midface rejuvenation with hyaluronic acid: a critical ... · midface rejuvenation with hyaluronic...
TRANSCRIPT
1
MidfacerejuvenationwithHyaluronicAcid:A critical appraisal of the vascular complication risks and thedevelopment of evidence based protocols for their prevention andmanagementMary-JaneSigridRowland-WarmannQueenMaryUniversity,London2016WordCount:16836excludingreferences
2
Abstract
Background:Fortunatelyrare,vascularcomplicationscausingskinnecrosisorblindnessare
feared consequences of hyaluronic acid filler treatment. Cases of skin necrosis from
intravascular injectionofhyaluronicacid fillersorvessel compressionhavebeen reported
worldwide,whereascasesofvisionlossfromhyaluronicacidfillersarelargelylimitedtoAsia.
TheauthorisgravelyconcernedabouttheinadequacyofcurrentUKlegislationsurrounding
dermalfillersandthelackoftrainingformanagementofvascularcomplications.
Objectives: A review of the literature was conducted to assess the risk areas for filler
treatment and compile evidence based methods for prevention and management in an
aestheticpracticesetting.
Methods: A comprehensive literature search was performed to include relevant articles
basedonspecifiedinclusionandexclusioncriteria.Thetypeoffillerwaslimitedtohyaluronic
acid,witharecordofcomplicationsasskinischaemia,softtissuenecrosis,visualdisturbance
orblindness.
ResultsandDiscussion:Tenarticlesrepresenting41patientswithvascularcompromisewere
identified.Datafromthesereportsincludeinjectionsite,symptoms,treatmentandoutcome.
Thetreatmentsiteassociatedwiththehighestincidenceofvascularcompromiseisthenose,
representing42%ofvisionlossand59%ofskinischaemiacases.62%ofskinischaemiacases
resolvedcompletely,irrespectiveoftimetopresentation.Themostsuccessfulmanagement
componentforskinischaemiaappearedtobeearlyinterventionwithhyaluronidase.None
of the twelve vision disturbance cases showed improvement compared with their initial
presentationwiththetreatmentmeasuresperformed.Forvisionlossthereappearstobeno
effectivemanagement.
Conclusions: Hyaluronic fillers arewidely used in aesthetic practice. It is imperative for
practitionerstobeabletoprevent,recogniseandmanagevascularcompromise.
3
Acknowledgements
I’dliketothankEd,mybusinesspartner,whohashelpedmeestablishourfacialaesthetics
practice.Andmycats,whoareawelcomedistractiontoanotherwiseupsettingtopic.
4
Contents
ListoffiguresListoftablesIntroduction DermalFillers Hyaluronicacid Hyaluronidase Complicationswithdermalfillerinjections Vascularcomplicationswithdermalfillertreatment Midfacerejuvenationandnosereshaping Theanatomyrelativetovascularcomplications Relatedstudies Currenttreatmentstandards
PracticeimplicationsExampleClinicalScenarioAimsandObjectives
MethodsResultsDiscussion Fillertypes Highriskareasandtheeffectonvessels Interventions Skinischaemia/necrosis Visionloss/disturbance Outcomes Practitionerstatus
Evidencebasedguidelinesforthepreventionandmanagementofvascularcomplications
Preventionprotocols Managementprotocols
TheEmergencykitFurtherstudiessuggested
ConclusionsReferences
5
ListofFigures
Figure1:Thevasculatureofthemidface
Figure2:Algorithmforthemanagementskinischaemia/necrosis
Figure3:reconstitutionanddosageofhyaluronidase(Hyalase/Wockhardt)
Figure4:Vascularcomplicationreportingform
ListofTables
Table1:Overviewof10selectedpapers,theirbasicoutcomesandweaknesses
Table2:Casecollectionfromkeypapers
Table3:Preventionprotocols
Table4:managementofvisionloss
Table5:managementofskinischaemia/necrosis
Table6:Emergencykitforvascularcompromise
6
Introduction
The worldwide use of dermal fillers for rejuvenation and facial reshaping is showing an
explosiveincreaseoverthelastdecade.
An 31% increase in dermal filler treatments, totalling 2,690,633 worldwide in 2014, was
recordedbyISAPSsince2013.Thissteadyincreasehasbeenseensince2011accordingto
theAmericanSocietyforAestheticPlasticSurgery,risingbyapproximately30%yearonyear
(Cavallinietal.,2016; ISAPS,2016). Closer tohome, in theUK,approximately1.5million
botulinumanddermalfillertreatmentswereperformedin2013,anincreaseof4%from2012
(Inglefieldetal.,2014). The reason for this increase is simple:dermal filler injectionsare
convenient,predictable,achievegoodresultsandarelargelysafe(Hsieh,Lin,Huang,&Yeh,
2015).
Theincreaseintheuseofdermalfillersworldwidewillmeananincreaseinthenumberof
complications,especiallysincethecomplexityoftheperformedtreatmentsisincreasingfrom
thesimpletreatmentofperiorallinesforwhichdermalfillerswereoriginallyapprovedbythe
FDA in 2003 (Abduljabbar&Basendwh, 2016;US FDA, 2016). Thus, it is imperative that
awarenessandknowledgetotreatcomplicationsareincreasedinabidtoimprovepatient
safety.
Itistooeasyandarroganttosaythatthebestwaytomanagecomplicationsisnottohave
any(Cavallinietal.,2016).Somecomplicationsareunpredictableandhappentoeventhe
most experienced clinicians. Selection of the appropriate filler, using it correctly and
conductingtheproceduresafelyareallfactorswhichmitigatetheriskofcomplications,as
improper technique and misuse of filler account for many easily avoided complications.
However,ifcomplicationsstilloccurthepractitionerhasadutytomanagethemtothebest
oftheirabilitiesintheinterestsofthepatient.
Vascular complications are certainly disproportionate to the expected outcome of the
procedure.Whilstthenumberofcasesofblindnessresultingfromdermalfillertreatmentis
approaching100worldwide,andtheincidenceofnecrosisisthoughttobelessthan0.1%of
7
allproceduresperformed,underreportingbycliniciansduetoembarrassmentmeansthatthe
trueincidenceisunknown(Cohenetal.,2015;DeLorenzi,2014;Sunetal.,2015).
DermalFillers
Whenselectingdermalfillersforpatientsinpractice,itisimportantthattheproductissafe
andeffective, not allergenic andgives reproducible results. Furthermore, they shouldbe
noncarcinogenic, nonmigratory and nonteratogenic (Alijotas-Reig, Fernandez-Figueras, &
Puig,2013).Onapurelyaestheticbasis,fillersshouldprovidesupportsandstructurewhere
thefeaturesofthefaceareaffectedbyvolumeloss(Montes,2012).
IntheUSA,theFDAregulateswhichdermalfillersmaybeused.Therearecurrently24FDA
approveddermalfillersintheUSA,ofwhich13arehyaluronicacid(USFDA,2016).InEurope
regulationismuchlessstrictandpractitionerscanchoosefrommanytypesoffillers–some
of lesserquality thanothersdue to lax regulationsoverproduction,distributionanduse.
There are over 160 types of filler and over 50manufacturers in the world. Low quality
productscancontributetothemostseriouscomplicationsandpractitionersshouldremain
vigilant,ascheapproductsareoftenthosewithleastscientificresearchintotheeffectson
thehumanbody(Cavallinietal.,2016).
Hyaluronicacid
Hyaluronic acid (HA) are biodegradable fillers that injected under the dermis to cause a
temporary change in appearance before being biodegraded (Funt & Pavicic, 2013).
Hyaluronic acid is a fundamental componentof theextracellularmatrixof cells (Cavallini,
Gazzola, Metalla, & Vaienti, 2013). It is a glycosaminoglycan (GAG) made of N-acetyl
glucosamineandglucuronic acid to forma lineardisaccharidepolymer, awater retentive
naturalsugarthathelpstohydrateandvolumisetheskin(Kassir,Kolluru,&Kassir,2011).
8
HAfillersarethemostcommonlyusedfillers,accountingfor78.3%ofallfillertreatmentsand
an increaseof253%since2000 (Abduljabbar&Basendwh,2016). Syntheticallyproduced
usingbacterial cultures,HA ismodifiedbyacross-linkingandstabilisingprocesswith1,4-
butanediol diglycidyl ether (BDDE) to achieve the appropriate permanence in tissues and
deliverproductsthatcanlastbetween4-12monthsonaverage(Cavallinietal.,2016).They
are described as either biphasic or monophasic, with a particle size of around 400µm.
BiphasicHAfillershavelittleornocross-linking,theirparticlesizedeterminingtheproduct
density (e.g. Restylane). MonophasicHA fillers are crosslinked to formhomogenous gels
whereHAconcentrationcombinedwithcrosslinkingdeterminesfillerdensity(e.g.Juvederm,
Belotero)(Cavallinietal.,2016).Onceintissue,HAexpandsduetoitshydrophilicproperties
(Grunebaum,BogdanAllemann,Dayan,Mandy,&Baumann,2009)
Wheninjectedintothetissues,HAcanprovidestructureandelasticitybybindingcollagen
and elastin (Kassir et al., 2011). The primary reason for its popularity in use is its
biocompatibility and reversibility (Abduljabbar & Basendwh, 2016), which is a preferable
propertywhentheimpactofHAontissuescancausedevastatingconsequences.However,
thispropertyisbynomeansamagicwand.
