Molecular-WeightHeparin;NSAIDS,Non-SteroidalAnti-InflammatoryDrugs;
occurswhen intraoperative anticoagulation is planned, PCC,ProthrombinComplexConcentrates;PF4,PlateletFactor4;PDE,
postponing surgery should beconsidered. Alternatively, Phosphodiesterase;PT,ProthrombinTime;SSRI,selectiveserotoninuptake
inhibitor;UFH,UnfractionatedHeparin;VTE,VenousThromboembolism
Background
The editors would like to thank Alain Borgeat
Thefirstnationalrecommendationsonneuraxialanaes-
(Switzerland),ElisabethGaertner(France,ESAScien-
thesia and antithrombotic drugs were published by
tific subcommittee 8), Daniela Filipescu (Romania,
the German Society for Anaesthesiology and Intensive
ESA Scientific subcommittee 6), Klaus Go¨rlinger Care in 19971 followed by the American Society of
(Germany, ESA Scientific subcommittee 6), Markus Regional Anesthesia (ASRA) in 1998,2 and Belgian
W.Hollmann(Netherlands,ESAScientificsubcommit- anaesthesiologists in 2000.3 Since then new anti-
tee8),SusanMallett(UK,ESAScientificsubcommit-
coagulant agents have been introduced and more
tee 6), and Andrew F. Smith (UK, Chair of the ESA
information regarding the risk of neuraxial regional
Guidelines Committee) for their help in critically
anaesthesiawithconcurrentanticoagulationisavailable.
appraisingthescientificcontentofthisarticle.Potential
Inaddition,ithasbecomeclearthatifantiplateletdrugs
conflicts of interest of the assessors are stated in the
are withheld after coronary stent implantation, the
Acknowledgementssectionattheendofthispaper.
risk of adverse perioperative cardiovascular events
increases. Newer and more comprehensive recommen-
dations are warranted.
TheEuropeanSocietyofAnaesthesiology(ESA)working-
From the Department of Anaesthesiology, Intensive Care and Pain Therapy,
Harlaching Hospital, Municipal Hospitals of Munich, Munich, Germany (WG), party on Neuraxial Anaesthesia and Anticoagulants,
Department of Anaesthesiology, University Hospitals, Katholieke Universiteit,
composed of academic physicians experienced in this
Leuven,Leuven,Belgium(EV),DepartmentofAnesthesiologyandIntensiveCare
Medicine, University Hospital, Munster, Germany (WG, HVA), Department of topic,hastheaimofprovidingguidelinestoassistEuro-
Anaesthesiology, Intensive Care and Pain Management, Vienna, Medical
pean anaesthesiologists in their daily clinical practice.
University,Vienna,Austria(SK),DepartmentofAnaesthesiologyandPostsurgical
CriticalCare,HospitalClinicUniversiari,Valencia,Spain(JVL)andDepartmentof Withregardtoperioperativecardiacrisk,thesamework-
AnaesthesiologyandIntensiveCare,Hotel-DieuUniversityHospital,Paris(CMS),
ing party has graded the level of recommendations
France
and the level of evidence for the European Society of
CorrespondencetoWiebkeGogarten,MD,PhD,DepartmentofAnaesthesiol-
ogy,IntensiveCareandPainTherapy,HarlachingHospital,MunicipalHospitalsof Cardiology(ESC)-ESAguidelines,usingthedefinitions
Munich,Munich,Germany of the Committee for Practice Guidelines of the ESC
Tel:[phone];fax:[phone];
e-mail:[email protected] (Table 1).4
0265-0215(cid:2)2010CopyrightEuropeanSocietyofAnaesthesiology DOI:10.1097/EJA.0b013e32833f6f6f
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
1000 Gogartenetal.
Table1 Classesofrecommendationsandlevelsofevidence
Classesofrecommendations
ClassI Evidenceand/orgeneralagreementthatagiventreatmentorprocedureisbeneficial
ClassII Conflictingevidenceand/ordivergenceofopinionabouttheusefulness/efficacyofthetreatmentorprocedure
ClassIIa Weightofevidence/opinioninfavourofusefulness/efficacy
ClassIIb Usefulness/efficacyislesswellestablishedbyevidence/opinion
ClassIII Evidenceorgeneralagreementthatthetreatmentorprocedureisnotusefuloreffectiveandinsomecasesmaybeharmful
Levelofevidence
LevelA Dataderivedfrommultiplerandomisedclinicaltrialsormeta-analyses
LevelB Dataderivedfromasinglerandomisedclinicaltrialorlargenon-randomisedstudies
LevelC Consensusofopinionoftheexpertsand/orsmallstudies,retrospectivestudies,registries
Due to the rarity of spinal epidural haematoma, recom- Collation of these cases in the USA allowed the risk of
mendationsregardingneuraxialregionalanaestheticpro- spinal epidural haematoma during concurrent adminis-
cedures with concurrent thromboprophylaxis, are not trationoflow-molecular-weightheparins(LMWHs)tobe
based on prospective randomised studies, but rather on calculated at 1:40800 for spinal anaesthesia, 1:6600 for
casereportsandexpertopinion.Thelatterisbasedmainly single-shotepiduralanaesthesia,and1:3100forepidural
on knowledge of the pharmacokinetics of the individual catheter anaesthesia.7 What appears to be a relatively
agents concerned. A rule of thumb adopted by most high incidence of bleeding was attributed to the twice
nationalsocietiesputsthetimeintervalbetweencessation daily administration of LMWH, and the lack of recom-
of medication and neuraxial blockade at two times the mendationsatthattimeregardingtimeintervalsbetween
elimination half-life of the drug. This approach has neuraxialpunctureorcatheterremovalandthrombopro-
recentlybeenrecommendedbyothers.5 phylaxis.TheresponseintheUSAhasbeentointroduce
recommendations that are stricter than those in place in
The decision for or against regional anaesthesia always
Europe,proposingavoidance ofLMWHtheentiretime
requiresacarefulrisk–benefitanalysis,notinganyhistory
epiduralcathetersareinplace.Thisdoesnotadequately
of bleeding, followed by a physical examination looking
take into account the increased perioperative risk of
for signs of increased bleeding tendency, for example
venousthromboembolism(VTE).14InEurope,thewide-
petechiae or haematoma6 (Class I, level A). Laboratory
spreadadoptionofasingledailydoseofenoxaparin40mg
tests, if indicated at all, should be appropriate to the
produced a lower incidence of complications. A retro-
individual (Class I, level A). Routine laboratory investi-
spectiveanalysisinSwedenfoundtheriskwas1:156000
gations do not always detect impaired coagulation. The
after spinal anaesthesia and 1:18000 after epidural
perioperativecessationofanticoagulantdrugstoimprove
anaesthesia,withbleedingoccurringmorerarelyinobste-
the safety of neuraxial block needs to be critically eval-
trics (1:200000) than in female orthopaedic patients
uated. An alternative anaesthetic technique should be
undergoing knee arthroplasty (1:3600).8 Risk factors
used if it is judged that the administration of the antic-
forspinalhaematomaafterneuraxialregionalanaesthesia
oagulant must not be interrupted (Class IIa, level C).