Hyaluronidase
Hyaluronidasecanbeusedtomanageunaestheticoutcomesandseriouscomplicationssuch
asvascularcompromiseafterdermalfillertreatment.
HyaluronidaseisanendoglycosidasewhichdepolymerisesHAfillersbyhydrolysisofthe1,4-
N-acetylglucosaminidic bond in hyaluronic acid (Cavallini et al., 2013; Hilton, Schrumpf,
Buhren,Bolke,&Gerber,2014).Originallylicensedtoimprovepermeationofintramuscular
andsubcutaneousinjectionsandinfusions,hyaluronidaseincreasesdrugdiffusionintothe
extracellularmatrixandincreasesbloodvesselpermeability(Cavallinietal.,2013;King,2014).
Theactionofhyaluronidaseisrapidandcompletewithin24-48hours.Thedisruptionthatis
causedtothedermalbarrieristransient,reformingafteraround48hours(Kassiretal.,2011).
9
Thefirstreportoftheuseofhyaluronidasetomanagedermalfillercomplicationswasin2007
by Hirsch et al (Hirsch, Cohen, & Carruthers, 2007) and early injection after a vascular
complicationsignificantlyreducestheriskandextentofnecrosis.ItnotonlydegradestheHA
filler,butcanalsoreducepressureontheoccludedvesselsandreduceoedema(Beleznay,
Humphrey,Carruthers,&Carruthers,2014). Dosages inthe literatureremainvaried,with
somestudiessuggestingbetween150U(Sunetal.,2015)to200Urepeatedafteronehour
(Cohenetal.,2015),andevenashighas1000U(Cavallinietal.,2016). It is thought that
althoughhyaluronidaseisassociatedwithahighincidenceofallergy,toolittlecanresultin
scarring(Cohen,2008).FillerwithhigherHAconcentrationsmayrequiremorehyaluronidase
(Cavallinietal.,2013).
Thereismuchhypeaboutthecomplicationratewiththeuseofhyaluronidase,theincidence
ofwhichisreportedataround0.05-0.1%.However,thishasbeenshowntoincreasetoas
highas30%whenhighdosesareadministered.ThemechanismofallergicreactionisTypeI
andTypeIVmediatedhypersensitivity,andcanresultinurticarial,erythemaandoedemabut
also,moreseriously,inrash,angioedemaandanaphylaxis(Cavallinietal.,2013;Cavalliniet
al.,2016). For this reason, it isoftenadvised to testhyaluronidaseon thepatientbefore
administering it;however, inthecaseof impendingnecrosis it isthoughtthat itwouldbe
moredamagingtohesitateanddirectadministrationofhyaluronidaseisadvised(Cohenet
al.,2015).
Thereason for its immunogenicity is thesourceofhyaluronidase. It isgained fromovine
testis and thus contains animal products, but also contains lactose components in the
hyaluronidasepowder.IntheUK,ovinehyaluronidaseisavailableasHyalase(Wockhardt)in
a1500unitampouleaspowder for reconstitutionwithsalineorwith lidocaine to further
vasodilation(Cohenetal.,2015;King,2014).InAmerica,ahumanrecombinantformulation
ofhyaluronidase,Hylenex,isavailablewhichissaferandlessimmunogenic(Cavallinietal.,
2013).
When using hyaluronidase in aesthetic practice, it is important to consider that it is
antagonised by furosemide, benzodiazepines, dopamine, alpha-adrenergic agonists,
antihistamines and phenytoin, some anti-inflammatories such as indomethacin and
10
antioxidantssuchasvitaminC(Cavallinietal.,2013;Cavallinietal.,2016;Glaich,Cohen,&
Goldberg,2006).
Hyaluronidaseisnotlicensedforcorrectingproblemsresultingfromfillertreatmentandis
usedoff-labelaccordingtoArticle87ofDirective2001/83/EC.Therefore,informedconsent
isrequiredfromthepatient,withathoroughoutlineofthemedicalandsurgicalalternatives
(Cavallinietal.,2016;Cohenetal.,2015;King,2014).
Complicationswithdermalfillerinjections
Non-surgicalcosmetic interventionssuchasdermal filler treatmentscarrywith themrisks
that the patientmust bemade aware of, as patient satisfaction is a significant outcome
measure.Often,sideeffectsaccompanytreatment,anddiscomfort,erythema,swellingand
haematomaarecommonplaceandtobeexpected (Kassiretal.,2011). Theseareusually
mild, transient and self-resolving, requiring no intervention except reassurance from the
treatingclinicianastotheirfiniteduration.
Complications from dermal fillers, on the other hand, are unexpected in nature and
disproportionatelynegativetotheintendedtherapeuticoutcomeeventhoughthetreatment
has been administered correctly and the product used normally. These can be
catastrophically severe and long lasting andmay result in permanent injury or aesthetic
deficitsifnottreatedcarefullyanddiligently(Cavallinietal.,2016;Y.Chenetal.,2014)
Vascularcomplicationswithdermalfillertreatment
Vascularcomplications,althoughconsideredtoberareatarateofaround0.1%,arethemost
frightening and serious consequences of dermal filler treatment (Sun et al., 2015).
Embolisationofavesselwithfillermaterialbydirectinjection,orcompressionofavesselby
the placement of product in close proximity, can progress to necrosis of tissue due to
ischaemia(Cohenetal.,2015;Kassiretal.,2011).Thiscanleadtoareasofskinsuccumbing
11
tooxygendeprivation.Similarly,blockageofthevesselsthatsupplytheeyeduetoaccessof
dermal fillerparticles to theocularcirculationcancausenecrosisof the retina, leading to
blindness.
Almost all filler materials in current and historical use have been linked to vascular
complications and blindness, including paraffin, silicone oil, hyaluronic acid,
polymethylmethacrylate,calciumhydroxylapatiteandfattissue(Hsiehetal.,2015).
In the case of occlusion of an artery, pain is often immediate and striking in naturewith
blanchingofthesupplytissueorlossofvisionoftheaffectedeye.Whenvisionlossoccurs,
partial or complete loss of light perception, limitedeyemovement and retinal cherry-red
spotsmayalsobeobserved(Hsiehetal.,2015).Whenvenousocclusionhasoccurred,itis
oftenaccompaniedbyadullachingpainwithdarkviolaceouspatchesatthevesseltributary
site(Kassiretal.,2011).Thepatientmayalsoreportthatthereispaindisproportionateto
theeffectedtreatment(Inglefieldetal.,2014).However,symptomsarenotalwaysthesame
and unusual or delayed presentations can occur as a result of swelling of the tissue or a
hydrophilicfillerleadingtocompression.Vascularcomplicationsaremorelikelytooccurin
areaswherethebloodflowislimited,orwherethereisalackofcollateralcirculationaffecting
thetissuesatrisk.Thisiswhythemidfaceandnoseareofparticularrelevancewhentreating
withdermalfillers.
Midfacerejuvenationandnosereshaping
Themidfaceissubjecttoage-relatedchanges,withlossofvolumeresultinginthedescentof
fat compartments and skin laxity, and deepened nasojugal and nasolabial folds. These
changesinvolumebringaboutanagedappearance,whichisoftenthepresentingcomplaint
ofapatientatanaestheticpractice.Replacementofvolumetothefatpadsofthefaceand
lendingimprovedstructurewithadvancingagecanresultinhighsatisfactionratesofupto
75%at2years,withpatientsreportingan improvement inperceivedageofupto5years
(Few,Cox,Paradkar-Mitragotri,&Murphy,2015).
12
Thenoseisnotanareaofthemidfacethatissubjecttoextremechangeswithage,butitisa
partthatisoftenthesourceofdissatisfactioninpatients.Whetheritistheshapeorsimply
adefectinthenosesuchasabumporanindentation,manypatientsseektoaugmenttheir
noses.Dermalfillertreatmentofthenoseisanexcellentsolutionforpatientswhowantto
camouflage small tomedium sized deformities of the nose but who are concerned with
financialexpense,aestheticrisk,surgerydowntimeorpermanenceoftheeffect(Humphrey,
Arkins,&Dayan,2009).Duetothenatureofthenosetissue–itisnonmobileandnotsubject
tothesamestressesasmobileandhighlyexpressionatepartsofthemidface–dermalfillers
canlastlongerthaninotherareas.
The midface and nose are considered amongst the highest risk areas for vascular
complicationsindermalfillertreatment.
Theanatomyrelativetovascularcomplications
Figure1:Thevasculatureofthemidface(from(D.Lazzerietal.,2012))
13
Thefacialartery,abranchoftheexternalcarotid,isthemajorvesselsupplyingbloodtothe
superficial face. At themandible, the facial artery crosses diagonally towards the nose,
runningatortuouspathatthenasolabialfold,beforeadvancingtothelateralnasalwallin
thealarcreaseandterminatingintheangularartery.Thisrunstowardsthemedialorbital
rim.Atthejunctionbetweenthefacialandangulararteriesliesthelateralnasalartery,which
suppliesthe78%ofthebloodtothenose(nasaltipandala),andbloodtotheremaining22%
ofthenoseviathedorsalnasalartery.Compressionorocclusionofthefacialarteryinthe
nasolabialfoldorfurtheronattheangularordorsalnasalarterycanleadtoalarnecrosisor
necrosisofthetipofthenose(Kassiretal.,2011).
Thefacialvein,whichrunsalongsidethefacialartery,liesatadistanceofaround2-3cmlateral
tothealaofthenoseandsimilarlyattheleveloftheoralcommissure.Thefacialveinalso
demarcatesthemedialborderofthemedialfatpadoftheface,whichistriangularandoften
subject to volumisationwith fillers (Chinnawong, Tansatit, Phanchart,& Rachkaew, 2015;
Cotofanaetal.,2015).Compressionorocclusionofthefacialveinoritstributariescanlead
to a dull ache, dusky grey/mottled appearance and progressive ischaemia of the tissues
(Inglefieldetal.,2014).
Theanastomosisofthedorsalnasalarterieswiththeophthalmicarteriesoccursatthemedial
canthusandfromherebloodcanaccessthecentralretinalarteriesviatheophthalmicarteries
(Kassiretal.,2011;Y.J.Kim,Kim,Song,Lee,&Yoon,2011).Theophthalmicarteryisabranch
of the internal carotid, and in turn supplies its orbital group, consisting of lacrimal,
supraorbital,posteriorethmoidal,anteriorethmoidal, internalpalpebral, frontalandnasal
arteries,andtheoculargroup,consistingoflongciliary,shortciliary,anteriorciliary,muscular
andcentral retinaarteries (S.N.Kimetal.,2014). Excesspressureon theplungerabove
systolicduringinjectionoffillerstothemidfacecanovercometheanastomosisbetweenfacial
andocularcirculationandrefluxthroughtotheophthalmicarteryoritsbranches(Tansatit,
Moon,Apinuntrum,&Phetudom,2015). Resulting ischaemiaof theophthalmicarteryby
blockagewith dermal filler leading to lack of blood supply of any of the branches that it
supplies can cause palsy of oculomotor nerves, blepharoptosis, exotropis, palsy of ocular
musclesandlossofvision(Kwonetal.,2013).Afurthercatastrophicconsequenceofpressure
overcomingtheanastomosisisaresultingcerebraleventfromfillerproductretrogradeflow
14
intotheinternalcarotidartery(Y.Chenetal.,2014;D.Lazzerietal.,2012).Furthermore,
highinjectionforcedepositshighervolumesoffilleratspeed,whichcanalsoleadtosudden
compressionofvessels,causing ischaemiaandsubsequentnecrosisofskin (Y.Chenetal.,
2014;D.W.Kimetal.,2011).
Theglabellahas limitedcollateralbloodflow,so it representsanotherareaathighriskof
vascular compromise. Branches from supratrochlear and supraorbital arteries supply the
glabellaandinferiorcentralforeheadandthenasalroot,withverysuperficialvessels(Glaich
etal.,2006;Kassiretal.,2011).Injectionoffillerintoeitherofthesecanresultindirectentry
intotheocularcirculationandresultantblindness,butalsoconsequencessuchasnecrosisof
theskinoftheforeheadandscalp(Kassiretal.,2011).
Relatedstudies
ThestudythatinspiredthisreviewisthecaseseriesreportbyBeleznayetalaboutvascular
complicationsindermalfillertreatment.Variousfillerswereanalysedandmethodsforthe
managementofvascularcomplicationsleadingtoimpendingskinnecrosisweredevised.This
studyisinterestingastheprotocolsdevisedweresimple,clearlysetoutandeasytofollow,
and the cases documented all had favourable outcomes (Beleznay et al., 2014). These
protocolscanbeutilisedinaestheticpracticeratherthanahospitalsettingandpiquedthe
author’sinterestindevisingmanagementprotocolsforbothskinandocularcomplications
fromfillerocclusionthatcanbeutilisedinpracticeintheUK.
Some similar reviews of the literature have been completed. In 2014, Carruthers et al
compiled a literature review of blindness. The focus was on prevention and therapy.
Autologous fat was identified as the main cause of filler blindness, but also that any
attempted interventions were unsuccessful. A technique for retrobulbar injection of
hyaluronidase was developed, in the hope that hyaluronidase injected directly into an
ophthalmicarterywouldcatabolisethecloggedvessel.Unfortunately,thistechniquewould
need to be completed by a neuroradiologist or ophthalmologist only and would not be
suitablefortheaestheticpractice(Carruthers,Fagien,Rohrich,Weinkle,&Carruthers,2014).
15
Lazzerietalcompletedaliteraturereviewin2012,inwhich29articlesrepresenting32cases
ofvascularcomplicationleadingtoblindnesswereanalysed.47%wereduetoautologousfat
injection,andprecautionprotocolsweresuggested.Treatmentprotocolsincludedlowering
theintraocularpressurebyincisionwithablade,IVdiuretics,carbogen(5%carbondioxide
and 95% oxygen) rebreathing alongside corticosteroids, but all proved to be fruitless (D.
Lazzerietal.,2012).
AfurtherreviewoftheliteraturewasconductedbyOzturketalin2013,where61casesof
vascularocclusionotherthanautologousfatwereconsidereddatingfrom1991to2012.This
studyassociatedthenosewiththehighestriskwith33.3%ofcomplicationshavingtreatment
at this site. A variety of fillers are considered and several complication types other than
vascularare includedandguidelinesforthemanagementoffillercomplications ingeneral
weredevelopedwhichincludedin-practicemanagementandreferral(Ozturketal.,2013).
It appears that no reviews have focused exclusively on the prevention and treatment of
vascular complications of HA filler in an aesthetic practice setting that include papers
publishedpast2012.SinceHAisthemostwidelyusedfiller,itseemspertinenttoconsider
casesthathaveresultedfromitsuse.Evidencebasedguidelinesthatareeasytoimplement
developed from this review canbeused to favourably influence theoutcomeof vascular
complicationsandinspirefurtherresearch.
Currenttreatmentstandards
Currently there are no reliable standard treatment protocols for vision loss as a result of
dermalfillertreatment(Hsiehetal.,2015)andpreventionisamuchmoreeffectivestrategy.
Managementofimpendingskinnecrosisisamuchmorewidelyappraisedtopic,withmany
papersoutliningguidelinesformanagement.Protocols,suchasthosepublishedbyInglefield
et al in 2014, or Beleznay et al in 2014 focus on the use of hyaluronidase,massage, hot
compressesandanti-inflammatorymeasuresasatreatment(Beleznayetal.,2014;Inglefield
16
et al., 2014). These recommendations appear to be realistic and the success rates and
outcomesadequate.
Practiceimplications
Patientscommittodermalfillereveryday,believingthatitisasimpleandquickprocedure.
Thisisindeedthecaseiftheprocedureisperformedsafely,byanexperiencedclinician,in
therightenvironment.IntheUK,thelegislationsurroundingtheuseofdermalfillersislax
andpatientsarepotentiallyatveryhighriskofharm.Practitionersdonotneedtoholdany
formalqualificationstopracticedermalfillers,anddonotneedtobehealthcareprofessionals
togainaccesstodermalfillerproductsoradvertisetoandtreatpatients(Keoghetal.,2013).
Asthemidfaceisthemostatriskareafordermalfillercomplications,soitwouldnotbeafar-
fetchedtaletosuggestthatavascularcomplicationcouldoccur inaestheticpracticefora
medically healthy patient undergoing cheek or nose augmentation. The patient would
requireimmediate,mediumtermandlongtermcare.Inordertogivethepatientthebest
possible care and mitigate any long term sequelae, it is important for all aesthetic
practitionerstobefamiliarwiththeprevention,identificationandmanagementofvascular
complications.
Itwassimplefortheauthortoachievetraining,completingaoneortwo-daycourseinorder
tobecertifiedcompetentintheuseofdermalfillers.However,initialtraininghadlittlefocus
onthepreventionandmanagementofdermalfillercomplications;indeedafterreviewingthe
coursematerial,itseemstohavenomentionofitwhatsoever.Itthereforeseemsthatmany
courses designed to offer practitioners a new scope of practice are leaving them poorly
equippedtodealwithanypotentialseriouscomplications,ofwhichnecrosisandblindness
arethemostworrying.
TheauthorisgravelyconcernedaboutthesafetyofmedicalaestheticsintheUK.Shechose
thetopicofcomplicationstodemonstratetherisksassociatedwithvariousproceduresand
17
also in the hope that it might be taken up as a guide for practitioners to mitigate any
unintentionalharmtheymaydotheirpatients.
InlightofthegovernmentstotalfailuretotakeonboardthefindingsoftheKeoghReport,
thereisalongwaytogobeforeaseriousregulatoryframeworkcanbeestablished.Agreat
dealofrelevantregulationinourindustrywasdirectlyorindirectlyinfluencedbyEurope,and
ourrecentbreakwiththeEUcanonlyfurtherharmtheprospectsforasensibleregulatory
regimeinthenearfuture.