were identified as lack of guidelines, administration of
Finally,thepresentguidelinesarenotintendedtobypass
antithrombotic agents, female sex, and difficult punctu-
clinical judgment. When the anaesthesiologist decides
res. Subsequent reports from various countries indicate
not to comply with these guidelines, the reasons should
thatspinalepiduralhaematomaafterneuraxialblockade
be noted in the patient’s chart.
occursin1:2700to1:19505patients,9–12withonereport
indicating that haematoma may be more common after
Risk of spinal epidural haematoma
lumbar (1:1341) compared to thoracic epidural anaes-
Spinal epidural haematoma often occur spontaneously,
thesia (0:10199).11
without any temporal relationship with neuraxial anaes-
thesia. The absolute risk of spinal bleeding during
The association of spinal haematoma with concurrent
concurrent thromboprophylaxis is not known, and
administration of antithrombotic drugs is not a new
following neuraxial blockade, events are too rare to
finding. Vandermeulen et al. reported as early as 199415
study in a randomised clinical trial. Relatively recent
that68%ofpatientswithspinalepiduralhaematomahad
caseseriessuggestthattheriskofspinalepiduralhaemor-
receivedanticoagulants,and20%hadthrombocytopenia
rhage is possibly much higher than was previously
or drug-induced platelet dysfunction, something Wulf
thought.7–13
confirmed later in 1996.16 Despite its frequent use,
After enoxaparin 30mg twice daily was introduced for aspirinisrarelymentionedasariskfactor,butfibrinolytic
thromboprophylaxis in the United States an alarming agents and Bechterew disease are implicated. In an
number of spinal epidural haematoma, some with per- analysis of 79 haematoma cases, coagulation disorders
manent paraplegia, were reported,7 triggering a warning were found in 72% of patients,17 and other risk factors
from the US Food and Drug Administration (FDA). included bloody or traumatic punctures and anomalous
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
EuropeanSocietyofAnaesthesiologyrecommendations 1001
anatomy, for example spina bifida, and Bechterew’s approach is poor. A single study including 1472 hip
disease. The risk of haemorrhage is lowest in spinal replacements, in whom dalteparin given 2h preopera-
anaesthesia, which employs fine needles, and highest tively was compared to dalteparin given 4h postopera-
in epidural catheter anaesthesia, which requires the lar- tively,21 found no difference in the incidence of VTE,
gest needle gauges available. Nearly half of all cases and patients in the preoperative group required signifi-
of bleeding occur during the removal of an epidural cantlymoretransfusions.Ameta-analysisofpreoperative
catheter,15andthisproceduremustberegardedascritical versus postoperative studies shows that LMWH given
as catheter insertion. 12h preoperatively does not reduce the risk of VTE
comparedtoapostoperativeregimen.22Themostrecent
SpinalepiduralhaematomaarenotrestrictedtoLMWH,
German guidelines on thromboprophylaxis, emanating
butcanoccurwithanyagentthatinterfereswithhaemo-
from a number of different specialties, and also the
stasis.Thetraditionalcoagulationscreenislargelyunaf-
ACCP, refer to preoperative administration only as an
fectedbyantithromboticagentsandisnothelpfulinthe
option, and not as a requirement.20,23 As antithrombotic
assessmentofbleedingrisk.Ifrecommendedanticoagu-
drugsincreasetheriskofspinalepiduralhaematomaafter
lantdosesandtimeintervalsbetweenadministrationand
neuraxial blockade, a postoperative start may be advan-
blockade are observed, the neuraxial procedure should
tageous, especially in patients also receiving aspirin
coincide with the lowest anticoagulant blood level. The
(Class IIb, level B).
smallertheamountprescribed,andthelongerthedelay
betweenadministrationandblockade,thelowertherisk
Antithrombotic drugs
of haemorrhage. Accordingly, recommended time inter-
Heparins
vals are based on the pharmacology of the individual
agents concerned. Assessing risk becomes more Unfractionated heparins
difficult when antithrombotic agents are used in combi- Thromboprophylaxis with low-dose heparin does not
nation, for example LMWH and aspirin, necessitating lead to an increased risk of bleeding after neuraxial
greater caution. blockade, provided that a time interval of 4–6h is
observed between heparin administration and puncture
In recent years, despite adhering to the recommen-
orcathetermanipulationandwithdrawal.Furtheradmin-
dations, spinal epidural haematoma have still occurred.
istration of low-dose heparin should be delayed for at
Thepatientsconcernedtendedtobeolderandsuffering
least1haftertheblock(ClassIIb,levelC).Coagulation
fromrenalimpairment,markingthisasanimportantrisk
studies are not required during prophylaxis with unfrac-
factor.18,19Mostdrugsusedforthromboprophylaxis,with
tionated heparin (UFH), except when treatment has
the exception of argatroban are eliminated by the renal
lasted 5 days or more, when a platelet count is needed
route, and will accumulate in those with renal impair-
to exclude heparin-induced thrombocytopenia (HIT)
ment,somethingthatoftengoesundetectedineveryday
(Class I, level B) (Tables 2 and 3).
practice.Therecommendedtimeintervalsthereforeonly
applytopatientswithnormalrenalfunction,andinthose WhereasdosesofUFHusedforvenousthromboprophy-
knowntohavereducedrenalfunction,eitherdoseadjust- laxisarerelativelysafe,anincreasedriskofbleedingdoes
ment or longer time intervals are required. occurattherapeuticdoses,whenpunctureandremovalof
acatheterarecontraindicated(ClassIII,levelC).If,after
TheAmericanCollegeofChestPhysicians(ACCP)inits
careful consideration, a neuraxial blockade or catheter
2008 recommendations state that appropriate patient
removal is planned, intravenous heparin administration
selection and caution is required when neuraxial block-
shouldbeinterruptedatleast4hearlierandtheactivated
adeisperformedinthepresenceofantithromboticdrugs.
partial thromboplastin time (aPTT) and/or anti-Xa
Theyadvisethatcautions applicable toneuraxialblock-
activity should have normalised before the procedure
ade should also be applied to deep peripheral nerve
(Class IIa, level C). UFHs have equal anti-IIa and
blocks,andifabloodypunctureoccurs,thromboprophy-
anti-Xa activity, and some laboratories might find it
laxis should be delayed.20
simpler to use a single test for both UFH and LMWH.