Itistheunfortunatecasethatactionfromgovernmentsandregulatorsdoesnotcomefrom
carefulplanningandconsiderationbutasaknee-jerk reactiontopublicandmediaoutcry
arisingfromseriousharmbeingdonetopatients.
Mostoftheworsecasesinthisresearchcomefromoverseasbutbeassuredthatsimilarly
appallingadverseeventswillstartappearingintheUKbeforelong.Itwillnotbethewealthy
orenlightenedpatientswhowillsufferbutthepatientsoflimitedmeansandinsight,whoare
arguablythosewhoneedprotectingthemost.
Askyourselfthisquestion;whatamIdoing,asathought-leaderandexpert inthisfieldto
mitigate risk? Similarly ask what am I doing to spread ideas that might protect both
practitionersandpatientsfromharm?
Illegalpractice isone issuebutsubstandardpracticemustalsobeaddressedasagrowing
problem.PerfectlygooddentistsandGPsare‘tryingtheirhand’atfacialaestheticsandare
stillverycapableofcausingharm,justassomeonewhopicksupaneedleforthefirsttimeis
capableofcausingharm.Theonlydifferencebeingthatpatientswillhavenoreasonnotto
trusttheabsoluteclinicalintegrityofamedicalpractitioner.
Everydaytheauthorseescomplicationsfromlegalandillegalpractitioners,theseverityand
frequencyofwhichareincreasingmonthonmonth.Oneofthemostharrowingfactsisthat
patients come to the author’s practice because their treating practitioners have often
disappeared,cannotbetracedorevendenyhavingbeenthetreatingpractitioneratall!
18
Theauthorencouragespractitionerstotakeuptheadvicefoundinthisresearchanduseitto
make sure they are offering patients a safe and effective service in a consistent and
appropriateway.Evidencebasedguidelineshavebeenproducedinsuchaformatthatthey
cansimplybetakenandputinsurgeriesforreferenceandguidanceinanincident.
Theauthoralsoencouragesreadersofthisresearchtojoinherincampaigningforasafeand
regulatedindustrybecausethepositionasitcurrentlystandsiswoefullyinadequate.
Thosewhosayfacialaestheticsisaprofitableenterpriseeitherlackcommercialinsightorlack
personalintegrity.Aestheticsisaboutthesubtlepursuitofperfectionandstrongclinicalskills
and judgement. Those undertaking it in the hope of making money should be actively
discouraged,fortheyrepresentthegreatestrisk.
ExampleClinicalScenario
Apatientattendsanaestheticpracticeseekingrejuvenationofthemidface,complainingof
volumelossandagedappearanceofthenasolabialfolds.Hyaluronicacidfillertreatmentis
suggestedandcommenced.Duringtreatmentthepatientcomplainsofpain.Immediately
after treatment the patient complains of further pain and a violaceous patch to the left
nasolabialfoldandtheleftsideofthenoseisobserved.
Immediateandlongtermmanagementstrategiesmustbeinvokedtoappropriatelymanage
thispatient.Whataretheeffectivewaystomanagethissituation?Couldanythinghavebeen
donetopreventthiscomplication?
AimsandObjectives
Theaimsof thisdissertationare toascertain thecurrent treatment recommendations for
casesofvascularcompromiseandlossofvisionasaresultofdermalfillerinjectionandto
searchthecurrentliteratureforcasesrelatingtovascularcompromiseandblindnessafter
19
treatment of the midface. A particular interest was the treatment guidelines for the
managementofbothvascularcompromiseandblindness,andwhethertherehavebeenany
changes in the commonlyacceptedprotocols todealwithvascular compromise since the
adventoftheuseofhyaluronidasein2007.Furthermore,sincethereisverylittleevidence
thatvisuallossasaresultofdermalfillertreatmentcanbeeffectivelytreatedintheclinic
setting,orindeedeveninthehospitalsetting,researchwasconductedtoassesswhetherany
useful treatment modalities for vision loss from HA fillers exist or attempts have been
documented.
The goal of this reviewwas to evaluate and construct evidence based guidelines for the
managementofvascularcomplications–bothimpendingskinnecrosisandblindness–after
treatmentwithhyaluronicacidfillersintheaestheticpracticesettingintheUK.Thehopeis
thattheywillbesimpleandaccessibletopractitioners,assisttheminthecasemanagement
andreduceasmallportionofthestressfeltbypatientandpractitionerinsuchapotentially
disastrouseventandassistinthemanagementoftheacutelycompromisedpatient.
Theaimsweredefinedasfollows;
• Toevaluatethehighriskareasfortreatmentofthemidface
• Toevaluatetheriskposedbyhyaluronicacidfillers
• Toassesstheoutcomeofinterventionstrategiesintheliterature
• Toformulateevidencebasedpreventionprotocolsforaestheticpractitionersinthe
UK
• Toformulateevidencebasedmanagementprotocolsforvascularocclusionthatcan
beeffectivelyusedinthepracticesetting
• TomakerecommendationsforimprovementintheUKpracticesetting
20
Methods
Thisisareviewoftherelevantliteratureinordertodrawconclusionsaboutcurrentaesthetic
practice,theriskstovasculatureandsightassociatedwithHAfillertreatmentofthemidface
andtoconstructevidencebasedguidelinesforthemanagementofsuchcomplications.The
complicationrateandtheirmanagementisanareathatisgainingsignificantinterestinthe
field of aesthetic medicine due to a steep increase in the number of patients treated.
Unfortunately,duetothelooselegislativeframeworkintheUKrelatingtotheuseofdermal
fillers,patientsareatasignificantriskofharm(Keoghetal.,2013).
AcomprehensiveMEDLINE,CochranedatabaseandPubMedelectronicdatabasesearchwas
performedtoincludejournalspublishedbetween2011and2016,inafive-yeardaterange.
AestheticMedicineisafastevolvingfieldandassuchanarrowdaterangehasbeenselected
forthisreason;HAfillerswerenotcommonlyuseduntil2003andtherapidincreaseintheir
usewasnotuntillater.Ozturketalconductedastudyondermalfillercomplicationswith
dateranges1991-2012,soa laterdaterangewouldserveasa fittingdevelopment,albeit
limitedtoHAfiller.Noempiricalresearchwasconductedforthefindingsinthisreview.
The initial broad search inPubMed (carriedout01Feb2016) includedkeywords “dermal
filler”and“complications”andresultedin51results.Abriefassessmentofthesearchresults
revealedthatmanyofthepapersincludeddifferenttypesofdermalfillers,butalsotechnique
specificpapers.
Thecriteriaforinclusionasakeypaperwasforprimarydatatobepublished,forexample
clinicaltrialsorcasestudies.Secondaryreviewsoftheliteraturewereexcluded.Aesthetic
medicine has a notoriously weak yet fast growing evidence base due to the worldwide
scientificandfinancialinterestinit.
The BestBETsmethodologywas used in order to formulate a three-part question for the
purposes of literature research. This was to include the target population, the possible
intervention,andtheoutcome.
21
ThefollowingBETwasdevised;
In[dermalfillertreatment]which[managementstrategies]resultin[improvedoutcomefor
vascularcomplications]
Thiswasthenrevisedtoamorefocusedtargetpopulationandintervention,asfollows;
In[patientsseekingrejuvenationofthemidface]howare[VASCULARcomplicationsmanaged]
aftertreatmentwith[hyaluronicaciddermalfillers]
Theoriginalsearchofdermalfillercomplicationswasrefinedto includethemidface. Ina
searchofMeSHterms,thisincludedthecheek,teartroughandnoseandwasdefinedas“the
portionofthefacecomprisingthenasal,maxillaryandzygomaticbonesandthesofttissues
covering these bones” (Mosby, 2013). Further narrowing of the search criteria was
performedtofocusonHyaluronicAcidfillersonly,astheseconstituteover75%ofallfillers
usedinmodernaestheticpractice(Abduljabbar&Basendwh,2016).
Only English journals were considered for the literature review. In order to formulate
evidencebasedguidelinesforthemanagementofvascularcomplicationsandtoassessthe
effectofdermalfillersandtheirriskinaestheticpractice,studiesnotconductedonhumans
wereexcluded.Theareaofinterestinthemidfaceistheanastomosisofthenasalvasculature
withtheocularvasculature,andthisisnotrepresentedinananimalmodel.Cadaverstudies
have been evaluated for the embolic channel, however these do not rovide useful
interventionsthatcanbeutilisedonapatient(Tansatitetal.,2015).Duetothenatureofthe
intervention, itwasthoughtunlikelyforrandomisedcontrolledtrialstobepublishedfora
humanpopulation.
Thesearchtermswererefinedasfollows:
(midface[AllFields]OR“midface”[AllFields]OR“mid-face”[AllFields]ORcheek[MeSHTerms]
ORcheek[AllFields]OR“teartrough”[AllFields]OR“nose”[AllFields])
AND
22
("hyaluronic acid"[MeSH Terms] OR ("hyaluronic"[All Fields] AND "acid"[All Fields]) OR
"hyaluronicacid"[AllFields])
AND
(“complication”[All Fields] OR “complications”[All Fields]OR “necrosis”[All Fields] OR
“blind*”[AllFields]OR“ocular”[AllFields]OR“vascular”[AllFields])
AND
("2011/05/02"[PDat]:"2016/05/02"[PDat]AND"humans"[MeSHTerms]ANDEnglish[lang])
42articleswerefoundintheabovesearch.