Timing of thromboprophylaxis Neuraxial blockade and intraoperative heparinisation
In most European countries thromboprophylaxis begins Intraoperative heparinisation does not necessarily
preoperatively; the exception to this is neurosurgery, represent a contraindication to neuraxial blockade. Rao
wherein it is started postoperatively. The reason for andEl-Etr24reportedin1981thatinpatientsundergoing
preoperative prescribing is the belief that thrombus for- vascular surgery, the risk of haemorrhage after epidural
mation occurs intraoperatively and that patients should anaesthesia and subsequent heparinisation is not
be protected during this period. In order to reduce increased if heparin is delayed for 1h after insertion
bleeding and to enable neuraxial blockade, LMWH and the activated clotting time (ACT) is maintained at
are usually administered the night before as opposed twicethebaselinevalue.Thiscontrastswithaveryhigh
to the morning of surgery. Scientific support for this 2% incidence of paraplegia when patients taking aspirin
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
1002 Gogartenetal.
Table2 Recommendedtimeintervalsbeforeandafterneuraxialpunctureorcatheterremovala
Timebeforepuncture/catheter Timeafterpuncture/catheter
manipulationorremoval manipulationorremoval Laboratorytests
Unfractionatedheparins(forprophylaxis, 4-6h 1h Plateletsduringtreatmentformore
(cid:2)15000IUperday) than5days
Unfractionatedheparins(fortreatment) i.v.4–6h 1h aPTT,ACT,platelets
s.c.8–12h 1h
Low-molecular-weightheparins(forprophylaxisb) 12h 4h Plateletsduringtreatmentformorethan5days
Low-molecular-weightheparins(fortreatment) 24h 4h Plateletsduringtreatmentformorethan5days
Fondaparinux(forprophylaxis,2.5mgperday) 36–42h 6–12h (anti-Xa,standardisedforspecificagent)
Rivaroxaban(forprophylaxis,10mgq.d.) 22–26h 4–6h (PT,standardisedforspecificagent)
Apixaban(forprophylaxis,2.5mgb.i.d.) 26–30h 4–6h ?
Dabigatran(forprophylaxis,150–220mg) Contraindicatedaccording 6h ?
tothemanufacturer
Coumarins INR(cid:2)1.4 Aftercatheterremoval INR
Hirudins(lepirudin,desirudin) 8–10h 2–4h aPTT,ECT
Argatrobanc 4h 2h aPTT,ECT,ACT
Acetylsalicylicacid None None
Clopidogrel 7days Aftercatheterremoval
Ticlopidine 10days Aftercatheterremoval
Prasugrel 7–10days 6haftercatheterremoval
Ticagrelor 5days 6haftercatheterremoval
Cilostazolc 42h 5haftercatheterremoval
NSAIDs None None
ACT,activatedclottingtime;aPTT,activatedpartialthromboplastintime;b.i.d.,twicedaily;ECT,ecarinclottingtime;INR,internationalnormalisedratio;IU,internationalunit;
i.v.,intravenously;NSAIDs,non-steroidalanti-inflammatorydrugs;s.c.,subcutaneously;q.d.,daily. aAlltimeintervalsrefertopatientswithnormalrenalfunction.bMaximum
prophylacticdosagesforlow-molecular-weightheparinsarelistedinTable3.cProlongedtimeintervalinpatientswithhepaticinsufficiency.
receivedintravenousheparinlessthan1hafterdiagnostic Manipulationorremovalofepiduralcathetersshouldbe
lumbarpuncture.25Theincreasedbleedingriskattached carried out at least 4h after the end of heparin admin-
to the coadministration of aspirin and heparin was con- istration and with normal aPTT, ACT, and anti-Xa
firmed by a Canadian research group. They calculated activity (Class IIa, level C).
that even if a 1h time interval between the neuraxial
If a bloody puncture occurs in patients in whom intrao-
block and subsequent intravenous heparinisation was
perative heparinisation is planned, it is recommended
observed,theriskofspinalepiduralhaematomainthose
that low-dose anticoagulation (5000IU heparin) should
taking aspirin was increased to 1:8500 after epidural
be avoided for 1–2h and that full intraoperative hepar-
anaesthesia and 1:12000 after spinal anaesthesia.26 In
inisationshouldbeavoidedfor6–12h,withtheoperation
the American Society of Anesthesiology closed claims
beingpostponedtothenextday,ifnecessary(ClassIIa,
analysis, spinal epidural haematoma occurred most fre-
levelC).Alternatively,toavoiddelays,epiduralcatheter
quentlyinvascularsurgicalpatients,suggestingthatthis
placementcanbecarriedouttheeveningbeforesurgery
particular patient group is at increased risk.27
(Class IIb, level C).
Table3 Doserecommendationsforvenousthromboembolism Cardiacsurgery
prophylaxisinhigh-riskpatients
Evidenceforthoracicepiduralanalgesiainpatientsunder-
Generic Max.prophylacticdoseperday going cardiac surgery is still equivocal with most studies
showing an improvement in pulmonary function,
Unfractionated Heparin(3(cid:3)5000IUoraPTTinnormal
heparin referencerange) improvedanalgesiaandlessarrhythmia,butnoreduction
Certoparin 1(cid:3)3000anti-XaUs.c. in length of ICU stay, time to discharge, myocardial
Dalteparin 1(cid:3)5000anti-XaUs.c. infarction or mortality.28,29 The accepted benefits need
Enoxaparin 1(cid:3)40mgs.c.
Nadroparin 2850anti-XaU(0.3ml)orweight-adjusted, tobecarefullyweighedagainstthepotentiallycatastrophic
max.5700anti-XaUs.c.(0.6ml) outcomeofhighthoracicspinalepiduralhaematoma.Ho
Reviparin 1(cid:3)1750anti-XaIUs.c.
Tinzaparin 1(cid:3)4500anti-XaUs.c. etal.30estimatedthemaximumprobabilityofhaematoma
Fondaparinux 1(cid:3)2.5mgs.c. formationinpatientsundergoingcardiacsurgerywithfull
Danaparoid 2(cid:3)750IUs.c.
heparinisation to be 1:1500 with epidural techniques. If
Desirudin 2(cid:3)15mgs.c.