Inordertoincludeandexcludepapersfromanalysis,titlesandabstractswerereadtocheck
forrelevance.Thefinaldateforinclusionandresearchwas02May2016.
Exclusionsweremadeonthefollowingbasis:
• 7paperswerenotrelatedtoaestheticmedicine
• 19 papers focused on techniques in the treatment of themidface rather than on
complicationsortheirmanagement
• 4paperswerenotrelatedtoHAfillers
• onepaperdealtwithacomplicationthatwasnotvascular
• onepaperdealtwith imagingtechniques indermal fillercomplicationsrather than
theirmanagement
Tenpaperswereincludedfromtheinitialsearch.Ofthese,twowerenotavailableonline.
Inordertomakesurethatallrelevantpapershadbeenfound,twofurthersearcheswere
conducted,featuringskinnecrosisandblindnessasthemainkeywords. Limitationswere
placedonthedaterangeonly,toincludepapersafter2011(afiveyearrange).
Furthersearch1:
“Skinnecrosisafterdermalfillerinjection”inPubmed
23
(("skin"[MeSHTerms]OR"skin"[AllFields])AND("necrosis"[MeSHTerms]OR"necrosis"[All
Fields])ANDafter[AllFields]AND("dermalfillers"[MeSHTerms]OR("dermal"[AllFields]AND
"fillers"[All Fields]) OR "dermal fillers"[All Fields] OR ("dermal"[All Fields] AND "filler"[All
Fields]) OR "dermal filler"[All Fields]) AND ("injections"[MeSH Terms] OR "injections"[All
Fields]OR"injection"[AllFields]))AND("2011/05/02"[PDAT]:"2016/05/02"[PDAT])
returned11papers
Furthersearch2:
“Blindnessafterdermalfillerinjection”inPubmed
("blindness"[MeSH Terms] OR "blindness"[All Fields]) AND after[All Fields] AND ("dermal
fillers"[MeSHTerms]OR("dermal"[AllFields]AND"fillers"[AllFields])OR"dermalfillers"[All
Fields] OR ("dermal"[All Fields] AND "filler"[All Fields]) OR "dermal filler"[All Fields]) AND
("injections"[MeSHTerms]OR"injections"[AllFields]OR"injection"[AllFields])
returned7papers
Twofurtherpaperswereidentifiedandincludedforconsiderationandevaluation.Afurther
evaluation of these papers and their references showed significant overlap in the source
papersandthesearchwasdeemedtohavebeencarriedouttoitsfullest.
Tenpaperswerefinallyselected,asfollows:
Beleznay,K.,Humphrey,S.,Carruthers,J.,&Carruthers,A.(2014).VascularCompromisefrom
SoftTissueAugmentation.JournalofClinicalandAestheticDermatology,7(9),6.
Clinicalcaseseries,Canada.
Chen, Q., Liu, Y., & Fan, D. (2016). Serious Vascular Complications after Nonsurgical
Rhinoplasty: A Case Report. Plast Reconstr Surg Glob Open, 4(4), e683.
doi:10.1097/GOX.0000000000000668
Singlecasereport,China/London.
24
Chen,Y.,Wang,W.,Li, J.,Yu,Y.,Li,L.,&Lu,N. (2014).Fundusarteryocclusioncausedby
cosmeticfacialinjections.ChinMedJ,127,1434-1437.doi:10.3760/cma.j.issn.0366-
6999.20133254
Clinicalcaseseries,China.
Kassir,R.,Kolluru,A.,&Kassir,M.(2011).Extensivenecrosisafterinjectionofhyaluronicacid
filler:casereportandreviewoftheliterature.JournalofCosmeticDermatology,10,
224-231.
Singlecasereport,USA.
Kim, S. N., Byun, D. S., Park, J. H., Han, S.W., Baik, J. S., Kim, J. Y., & Park, J. H. (2014).
Panophthalmoplegia and vision loss after cosmetic nasal dorsum injection. J Clin
Neurosci,21(4),678-680.doi:10.1016/j.jocn.2013.05.018
Singlecasereport,SouthKorea.
Kim,Y.J.,Kim,S.S.,Song,W.K.,Lee,S.Y.,&Yoon,J.S.(2011).Ocularischaemiawithhypotony
after injectionofhyaluronicacidgel.OphthalPlastReconstrSurg,27(6),e152-155.
doi:10.1097/IOP.0b013e3182078dff
Singlecasereport,SouthKorea.
Kwon, S. G., Hong, J.W., Roh, T. S., Kim, Y. S., Rah, D. K., & Kim, S. S. (2013). Ischemic
oculomotor nerve palsy and skin necrosis caused by vascular embolization after
hyaluronic acid filler injection: a case report. Ann Plast Surg, 71(4), 333-334.
doi:10.1097/SAP.0b013e31824f21da
Singlecasereport,SouthKorea.
Park,S.W.,Woo,S.J.,Park,K.H.,Huh,J.W.,Jung,C.,&Kwon,O.K.(2012).Iatrogenicretinal
arteryocclusioncausedbycosmeticfacialfillerinjections.AmJOphthalmol,154(4),
653-662e651.doi:10.1016/j.ajo.2012.04.019
Clinicalcaseseries,SouthKorea.
25
Park,T.H.,Seo,S.W.,Kim,J.K.,&Chang,C.H.(2011).Clinicalexperiencewithhyaluronic
acid-filler complications. J Plast Reconstr Aesthet Surg, 64(7), 892-896.
doi:10.1016/j.bjps.2011.01.008
Clinicalcaseseries,SouthKorea.
Sun,Z.S.,Zhu,G.Z.,Wang,H.B.,Xu,X.,Cai,B.,Zeng,L.,...Luo,S.K.(2015).ClinicalOutcomes
ofImpendingNasalSkinNecrosisRelatedtoNoseandNasolabialFoldAugmentation
with Hyaluronic Acid Fillers. Plast Reconstr Surg, 136(4), 434e-441e.
doi:10.1097/PRS.0000000000001579
Clinicalcaseseries,China.
26
Results
TenstudieswerefinallyselectedasrepresentingvascularcomplicationsfromHAfillerinthe
midface.Asthesevascularcomplicationscanresultintissuenecrosisandblindness,studies
wereselectedthatshowedinstancesofboth.
Outofthetenpapersselected,fiveinvolvedocularcomplicationsasaresultofdermalfillers
inthemidfaceandfiveinvolvedskincompromise.Aninitialreviewofthepapersinvolved
ascertainingwhich elementswould prove useful to the central question of how vascular
complicationswithHAfillersariseandhowtomanagecomplicationsinthemidface.
TheinitialreviewoftheincludedpaperscanbefoundinTable1.
AllpapersfoundwereCEBMLevel4evidence,withmostbeingsingleormultiplecasereports.
Norandomisedcontrolledtrialswerefound.Duetothefortunatelyrareincidenceofvascular
complicationsindermalfillerpractice,samplesizesaresmall,rangingfromreportsofasingle
patientinpracticeto13patientsintheocularcomplicationsgroupand28casesinafurther
study including vascular compromise affecting skin. However, with ocular complications
papersitwasfoundthatwhilstthesinglecasesfocusedonHAfillersinthemidface,papers
withmultiplecasesincludedanarrayoffillerproducts,includingautologousfatandcollagen
(Y. Chen et al., 2014; S. W. Park et al., 2012). The studies that dealt with vascular
complications involving skin only were similar, with the Beleznay et al reporting on a
combinationofHA,CaHAandparticulatefillers(Beleznayetal.,2014).Furthertothis,Park
etalexploredalltypesofcomplicationsinamuchlargersamplesizeof28treatedwithHA
filleronly,ofwhichthreewerefoundtobevascularcomplicationsaffectingtheskin(T.H.
Park,Seo,Kim,&Chang,2011).
Inorder tocloserevaluatethedatagained fromthe includedstudies, itwasnecessary to
assimilatetherelevantcasesfromeachstudyinonetable.Factorsthatwereconsideredto
beimportantfortheevaluationofthedataandsubsequentformationofanyguidelineswere
thefollowing:
27
• Injectionsite– inordertoevaluateprocedureriskandformulateprotocols forthe
preventionofvascularcomplicationsanassessmentofthesitesmostaffectedmust
bemade
• Practitioner – itwould be interesting to ascertainwhether certain professions are
associatedwithahigherriskinvascularcomplicationandoutcome
• Injectionvolume–highervolumesinjectedareassociatedwithincreasedriskofvessel
compression,increasingtheriskofischaemia
• Injectiontechnique–usinglargerboreneedlesisassociatedwithanincreasedrisk
• Symptoms–torecognisethecomplicationinrelationtotheareatreatedandwhether
thereisastandardpresentationorwarningsignforvascularcompromise
• Timetopresentation–delayingpresentationforemergencytreatmentmayresultin
worseprognosis
• Diagnosis – a diagnosis should include the vessel affected and the result of the
occlusion.Theremaybeadifferencebetweentheinterventionandoutcomeifthe
vesselsuffersfrankocclusionorcompression
• Treatment–inordertoformulateevidencebasedguidelinesforthemanagementof
vascularcompromise,anyeffectivetreatmentmodalitiesmustbeassessed
• Outcome– to judgewhetheran interventionor treatmenthasbeeneffective, the
outcomemustbemeasured
Thesubjectsisolatedfromthetenstudiesthatmeetthecriteriaofvascularcompromisewith
HAfillerinthemidfacecausingskinorocularsymptomsarepresentedinTable2.