Rivaroxaban 1(cid:3)10mgp.o. full anticoagulation during cardiopulmonary bypass is
Apixaban 2(cid:3)2.5mgp.o. planned,someguidelinesadvocateperformingtheblock
Dabigatran 1(cid:3)220mg(firstdose110mg)
thedaybeforesurgery,orfollowingatraumaticpuncture,
1(cid:3)150mgp.o.intheelderlypatient>75years
(firstdose75mg) that surgery is delayed.14,31,32 In many cardiac surgical
centres, aspirin and clopidogrel are administered in con-
aPTT, activated partial thromboplastin time; max., maximum; p.o., orally; s.c.,
subcutaneously. junction with therapeutic postoperative heparinisation,
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
EuropeanSocietyofAnaesthesiologyrecommendations 1003
increasing the risk associated with catheter removal. If catheter removal thereforedoesnotinterferewith neur-
patientsarenotdirectlyextubatedintheoperatingroom, axial blockade (Class IIa, level C).
close neurological surveillance is often delayed. Unde-
If thromboprophylaxis with LMWH is prescribed in a
tected cord compression reduces the chance of early
twice-dailyschedule(e.g.enoxaparin2(cid:3)30mgperday),
haematomaevacuation,andthehopeoffullneurological
compared to a once daily regimen, the risk of epidural
recovery.
haematoma may be increased because the trough levels
Analgesiawithasingledoseofspinalopioidmaydecrease ofanti-Xaactivityarehigher.42Inthissituation,onedose
the risk of neuraxial bleeding compared to a catheter ofLMWHshouldbeomittedcreatinga24htimeinterval
technique, but needs to be performed directly before before catheter removal and thesubsequent dose (Class
surgery and not in advance. Theabsence of local anaes- IIb, level C).
theticinfusionremovestheabilitytoadjustthequalityof
Similarly, when therapeutic doses of LMWH are being
analgesia and significantly limits the benefits.29
administered once or twice daily, catheter placement or
Becauseneuraxialblockadeincardiacsurgeryconfersno removalshouldalsobedelayedforatleast24hafterthe
majoreffectonmorbidityandmortalityandhassignificant last dose (Class IIa, level B). Whether a 24-h interval is
risks,itisarguablewhetherspinalandepiduraltechniques acceptable in relation to the risk of VTE needs to be
arejustifiedatallandperhapsshouldbeabandonedinthis considered on an individual basis. When the risk of
particularpatientgroup33(ClassIIb,levelC). thrombosis is high, for instance, with mitral or double
mechanical valve replacement, one should refrain from
neuraxial blockade and continue the administration of
Low-molecular-weight heparins
LMWH (Class III, level B). Following spinal/epidural
LMWHs are used for both prophylaxis and treatment
puncture, or after removal of a spinal/epidural catheter,
of VTE.34 Coagulation variables ACT and aPTT can
thenextdoseofLMWHshouldbedelayedforatleast4h
remainunaffectedbyLMWHandmonitoringisunhelp-
(Class IIa, level C).
ful(ClassI,levelA).Instead,theanticoagulanteffectof
LMWH is easily assessed by measuring plasma anti-
Anti-Xa agents
factor Xa activity (Class I, level A). The advantages of
Fondaparinux
LMWH lie in their high bioavailability (approximately
Fondaparinux is a synthetic indirect Xa inhibitor with
100%) after subcutaneous administration and their
potent anticoagulant activity. It can be monitored using
longerhalf-lifeof4–7h,whichtogethermakeonce-daily
anti-factor Xa activity. Platelet aggregation is not
administration feasible. For thromboprophylaxis, they
affected. Although antibodies against the platelet factor
are the treatment of choice, being superior to UFH in
4 (PF4)/ heparin complex(cid:4) that is responsible for HIT,
high-risk groups such as hip or knee replacement and
may form during its administration, in the absence of
traumapatients,butwithoutincreasingtheriskofbleed-
heparintheydonotbindwelltoPF4.Todate,thereare
ing.35 Although individual LMWH differ with regard
only two reports of a possible association of HIT with
to molecular weight and pharmacokinetics, and have
fondaparinux.43,44 Despite these reports, platelet count
been approved for different indications, their clinical
monitoring is not recommended and fondaparinux has
efficacyissimilar.ComparedtoUFH,thereisa10-fold
been successfully used for the treatment of HIT.45 The
reduction in the risk of HIT.36 Nevertheless, they are
ACCP currently suggests fondaparinux as an alternative
contraindicated in HIT due to the 90% level of cross-
anticoagulant for thromboprophylaxis in patients with a
reactivity.
history of HIT (grade 2C).46 Individual studies and a
AftersubcutaneousadministrationofLMWH,maximum meta-analysis have shown that it is superior to LMWH
efficacy levels are reached in approximately 3–4h, and in prevention of asymptomatic VTE following hip or
theterminaleliminationhalf-lifeinpatientswithnormal kneereplacementsurgeryandhipfractures.47Incontrast
renalfunctionis4–6h37,38(ClassI,levelA).Inthosewith to the European practice of starting VTE prophylaxis
severe renal insufficiency, anti-Xa activity reaches a preoperatively, fondaparinux is initiated at least 6–8h
higher maximum level and the elimination half-life aftertheendoftheoperationtoavoidcomplicationsfrom
can increase by up to 16h. In contrast to UFH, LMWH surgical bleeding (Class I, level A). The recommended
show a higher degree of fibrinolytic activity and less prophylactic dose is 2.5mg. Higher doses will increase
platelet interaction.39 This is reflected in greater throm- the rate of complications from haemorrhage without
bus regression in the treatment of deep venous throm- reducing the rate of VTE and are only approved for
bosis (DVT).40 To avoid bleeding complications, there therapeutic anticoagulation. Due to its long half-life of
should be a time interval of at least 12h between sub- 18h, fondaparinux is administered once daily and may
cutaneous administration of LMWH at prophylactic accumulate significantly in patients with impaired renal
doses and neuraxial blockade or removal of an epidural function(ClassI,levelA).Eveninpatientswithnormal
catheter15,41 (Class IIa, level C). Administration of renal function, a stable plateau is achieved only after
LMWH the evening before surgery or the night before 2–3 days.48 The manufacturer recommends reducing
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
1004 Gogartenetal.
the dose of fondaparinux to 1.5mg per day in patients avidin is very short (2min), in preliminary studies a
with moderate renal insufficiency (creatinine clearance rebound effect on anticoagulation was not seen during
20–50mlmin(cid:5)1)anditiscontraindicatedinpatientswith 5 days of observation. Nevertheless, a rebound effect
severe renal insufficiency (creatinine clearance below maybeexpectedwithredistributionofidrabiotaparinux
20mlmin(cid:5)1). from tissue sites. The concept is of great interest and
studies continue, but caution is necessary as, in theory,
Plateletfactor4 avidin carries the risk of an allergic reaction. Idrabiota-
PF4isasmallprocoagulantcytokinethatisreleasedfromthe parinuxhasnotbeendevelopedforperioperativethrom-
a-granules in platelets during aggregation. It binds to boprophylaxis,andtherearenodataregardingneuraxial
heparin with highaffinity,andthe PF4/heparin complexis anaesthesia, which is contraindicated pending further
the source of the antigen in HIT. Autoantibodies to PF4 studies (Class III, level C).
have been found in a thrombotic state similar to HIT, but
without administration ofheparin.
Rivaroxaban (Xarelto)
Rivaroxaban acts through inhibition of factor Xa and is
This agent has advantages in neuraxial regional anaes-
currently approved for the prevention of VTE in hip
thesia,becausecoagulationisnotaffectedatthetimeof
and knee replacement surgery. Rivaroxaban is elimi-
initiation of the block. However, rising plasma levels
nated through the kidney (33% active drug) and the
during the initial days of treatment and accumulation
liver, making accumulation less likely than with antic-
inpatientswithrenalinsufficiencyneedtobetakeninto
oagulants exclusively eliminated through the kidney.
account during removal of an epidural catheter. The
Rivaroxaban prolongs the prothrombin time (PT) in a
EXPERT study with a total of 5387 patients included
dose-dependent manner, but until further data are
1428 undergoing regional anaesthesia procedures, and a
available monitoring with PT is not recommended.
single dose of fondaparinux was omitted the evening
Similar to most other anticoagulant drugs, no specific
beforecatheterremoval.49Thisprovidedatimeinterval
antidote is available. The first 10mg dose of rivarox-
of36hbeforecatheterremovaland12hbetweencatheter
abanisadministered6–8haftersurgery.Incomparison
removalandthenextdoseoffondaparinux.Omittingone
with enoxaparin (1(cid:3)40mg), rivaroxaban was superior
doseoffondaparinuxdidnotincreasetheriskofVTEbut
in the prevention asymptomatic and symptomatic
did contribute to the safety of neuraxial blockade. No
VTE.52 A subsequent study has also shown improved
cases of spinal epidural haematoma occurred, but its
efficacyinthetreatmentofDVTcomparedtoheparins
incidence is so low that this study, like most others,
and vitamin K antagonists.53 Although initial studies
lacked sufficient power to make firm conclusions in
with rivaroxaban showed no increased bleeding risk
this regard.
compared to enoxaparin, an analysis performed by the
Neuraxial regional anaesthesia should not be performed FDA did identify an increased risk for non-major
when therapeutic doses of fondaparinux (5–10mg per clinical bleeding events.54 The plasma half-life of
day)areemployedduetothesubstantialriskofaccumu- rivaroxaban is 5–9h and is not significantly prolonged
lation (Class III, level C). in patients with moderate renal impairment, but
according to the manufacturer it is prolonged to 11–
Idrabiotaparinux 13h in the elderly. A time interval of 22–26h between
Like fondaparinux, idraparinux is a pentasaccharide the last dose of rivaroxaban (10mg) and catheter with-
but with a half-life of at least 135h, and possibly up to drawal is thus required (Class IIa, level C). After
66 days.50 It is administered just once weekly by sub- catheter withdrawal the next dose of rivaroxaban
cutaneous injection. Its clearance is exclusively renal. maybegivenafter4–6h(ClassIIb,levelC).Available
PhaseIIIstudiesfollowingDVT,pulmonaryembolism, experience with neuraxial blockade is very limited.
andinatrialfibrillationpatientshavereportedanexcess Extreme caution is therefore recommended when
of major bleeding compared to vitamin K antagonist usingrivaroxabaninthepresenceofneuraxialblockade
treated patients.51 Its very long half-life may lead to (Class IIb, level C).
accumulation, especially in the elderly and in renal
insufficiency.Thisandtheproblemswithhaemorrhage Apixaban
havehaltedthedevelopmentofidraparinux,whichhas Apixabanisanoral,reversible,directfactorXainhibitor
been replaced with a new form coupled to biotin. The relatedtorivaroxaban.Itsbioavailabilityrangesfrom51
half-life of idrabiotaparinux is also long, it is mainly to 85%, and its inhibition constant Ki (0.08nmoll(cid:5)1) is
excreted through the kidneys, but unlike fondaparinux betterthanthatofrivaroxaban.Thehalf-lifeisbetween
andidraparinux,anantidoteisavailable.InaphaseItrial, 10 and 15h,55 and elimination follows multiple path-
the anti-Xa activity of biotinylated idraparinux was ways, with only 25% excreted renally and 75% by
immediately reversed by an intravenous infusion of hepaticandbiliarymetabolism,andintestinalexcretion.
avidin,aproteinfromheneggwhitewithahighaffinity A phase III study comparing a 2.5mg dose twice daily,
tobiotinthatiswelltolerated.Althoughthehalf-lifeof totwicedailyenoxaparin30mginorthopaedicsurgery,
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
EuropeanSocietyofAnaesthesiologyrecommendations 1005
hasbeencompleted.Ithasasimilarefficacybutappears catheters,ifneuraxialblocksareperformedatall(Class
to cause less bleeding when started 12–24h after III, level C).
surgery.56 Apixaban has currently no formal approval
for use. As with most new anticoagulants, no specific
Direct thrombin inhibitors
monitoring or antidote is available. Applying the same
Other agents for both perioperative thromboprophylaxis
rules to apixaban as to other anticoagulants 2 ((cid:3) half- and therapeutic anticoagulation include direct selective
life)wouldyieldatimeintervalof26–30hbetweenthe
thrombininhibitors.Incontrasttoheparins,theycanalso
last dose of apixaban (2.5mg) and catheter withdrawal
inactivatefibrinalreadyboundtothrombin,thusinhibit-
and suggests that at least one dose should be omitted
ing further thrombus growth. When bound to thrombin,
(ClassIIb,levelC).Aftercatheterwithdrawal,thenext
proteolyticpropertiesareinhibitedwithoutantithrombin
dose of apixaban may be given 4–6h later (Class IIb,
or other cofactors being necessary. Thrombin inhibitors
level C). As with all new anticoagulant drugs, experi-
influence, to various extents, all functional haemostasis
ence with neuraxial blockade is limited and most
testsbasedonfibrinformation,particularlyaPTT,which
patients received only single-shot spinal anaesthesia.
isusuallyusedforlaboratorycontrols.Theecarinclotting
Extremecautionisthereforerecommendedwhenusing
time (ECT) is more specific and should be used with
neuraxial blockade in the presence of apixaban (Class
therapeutic doses of thrombin inhibitors. The most
IIb, level C).