Itsoonbecameclearthatthequalificationofthetreatingpractitionerwasrarelyrecorded.
SNKimetal,YJKimetalandKwonetalrecordedageneralsurgeonandtwophysicians(exact
specialtyundisclosed)asthetreatingpractitionersoutof12consideredocularcomplications.
Fromthegroupofvascularcomplicationsaffectingtheskinandatotalof28cases,onlytwo
paperslistedtreatmentproviders,withKassiretalstatingtheprocedurewascompletedby
aplasticsurgeonandChenetalmadeapointofadvisingthatthelicenseandregistration
statusofthepractitionerwasunknown,withoutmakingreferencetowhetheritwasinfacta
medicalprofessional.Thuswithsuchlimiteddataontheperformers,itwouldbedifficultto
28
evaluatetheriskassociatedwithdifferentprofessionalsundertakingaestheticproceduresin
themidface.Assuch,thiscriterionwasremoved.
The classification of injection techniquewas also removed. It iswell documented that a
smallergaugeneedleisthoughttocausealowerincidenceofvascularevents,withtheuse
ofacannulareducingthisriskfurther.However,fromtheevidenceinthesestudiesalone
thisconclusioncannotbedrawn,astheinjectiontechniquewasonlyrecordedtwiceforthe
ocularcomplicationspatientsbyChenetalrecordingtheuseofa30Gneedle,andBeleznay
et al recordingneedlegaugeof28G to30G in their treatment. Themethodof injection,
whetherretrogradeoranterograde,orwhethertheproductwasplacedinabolus,fanningor
otherpattern,wasnot recorded inany studies. As such itwouldbedifficult toestablish
evidencebasedguidelinesonthesecasesalone(Beleznayetal.,2014;Y.Chenetal.,2014).
Itisclearthatthereisagreatvariationontheinclusioncriteriaforeachofthestudies.Itis
also clear thata large studyof thevascular complicationsaffectingboth sightand skin in
relationtoperformingtreatmentwithHAfillersinthemidfacehasnotbeendoneonascale
that would normally net reliable data for the formulation of proper evidence based
guidelines.However,asaestheticmedicineissuchafastevolvingandasyetunexploredfield,
the data from the individual cases,when collected and evaluated togetherwill provide a
usefulframeworkfromwhichtodrawupguidelinesforcomplicationmanagement.
29
Table1:Overviewofthetenselectedpapers,theirbasicoutcomesandweaknesses
Author/Date/Category
PatientGroup Study Type(level ofevidence)
Outcomes
KeyResults StudyWeaknesses
(Y. Chen et al.,
2014)
VASCULAR /
OCULAR
Study Number:
1
13patientswith
fundus artery
occlusion from
dermal filler
treatment
resulting in loss
of vision,
women, age
range23-47
CEBMLevel4
Casereports
Retrospective
Typeoffiller
Autologousfat7/13cases
Collagen1/13cases
HA5/13cases
• Smallstudysample
• Follow up period
shortorinconsistent
• No evidence of
injection technique
orvolumeinjected
• Statisticalassessment vague
due to high number
of variables,
tabulatedonly
• No treatment or
preventionstrategies
• Vague description ofsiteas“frontal”
Symptoms
and
presentation
Immediatepartialorcompletevisionlossunilateral
Ophthalmicarteryocclusionpresentswithseverepain
Associated
Site and
vessel
Frontalarea5/13,Periocular2/13,Temple2/13,Nose4/13
Ophthalmicarteriesandbranches
Resolution
Noimprovementin10/13cases(7autologousfat,1collagenand
2HA)
NoimprovementifOphthalmicarteryocclusion
3patientsinHAfillergroupshowedimprovement
Treatment /
intervention
Nitroglycerin,massage,eyedropstolowerintraocularpressure,
aspirin,prednisolone.
Norelationshipbetweentreatmentandsymptomresolution
(S.N.Kimetal.,
2014)
VASCULAR /
OCULAR
Study number:
2
Single case,
female, age
unknown
(“healthy young
woman”
CEBMLevel4
Casereport
Retrospective
Typeoffiller HA • Singlecase• No dosage or
treatmentoutlined
• No conclusions for
treatment can be
drawn
Symptoms
and
presentation
Immediatevisionlossunilateral
Severepain
Associated
site and
vessel
Nose:non-surgicalrhinoplasty
OcclusionofOphthalmicartery
Resolution Noimprovement
30
Author/Date/Category
PatientGroup Study Type(level ofevidence)
Outcomes
KeyResults StudyWeaknesses
Treatment
and
Intervention
Nonerecorded
(Y. J.Kimetal.,
2011)
VASCULAR /
OCULAR
Study number:
3
Single case,
female,30.
CEBMLevel4
Casereport
Retrospective
Typeoffiller
HA • Singlecase• No discussion or
exploration of
treatmentmodalities
Symptoms
and
presentation
Visionlossunilateral
Painnotspecified
Associated
site and
vessel
Nose:non-surgicalrhinoplasty
Centralretinalarteryocclusion
Resolution Noimprovement
Treatment
and
Intervention
IVMethylprednisolone1gperday3D,Oralprednisolone,aspirin
100mg,dressingskinlesion
Norelationshipbetweentreatmentandresolution
(S. W. Park et
al.,2012)
VASCULAR /
OCULAR
Study number:
4
12 cases,
female, age
range18-66
CEBMLevel4
Casereports
Retrospective
Typeoffiller Autologousfat7/12cases
Collagen1/12cases
HA4/12cases
• No attempt at
treatment
• Shortfollowupof16days and under in 5
cases
• Mixed retrospective
studynotjustHAfiller
Symptoms&
presentation
Lossofvisionunilateralwithorwithoutpain.
Associated
site and
vessel
Glabella7/12cases,Nasolabialfold4/12cases,CombinedG/NL
1/12cases
Ophthalmicarteryocclusion7/12cases(1/7HA)
Centralretinalarteryocclusion2/12
Branchretinalarteryocclusion3/12(3/3HA)
Resolution IAT(urokinase)didnotimprovesymptoms
No relationship between intervention and improvement in
symptoms;NoimprovementinOAOgroupandnoimprovement
inpatientswithseverelydecreasedsight
31
Author/Date/Category
PatientGroup Study Type(level ofevidence)
Outcomes
KeyResults StudyWeaknesses
Treatment
and
intervention
Intra-arterialthrombolysis4/12cases
Anteriorchamberparacentesis4/12cases
Conservativemanagement3/12cases
Nohyaluronidase
(Kwon et al.,
2013)
VASCULAR /
OCULAR
Study number:
5
Single case,
female,20.
CEBMLevel4
Casereport
Retrospective
Typeoffiller
HA • Singlepatientstudy• Use of unlicensed
treatments
Symptoms
and
presentation
Lossofvisionunilateral,skinnecrosis
Warningsignignoredbypractitioner
Associated
site and
vessel
Nose:non-surgicalrhinoplasty
Occlusionofretinalartery
Resolution Improvedvisualacuityfrom20/70to20/30
Treatment /
intervention
Nicergorline
(Kassir et al.,
2011)
VASCULAR /
SKIN
Study number:
6
Single case,
male,52.
CEBMLevel4
Casereport
Retrospective
Typeoffiller HA • Singlepatientonly• Reference to
unrelatedstudies• Mechanism of
vascular event not
identifiedexactly• Appreciation of
common treatment
modalities which
have not been
implemented
Symptoms
and
presentation
Skinnecrosis:Latepresentationafterlargefillervolume
Associated
site and
vessel
Compressionoffacialartery,transversefacialarteryandbuccal
branchofmaxillaryartery
Resolution Healedwithscarring
32
Author/Date/Category
PatientGroup Study Type(level ofevidence)
Outcomes
KeyResults StudyWeaknesses
Treatment
and
intervention
Conservative:antibiotics,scarmanagement
(Sun et al.,
2015)
VASCULAR /
SKIN
Study number:
7
20 cases with
nose and
nasolabial fold
augmentation,
1 male (21), 19
female(21-52)
CEBMLevel4
Casereports
Retrospective
Typeoffiller HA • Dosagesnotrecorded• Experimental
medicine / herbal
remediesusedwhich
are not licensed in
US/EU
Symptoms
and
presentation
Impendingskinnecrosisofnoseandglabellartissues
Associated
site and
vessel
Nose:nonsurgicalrhinoplastyin15/20cases
NLfoldaugmentation5/20cases
Resolution 13/20caseswithfullrecovery
earlytreatmentshowedbetteroutcome
Treatment
and
intervention
Previouslyundocumentedinterventions
Treatment with hyaluronidase all except 2 late presentation
cases
(T.H.Parketal.,
2011)
28 cases, all
filler
CEBMLevel4
Typeoffiller
HA
33
Author/Date/Category
PatientGroup Study Type(level ofevidence)
Outcomes
KeyResults StudyWeaknesses
VASCULAR /
SKIN
Study number:
8
complications.