important side effect of higher doses of direct thrombin
inhibitors, particularly when used in combination with
other antithrombotic agents or platelet aggregation
InhibitionconstantKi inhibitors, is increased bleeding.59 There is no specific
The inhibition constant Ki represents the concentration of
antidote, but hirudins and argatroban can be eliminated
inhibitor that is needed to reduce the maximal rate of
by dialysis.
reaction to half that of the uninhibited value. The lower
theKi,thelowertheconcentrationofinhibitorneeded,and
themore potentthe inhibition. Desirudin, lepirudin
The recombinant hirudins, desirudin and lepirudin, are
firstgenerationdirectthrombininhibitorsthatareadmi-
Danaparoid(Orgaran)
nistered parenterally. They are indicated for thrombo-
Danaparoidisaglycosaminoglycanmixtureconsistingof
prophylaxis(desirudin)andVTEtreatment(lepirudin)in
84% heparin sulphate, 12% dermatan sulphate and 4%
patients with HIT type II.
chondroitin sulphate. Its effect occurs mainly through
antithrombin-mediatedinhibitionoffactorXa.Danapar- Incontrasttoheparins,aprolongedaPTTappearstobe
oid is used for VTE prophylaxis and treatment in HIT required for effective thromboprophylaxis. Following a
type II, although cross-reactivity with heparin-induced singleintravenousorsubcutaneousinjectionofdesirudin,
antibodies can occur in 10% of patients. Efficacy and there is a fast rise in the aPTT, which is measurable
bleeding risks have been compared with LMWH,57 but within 30min and reaches a maximum after 2h.60 The
clinicalexperienceislimited.Astheterminalhalf-lifeof aPTT is still prolonged 8h after subcutaneous adminis-
danaparoid may be markedlyprolonged inpatients with tration of low-dose hirudin (prophylactic dose 2(cid:3)15mg
renalinsufficiency,doseadjustmentsarenecessary(Class per day).61 The elimination half-life is 2–3h and is
IIa, level B). Cases of severe bleeding have been markedly prolonged in patients with impaired renal
observed with danaparoid. There is no antidote, and it function.
cannot be haemofiltered, but it can be removed using
In general, it is advisable to wait at least 8–10h, and
plasmapheresis.Coagulationmonitoringisonlypossible
longer if possible, between the administration of these
using anti-Xa activity. Despite its very long half-life of
agentsandneuraxialpuncture,andtoavoidcombinations
22h, the agent is administered twice daily for throm-
with other antithrombotic agents (Class I, level C).
boprophylaxis, so that genuine trough levels are prob-
ably notachieved. Recommendationsfor VTEprophy- Hirudinsaccumulateinpatientswithrenalinsufficiency.
laxis state that danaparoid should first be administered For desirudin, prophylactic doses should be monitored
2h preoperatively. Although neuraxial blocks were car- with aPTT in patients with creatinine clearance levels
ried out in a very small number of patients 1h after between30and90mlmin(cid:5)1,andinthosewithcreatinine
danaparoid administration, this approach is strongly clearancelevelsbelow30mlmin(cid:5)1,itiscontraindicated.
discouragedas,aftersubcutaneousinjection,significant The doses of lepirudin once approved are no longer in
plasma levels are obtained at this time.58 Instead, pre- use; bolus administration is now avoided and the initial
operative danaparoid administration should be avoided treatmentinpatientswithnormalrenalfunctionisstarted
whenneuraxialanaesthesiaisplanned(ClassI,levelC). at0.1mgkg(cid:5)1h(cid:5)1.Forlepirudin,adosereductionofup
Due to its very long half-life and accumulation in to 85% is recommended in patients with severe renal
patientswithrenalinsufficiency,itispreferabletocarry insufficiency. Treatment should be monitored using
out single-shot spinal anaesthesia and avoid the use of aPTT or ECT. Following several days of lepirudin
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
1006 Gogartenetal.
administration antibody formation may develop in postoperatively, followed by 220mg on subsequent
approximately 40% of patients, delaying elimination days. In patients with renal impairment, the respective
andleadingtosubstantialandunpredictableprolongation dosing recommendations are 75 and 150mg. Dabiga-
of its activity.62 Bleeding complications are therefore tran prolongs the aPTT, and at doses recommended
frequent. for thromboprophylaxis, this effect is significantly pro-
nounced in renal failure65 because elimination is
Althoughdesirudin(Revasc)hasbeenadministeredina
through the kidneys with potential for accumulation.
small number of patients immediately after neuraxial
The efficacy of dabigatran (220mg) in the prevention
puncturewithoutdevelopmentofspinalepiduralhaema-
of VTE is comparable to enoxaparin (1(cid:3)40mg) with-
toma,thisisnotadviseddue totheagent’s pharmacoki-
out increasing bleeding.66 In the initial studies with
netics(Class IIa, level B).A delay ofat least 2–4h after
dabigatran, neuraxial blockade was performed in
neuraxialregionalanaesthesiashouldbeobserved(Class
approximately 70%, but all epidural catheters were
IIa, level C). Time intervals of 8–10h before puncture
removed at least 4–6h before the first dose. There
apply only in patients with normal renal function and
are no reports of dabigatran use and indwelling epi-
without antibody formation, and it is therefore recom-
dural catheters. The 12–17h half-life of dabigatran in
mended that aPTT should be checked before puncture
healthy patients would suggest a time interval of 34h
(Class IIa, level C). A case of spontaneous epidural
between the last dose of dabigatran and catheter
haematoma during lepirudin treatment has been
manipulation or withdrawal, however, the manufac-
reported.63 Due to side effects resulting from antibody
turer advises against the use of dabigatran in the
formation, and anaphylactic reactions at re-exposure, in
presence of neuraxial blockade (Class III, level C).
somecountriesbothdrugshavebeenreplacedbyfonda-
This warning may have medicolegal consequences if a
parinux for thromboprophylaxis and by argatroban for
spinal epidural haematoma occurs.
treatment of acute HIT.
Argatroban (Argatra) Vitamin K antagonists (acenocoumarol, fluindione,
Argatroban is a reversible direct thrombin inhibitor, phenprocoumon, warfarin)
whichhasbeenapprovedforthetreatmentofHITtype Therapeutic anticoagulation with vitamin K antagonists
II.Argatrobanisadministeredintravenouslyandiselimi- remains an absolute contraindication to neuraxial block-
natedexclusivelybytheliver,makingrenalinsufficiency ade. As it can take several days after these drugs have
anindicationforitsuse.Therecommendedintravenous been withdrawn before coagulation returns to normal,
dose is 0.5–2.0mgkg(cid:5)1min(cid:5)1 in patients with normal progress has to be checked using the international nor-
organ function, and the dose should be adjusted to malised ratio (INR). The process can be accelerated by
maintain the aPTT between 1.5 and 3 times normal. administering vitamin K, prothrombin complex concen-
Dosereductionisrequiredinthecriticallyillandinthose trates (PCC), or, if PCC is unavailable, fresh frozen
with heart failure or impaired hepatic function. When plasma. This requires an appropriate indication and
hepaticfunctionisgood,normalisationoftheaPTTtakes should follow an individual risk-benefit analysis. The
only2–4haftertheendofargatrobaninfusion,duetothe useofcoagulationfactorspurelytonormalisecoagulation
short half-life of 35–45min.64 before neuraxial anaesthesia, is inappropriate (Class III,
level C).
Ifneuraxialblockadeisconsidered,adistinctionneedsto
be made between patients with a history of HIT, who In contrast to most European countries, in the United
require only thromboprophylaxis, usually with low-dose States postoperative use of vitamin K antagonists for
danaparoid, or fondaparinux (off-label) subcutaneously, thromboprophylaxis is widespread, but they are less
and those with acute HIT type II, in whom therapeutic effective than UFH or LMWH. Horlocker et al.67
anticoagulation is required. In the latter, there is a high described 188 patients with epidural analgesia, who
riskofVTEifanticoagulationisinterrupted.Inpractice received postoperative low-dose thromboprophylaxis
thisgroupfrequentlysuffersfrommultipleorganfailure with warfarin. Epidural catheters were removed during
that includes coagulation, and they may be ventilator warfarin therapy, without incident. The small number
dependent, making neuraxial blockade inadvisable of patients included, and the early removal of epidural
(Class III, level C). catheters before the onset of effective anticoagulation
(INR <1.4), prevent any worthwhile comment on the
Dabigatran(Pradaxa) safety or otherwise of this practice. Withdrawal of
Dabigatran is an oral reversible monovalent thrombin epidural catheters when vitamin K antagonists have
inhibitor that has recently been approved for VTE already taken full effect is not recommended, and
prophylaxis in patients undergoing hip or knee repla- accordingly they should be administered only after
cement.55 The oral prodrug dabigatran etexilate is the catheter has been removed (Class IIa, level B).
metabolised by plasma esterases into dabigatran. Another analysis of 950 patients undergoing epidural
The first dose of dabigatran 110mg is given 1–4h analgesia, in whom vitamin K antagonists were started
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
EuropeanSocietyofAnaesthesiologyrecommendations 1007
preoperatively, similarly found no complications. How- relatedtoaspirin.Pregnancyisassociatedwithageneral
ever, no detailed information was provided regarding low-gradeactivationofcoagulation,andaspirinwastaken
coagulation status.68 aloneratherthanincombinationwithotherthrombopro-
phylaxis.Thesetwofactorswouldhavereducedtherisk
These two studies apart, there are several reports of
ofepidural haematomainthiscohort.8Horlockeretal.76
spinal epidural haematoma following neuraxial regional
includedatotalof924orthopaedicpatientsintheirstudy,
anaesthesia with concurrent use of vitamin K
193ofwhomhadtakenaspirinpreoperatively.Intakeof
antagonists.17 In view of the high rate of bleeding com-
aspirin was defined as intake within the previous week,
plications in patients receiving therapeutic doses, this is
although the extent of platelet aggregation inhibition
notsurprising.IntheUnitedStates,asinEurope,thera-
declinesmeasurably3daysafterwithdrawal.70Thenum-
peuticanticoagulationwith oralvitaminKantagonistsis
ber of neuraxial blocks in patients receiving aspirin was
regarded as a contraindication to neuraxial blockade.
small. Only 22 of 1000 patients received concurrent
However, in contrast to Europe, despite the problems
thromboprophylaxis with heparin, and no epidural
associatedwithcatheterremoval,perioperativethrombo-
haematomas were observed. In another study including
prophylaxiswithwarfarinisstillanoptionintheUnited
1035 patients with no accompanying thromboprophy-
States.14
laxis, again no epidural haematoma occurred in 158
patients who had taken aspirin within the previous
Platelet aggregation inhibitors week.77
Acetylsalicylicacid (aspirin) and theriskofbleeding
Although the administration of aspirin alone does not
Even after a single dose, aspirin leads to irreversible
appear to increase haematoma formation, a higher rate
inhibitionofplateletfunctionasaresultofcyclo-oxyge-
of complications has been observed in both surgical and
naseinhibition.Duetothelackofanucleus,plateletsare
medical patients when heparins were administered con-
notabletosynthesisenewcyclo-oxygenase,sothatthe
currently.25,26Becausepreoperative,versuspostoperative,
effect persists for the entire lifetime of the platelets,
thromboprophylaxis is not proven to be beneficial,21 a
usually 7–10 days. However, healthy bone marrow will
cautionary approach in the presence of aspirin would be
replace more than 30% of the irreversibly inhibited
tostartVTEprophylaxispostoperatively(ClassI,levelB).
platelets within 3–4 days.69,70 With a normal platelet
count, this is usually sufficient to return haemostasis to Non-steroidal anti-inflammatory drugs (NSAIDS) also
normal. Bleeding effects of aspirin appear to be dose- inhibit cyclo-oxygenase and thus platelet aggregation
dependent, with more adverse events observed in in a reversible manner that is proportional to the half-
patients receiving more than 100mg daily.71 Analyses life of the agent used. With the exception of tenoxicam
of medical patients have shown that the risk of spon- and piroxicam, platelet function normalises within 12–
taneous bleeding in those taking aspirin is doubled, 24h after withdrawal.78 To avoid any negative effect of
although the risk remains generally very low. A total NSAIDS on platelet function and neuraxial block, it is
of800patientsannuallywouldneedtobetreatedforone sufficient to miss a dose the evening before a planned
additional bleeding event to occur.72 A slight increase procedure or catheter removal. For selective COX-2
in bleeding complications is also observed in surgical inhibitors, there is no evidence of any relevant effects
patients receiving aspirin, but transfusion is rarely onplateletaggregationcapacityoranincreasedbleeding
required.Incardiacsurgery,theriskofrelevantbleeding tendency.79 The non-opioid analgesics paracetamol
isincreased1.4-fold,makingaspirinlessimportantthan (acetaminophen) and metamizole (dipyrone) have not
thedurationofextracorporealcirculationorthepresence been linked to spinal epidural haematoma to date.
ofrenalinsufficiency.73Inameta-analysisofnon-cardiac
On the basis of the available data, NSAIDS, including
surgicalprocedures,1.5-foldhigherbleedingrateswere
aspirin,whengiveninisolation,donotincreasetheriskof
reported in patients receiving aspirin. With the excep-
spinalepiduralhaematomaandarenotacontraindication
tion of intracranial procedures, prostatic resections and
toneuraxialblock(ClassIIb,levelC).Spinalanaesthesia
tonsillectomies, no severe cases of bleeding were
has better support than epidural (Class IIb, level C).