3/28 skin
necrosis
Casereports
Retrospective
Symptoms
and
presentation
Various,but3/28patientsskinnecrosis • All types of
complications
considered
• Number of vascular
complication cases
consideredlow
• Conventionaltreatment
approachesnotused
Associated
site and
vessel
Nose,nasalsidewallandmentumforsitesofnecrosis
Resolution Notdocumentedfornecrosiscases
Treatment
and
intervention
Antibioticsandsurgicalexcision
(Beleznayetal.,
2014)
VASCULAR /
SKIN
Study number:
9
12 cases with
vascular
compromise
from soft tissue
augmentation,
aged26-69
CEBMLevel4
Casereports
Retrospective
Typeoffiller Various,including5/12withHA
• Singlecentreresults• Long duration of
study(10yearspan)–
developments in
treatment and
products in this time
+++ but also shows
risk related to total
number of
treatments
Symptoms
and
presentation
Immediate (hours) or delayed (1-5) violaceous reticulated
patches
Associated
site and
vessel
Cheek,NLF,alarcrease
Resolution Complete,withoutscarringinallHAcases
Treatment
and
intervention
Warmcompress,hyaluronidase,massage,GTNpaste2%,aspirin
andprednisolone,dailyfollowup
(Q.Chen,Liu,&
Fan,2016)
VASCULAR /
SKIN
Single case,
female,32.
CEBMLevel4
Casereports
Retrospective
Typeoffiller
HA • Singlecase• No protocol that hasbeentested
• Pooroutcomes
Symptoms
and
presentation
Intensepainandskinblanching,followedbynecrosisofskin
34
Author/Date/Category
PatientGroup Study Type(level ofevidence)
Outcomes
KeyResults StudyWeaknesses
Study number:
10
Associated
site and
vessel
Nose:non-surgicalrhinoplasty • Contradiction of
commonly used
methods for
resolutionofHAfiller
complication without
significantevidence
Resolution
Healingwithscarring
Treatment
and
intervention
Surgical decompression nasal tip, suction drainage, hyperbaric
oxygen,thereafterwoundmanagement
35
Table2:casecollectionfromkeypapers
Reference Paper category Age Sex Inject.
site Injection
vol. Symptom Time to presentation Diagnosis Treatment Outcome
Y. Chen et al., 2014 ocular 44 F Nose unknown headache, vision loss immediate anterior ischaemic optic
neuropathy
nitroglycerin, digital massage, eye drops to lower intraocular
pressure, aspirin, prednisolone.
small improvement BCVA
Y. Chen et al., 2014 ocular 45 F Periocular 1.0ml vision loss although not
fully described immediate central retinal artery occlusion
reduced sight but not loss, no improvement
Y. Chen et al., 2014 ocular 25 F Frontal
area 2.1ml vision loss although not fully described immediate unknown resolution from hand movement to
improved visual acuity
Y. Chen et al., 2014 ocular 38 F upper
eyelid 0.6ml Ptosis, ophthalmoplegia,
dizzy, vomiting, vision loss
immediate ophthalmic artery occlusion no light perception - no improvement
Y. Chen et al., 2014 ocular 23 F Nose 0.9ml
Ptosis, ophthalmoplegia, dizzy, vomiting, vision
loss immediate ophthalmic artery occlusion no light perception - no
improvement
S. N. Kim et al., 2014 ocular NA F Nose unknown Ptosis, ophthalmoplegia,
vision loss right eye unknown central retinal artery occlusion High dose IV corticosteroids
no light perception - improvement in lateral eye movement but not
sight
Y. J. Kim, Kim, Song,
Lee, & Yoon, 2011
ocular 30 F Nose 0.8ml
(symptoms after
0.2ml)
toothache, headache, pain in left upper face,
vision loss left, necrosis of skin glabella and nose
immediate central retinal artery occlusion
IV methylprednisolone, aspirin 100mg, wound dressing.
no scarring, eyeball movement returned to normal, permanent
vision loss
S. W. Park et al., 2012 ocular 32 F Nasolabial
/ glabella unknown
vision loss right eye, ocular pain, ptosis,
exotropia, ophthalmoplegia, corneal
oedema
immediate ophthalmic artery occlusion intra-arterial thrombolysis no light perception - no improvement
S. W. Park et al., 2012 ocular 26 F Nasolabial unknown inferior visual field defect,
no ocular pain 2 weeks branch retinal artery occlusion not recorded visual acuity 1 at presentation, no
change (20/20)
36
Reference Paper category Age Sex Inject.
site Injection
vol. Symptom Time to presentation Diagnosis Treatment Outcome
S. W. Park et al., 2012 ocular 26 F Glabella unknown
inferior and central visual field defect, no ocular
pain 5 hours branch retinal artery
occlusion massage and anterior chamber
paracentesis visual acuity decreased from 0.7 to
0.15 over review period
S. W. Park et al., 2012 ocular 26 F Nasolabial unknown inferotemporal visual field
defect, no ocular pain 3 weeks branch retinal artery occlusion not recorded visual acuity 1 at presentation, no
change (20/20)
Kwon et al., 2013 ocular 20 F Nose unknown
partial visual disturbance right eye, orbital pain,
nausea, vomiting, headache,
blepharoptosis, diplopia, dizziness
immediate branch retinal artery occlusion
Aspirin, Nicegorline, eye drops, IV steroids, hyaluronidase for skin lesion, wound care, IV
antibiotics
visual acuity improvement from 0.3 to 0.6, partial resolution
blepharoptosis, partial resolution limitation of eyeball movement,
minimal scarring of skin
Kassir, Kolluru, &
Kassir, 2011 skin 52 M Cheek
scar 2.0ml persistent pain, slough, necrosis, poor capillary
refill 5 days
compression of facial, transverse facial, buccal branch of maxillary artery
supply areas
topical antibiotic, IM antibiotic, wound care (silicon) healing with scarring
Sun et al., 2015 skin 24 F Nose unknown Nose skin impending
necrosis 7 days compression of vessel causing ischaemia No hyaluronidase Necrosis of skin, healing with
scarring
Sun et al., 2015 skin 22 F Nose unknown Nose and NLF impending
necrosis 7 days embolism of vessel
Hyaluronidase, antibiotics, Tanshinone, Papaverine, topical
magnesium sulphate, infrared radiation, hyperbaric oxygen,
aspirin.
Necrosis of skin, healing with scarring
Sun et al., 2015 skin 24 F Nose unknown Nose skin impending
necrosis 4 days compression of vessel causing ischaemia Necrosis of skin
Sun et al., 2015 skin 28 F Nose unknown
Nose, NLF, glabella, forehead impending
necrosis 1 day embolism of vessel Necrosis of skin, healing with
scarring
Sun et al., 2015 skin 25 F Nasolabial unknown Nose, NLF, lip reticulated
erythema skin 3 hours embolism of vessel Full recovery
Sun et al., 2015 skin 25 F Nose unknown Nose reticulated
erythema skin 1 day compression of vessel causing ischaemia Full recovery
37
Reference Paper category Age Sex Inject.
site Injection
vol. Symptom Time to presentation Diagnosis Outcome
Sun et al., 2015 skin 38 F Nose unknown Nose, glabella, forehead
reticulated erythema 1 day embolism of vessel Full recovery
Sun et al., 2015 skin 24 F Nose unknown
Nose, NLF, glabella, forehead impending
necrosis 9 days embolism of vessel No hyaluronidase Necrosis of skin
Sun et al., 2015 skin 25 F Nose unknown Nose reticulated
erythema skin 2 days embolism of vessel
Hyaluronidase, antibiotics, Tanshinone, Papaverine, topical
magnesium sulphate, infrared radiation, hyperbaric oxygen,
aspirin.
Full recovery
Sun et al., 2015 skin 52 F Nasolabial unknown Nose, NLF reticulated
erythema 1 hour embolism of vessel Full recovery
Sun et al., 2015 skin 21 M Nose unknown Nose reticulated
erythema skin, pustules 2 days compression of vessel causing ischaemia Full recovery
Sun et al., 2015 skin 22 F Nose and
nasolabial unknown Nose, NLF, glabella, forehead reticulated
erythema immediate embolism of vessel Full recovery
Sun et al., 2015 skin 31 F Nose unknown Nose reticulated
erythema skin, pustules 2 days compression of vessel causing ischaemia Full recovery
Sun et al., 2015 skin 21 F Nose unknown Nose, glabella, forehead
reticulated erythema 5 days embolism of vessel Full recovery
Sun et al., 2015 skin 35 F Nasolabial unknown Nose, NLF reticulated
erythema, pustules 1 day embolism of vessel Full recovery
Sun et al., 2015 skin 39 F Nasolabial unknown Nose, NLF, lip reticulated
erythema skin 3 days embolism of vessel Full recovery
Sun et al., 2015 skin 33 F Nasolabial unknown Nose, NLF, lip reticulated
erythema skin 2 days embolism of vessel Full recovery
38
Reference Paper category Age Sex Inject.
site Injection
vol. Symptom Time to presentation Diagnosis Outcome
Sun et al., 2015 skin 34 F Nose unknown Nose skin impending
necrosis 5 days compression of vessel causing ischaemia Necrosis of skin
Sun et al., 2015 skin 26 F Nose unknown Nose skin impending
necrosis 2.5 days compression of vessel causing ischaemia
Necrosis of skin, healing with scarring
Sun et al., 2015 skin 37 F Nose unknown Nose, NLF reticulated
erythema 1 day embolism of vessel Full recovery
T. H. Park, Seo, Kim, & Chang, 2011
skin ? ? Nose unknown tissue necrosis 3 months not made Oral antibiotics not described
T. H. Park, Seo, Kim, & Chang, 2011
skin ? ? Nasolabial unknown alar necrosis 7 days not made Oral antibiotics not described
Beleznay, Humphrey, Carruthers,
& Carruthers,
2014
skin 69 ?