observed.74
Only three studies consider the safety of neuraxial Aspirinand the riskofadverse cardiovascularevents
regional anaesthesia during aspirin administration.75–77 In patients with a history of acute coronary syndrome,
Although the Collaborative Low-dose Aspirin Study in stroke, or peripheral arterial occlusive disease, aspirin
Pregnancy (CLASP) included a total of 9364 pregnant reduces the risk of a recurrent cardiovascular event by
women, only 2783 of the women underwent epidural around30%,andmortalitybyapproximately15%.80Case
analgesia. Of the latter, 1422 had taken aspirin during series in recent years suggest that morbidity and
their pregnancy, and only half had continued it up to mortality,particularlyinpatientswithrecentlyimplanted
delivery, leaving approximately 700 women for analysis. coronarystentsorunstablecoronarysyndromes,ismark-
Noneofthewomeninthisstudyexperiencedaproblem edly increased if aspirin is stopped before a surgical
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
1008 Gogartenetal.
procedure.74,81,82 A rebound phenomenon has also been initial bolus administration of 300–600mg. Recovery of
discussed.83Theriskoflate stentthrombosisis greatest platelet function occurs only 6–7 days after the end of
in patients with drug-eluting stents. In summary, the clopidogrel administration,89 so that neuraxial anaesthe-
preoperativewithdrawalofaspirinisunnecessaryinmost siashouldonlybeperformedatleast7daysafterthelast
cases and is associated with a high risk of acute throm- intake (Class IIa, level C).
bosis(ClassIIa,levelC).Itisrecommendedthatpatients
with acute coronary syndromes or stent implantation Thienopyridines and the riskofbleeding
should continue to take aspirin on a lifelong basis.84 Clopidogrel treatment in patients undergoing cardiac
Those with drug-eluting stents should only stop aspirin surgery may cause severe surgical bleeding, with a 2.5-
beforeaplannedoperationwhenthereisalife-threaten- foldincreaseintheneedfortransfusionanda5–10-fold
ing bleeding risk such as in neurosurgical procedures increase in the risk of repeat surgery, as well as a pro-
(Class IIa, level C). It should be noted that concurrent longed course of intensive therapy.90–92 Severe peri-
intakeofibuprofenmayleadtoareductionintheefficacy operativebleedingisclearlymorefrequentwithclopido-
of aspirin79 and that all NSAIDS increase the risk of grelthanwithaspirin.Theincidenceofseverebleeding
adverse cardiovascular events. The ACCP does not cur- is increased by simultaneous treatment with vitamin K
rentlyrecommendassessmentofplateletfunctionbefore antagonists,dextrans,orheparins.86Theextentofbleed-
invasive procedures because there is no apparent corre- ing complications in other surgical procedures without
lation with bleeding (Class II, level C).85 full intraoperative heparinisation has not been ade-
quatelyinvestigated.Immediateimprovementofhaemo-
Thienopyridines stasis is only possible by administering platelets (Class
Ticlopidine(Ticlid)andclopidogrel(Iscover,Plavix)are IIa, level C).
platelet aggregation inhibitors belonging to the thieno- As there have already been reports of spinal epidural
pyridines group. They act by antagonising adenosine haematomafollowingneuraxialblockadeduringclopido-
diphosphate (ADP) at the platelets’ purine receptors. greladministration,18,19currentadviceisagainstallsuch
ADP-inducedaggregationisnon-competitivelyandirre- proceduresunlessthetreatmenthasbeeninterruptedfor
versibly inhibited, whereas arachidonic acid metabolism 7 (clopidogrel) to 10 days (ticlopidine).
is unaffected. The two agents undergo hepatic conver-
sion into active metabolites in vivo. Consequently, sev- Thienopyridines and the riskofadversecardiovascular
eral days are required before full effect is reached, and events
the process is dose-dependent. The active metabolites Compared to aspirin, clopidogrel and ticlopidine are
are eliminated by the kidneys.86 Finally, a reduction in better at preventing ischaemic cerebral infarction, myo-
ADP-dependent activation of the glycoprotein IIb/IIIa cardial infarction, and vascular deaths in general.93
receptor takes place, causing a reduction in fibrinogen Patients with unstable coronary syndromes, previous
fixation and platelet cross-linking. percutaneous coronary interventions, and stent implan-
tations,benefitfromlong-termdualplateletaggregation
inhibitionwithaspirinandclopidogrel.94Ifthistreatment
Ticlopidine
combination is prematurely withdrawn following coron-
Maximum aggregation-inhibiting effect is achieved after
ary intervention, the risk of acute stent thrombosis and
8–11daysonaticlopidine500mgperdayregimen.The
myocardialinfarctionissubstantiallyincreased,withhigh
eliminationhalf-lifeis24–32h,butbecauseplateletfunc-
mortality.95 This also appears to be the case even when
tion inhibition is irreversible, the effect is still evident
more than 72h after its withdrawal.87 With long-term perioperative bridging is carried out using heparin, and
the platelet aggregation inhibitors are only withdrawn
administration, the elimination half-life increasestoover
very briefly.96 Patients with a drug-eluting stent are at
90h, and following cessation, 10 days should be allowed
before normalisation can be expected.88 In contrast to riskforaparticularlylongperiod,duetolateandincom-
plete endothelialisation.97 Consequently the American
clopidogrel, ticlopidine can lead to neutropenia in up to
Heart Association currently recommends that drug-elut-
1% of patients, limiting widespread use of the drug and
ing stents are only used provided no surgery is planned
making regular blood count checks necessary in the
withinthefollowing12months,andthepatientsshowa
initialweeks.
high degree of compliance.98 A cardiologist should be
Neuraxial regional anaesthesia should not be performed consultedbeforeanyinterruptionofplateletaggregation
until10dayshaveelapsedbetweenthelastingestionand inhibition, and clopidogrel with aspirin should be admi-
the procedure (Class IIa, level C). nisteredduringtheperioperativeperiod(ClassI,levelC).
With bare metal stents, dual platelet aggregation inhi-
Clopidogrel(Iscover, Plavix) bitionshouldbeadministeredforatleast4–6weeksand
Following oral administration of clopidogrel 75mg, the with both types of stent, the administration of aspirin
maximumplateletfunction-inhibitingeffectisobserved shouldbecontinuedonalifelongbasisandperioperative
after3–7days,orafterapproximately12–24hfollowing interruption should be avoided.98
EuropeanJournalofAnaesthesiology 2010,Vol27No12
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.