Pre-jowl sulcus, lip
corner, right
lateral cheek
1.0ml violaceous reticulated patch right NLF, pain
right upper lip, numbness, ecchymosis
2 days vascular compromise -
presumed compression of vessel or embolisation (if
immediate) Massage, warm compress complete resolution
Beleznay, Humphrey, Carruthers,
& Carruthers,
2014
skin 65 ? Nasolabial 1.0ml violaceous reticulated patch left NLF, pain 2 hours
vascular compromise - presumed compression of vessel or embolisation (if
immediate) Massage, hyaluronidase complete resolution
Beleznay, Humphrey, Carruthers,
& Carruthers,
2014
skin 31 ? Nasojugal 2.0ml violaceous reticulated patch right cheek, pain 5 days
vascular compromise - presumed compression of vessel or embolisation (if
immediate)
Massage, warm compress, hyaluronidase complete resolution
Beleznay, Humphrey, Carruthers,
& Carruthers,
2014
skin 50 ? Nasolabial 2.0ml violaceous reticulated patch left NLF and alar
crease, pain 1 day
vascular compromise - presumed compression of vessel or embolisation (if
immediate)
Massage, warm compress, hyaluronidase, prednisolone complete resolution
39
Reference Paper category Age Sex Inject.
site Injection
vol. Symptom Time to presentation Diagnosis Treatment Outcome
Beleznay, Humphrey, Carruthers,
& Carruthers,
2014
skin 58 ? Nasolabial and cheek
0.5ml HA but also
3ml CaHA to NLF
and cheek
violaceous reticulated patch and blanching left
NLF immediate
vascular compromise - presumed compression of vessel or embolisation (if
immediate)
Massage, warm compress, nitroglycerin paste,
hyaluronidase, prednisolone, aspirin
complete resolution
Q. Chen, Liu, & Fan,
2016 skin 32 F Nose unknown
grey discolouration forehead, nasal sidewalls
(bilat), nasal triangles (bilat)
2 days vascular compromise surgical decompression, suction
drainage, hyperbaric oxygen, wound management
healing with extensive scarring
40
Discussion
Vascular complications are fortunately rare, and occur at a rate of around 0.001% of all
performeddermalfillertreatments(Cohenetal.,2015).
Beleznayetalreportariskrateofaround0.05%forvascularcomplicationsintheirpractice,
whichmaybeduetoahighercomplexityofproceduresundertaken(Beleznayetal.,2014).
However,whentheadverseeventrequiresreferral toacentresuchas thatbyParketal,
around10%ofallthosereferredarenecrosiscases(T.H.Parketal.,2011).
Whilstarterialocclusionisofsuddenonset,oftenaccompaniedbyseverepainandblanching
of the affected skin, venousobstruction can result in delayedpresentationof violaceous,
reticulatedpatchesontheareaoftissuedrainedbytheobstructedvein(Daines&Williams,
2013).Visionlossisoftentheformer,whereanembolusoccludesanocularsupplyartery
leadingto ischaemiaoftheorbitandassociatedmuscles,whereascomplications involving
theskincanbeeitherduetofillerembolusaffectinganartery,orcompressionofanarteryor
veincausingischaemiaofthetissuewithdelayedonset.Pressurenecrosiscanresultfrom
thesheervolumeoffillerintroducedintothetargettissue,suchasthecasebyKassiretal
where2.0mlofHAfiller injectedintoanalreadypoorlyperfusedscarresultedinnecrosis,
furtherexacerbatedbydelayedexpansiontoHAfillerasaresultofitshydrophilicproperties
(Grunebaumetal.,2009;Kassiretal.,2011;Sunetal.,2015).Thetimetopresentationisa
useful indicatorof thenatureof thevascularaccident– inthestudybyBeleznayetal for
example,threeoutofthefiveHAfillercasespresentedat24hoursormore,suggestingthat
thecomplicationisduetocompressionorembolizationofavein,whichisofsloweronset
(Beleznayetal.,2014).Sunetalspecifiedthedifferencebetweenvesselembolizationand
compression in their findings, however how these were distinguished has not been
elaboratedupon–itisassumedthatimagingprocedureswereusedtoidentifythecauseof
thecompromise inadditiontotheareasuppliedbeingdemarcatedbytheareawhichthe
vesselshouldsupply.Asfarastheauthorisaware,thereisnohardandfastwaytoeasily
distinguish between embolization and compression at the point of diagnosis. However,
compression may be easier to resolve with liberal injection of hyaluronidase to relieve
pressureonthevesselinquestion.
41
Allstudiesexaminedwereretrospectivecasereviews.Chenetal,ParketalandBeleznayet
al all reported larger sample size in the papers analysed, with 13, 12 and 12 cases,
respectively. However, these studies all considered various types of fillers, including
autologousfat,collagen,unidentifiedparticulatefillersandHA(Beleznayetal.,2014;Y.Chen
etal.,2014;S.W.Parketal.,2012).TheonlystudytofocusentirelyonHAfillercomplications
wasbySunetal,with20subjectsrecordedinaretrospectivecaseseriesexaminingimpending
necrosisoffacialtissues(Sunetal.,2015).Conclusionsrelevanttotheclinicalquestioncould
onlybedrawnfromcasesthatweretreatedwithHAfillersinthemidface,whichhavebeen
presentedinTable2andwhichwillbereferredtohenceforth.Therewereatotalof12ocular
complicationsand29skincomplicationsassociatedwithHAfillertreatmentinthetenstudies
includedforanalysis.Allcaseswereaccountedforfrompresentationtooutcome.
Fillertypes
Therelevanceofdifferentfillertypesisintheircapacitytooccludevesselsoftheeye.The
particle sizeofHA fillers is around400µm,which isof sufficient size toblock the smaller
vesselssuchastheretinalarteriesmeasuringaround160µmdiameterbutnotlargeenough
to block the main ophthalmic arteries at around 2mm diameter, unless a large bolus is
injectedatspeed. Incasesofvision lossresultingfromHAfiller,theretinalarteryand its
brancheswereinvolvedin64%wheretheaffectedvesselwasdocumented.Chenetaland
Park et al compared this with the occlusion of vessels by other types of filler, especially
autologousfatwheretheparticlesizeisextremelyvariableandoftenlarger,andnotedthat
autologousfatoccludedlargervesselssuchastheophthalmicarterymorefrequently. HA
fillerocclusionoftheocularvesselsisthereforeassociatedwithbetteroutcomeastheinitial
blockageofthevesselislikelytobeasmallerone,supplyinglesstissue(Y.Chenetal.,2014;
S.W.Parketal.,2012).
ThestudiesbyParketalandChenetalconsiderverysimilarcasesofocularcomplications.
Whencomparing the two, the incidentsofocular complications for thedifferent typesof
fillers resulted insimilarpercentages for incidentsofHAversusAutologous fat fillers. HA
42
fillersrepresentfiveandfourcasesineachseries,respectively.Autologousfataccountsfor
sevencasesoutofthe13casesbyChenandsevenoutof12casesbyPark(Y.Chenetal.,
2014;S.W.Parketal.,2012).Autologousfatthereforeaccountsformorecasesofvisionloss
inbothreferralcentres,eventhoughthenumberofHAfillertreatmentsfaroutweighthe
numberofautologousfattreatments.Comparingthiswiththe2014ISAPSdata,therewere
965,727 treatments with autologous fat versus 2,690,633 with HA filler. However, the
instancesofreportedvisionlossforautologousfatinthestudiesbyChenetalandParketal
arehigherthanHAfillerbyafactorofalmost2.Ifthepatientgroupinthesetwostudiescan
beusedasarepresentativeaverageofwhatreallyoccursintermsofcomplicationrate,then
extrapolating these values relative to the ISAPS data means that autologous fat is
approximatelyfivetimesmorelikelytocauseblindnessthanHAfillers(ISAPS,2016).
Highriskareasandtheeffectonvessels
Theresultsgainedfromtheincludedpapersgaveaveryclearindicationofthehighriskareas
fortreatmentwithdermalfillers.SunetalintheirstudyfocusingonHAfillers,75%ofthe
cases of skin necrosiswere due to treatment of the nose and the remaining 25%due to
treatmentofthenasolabialarea,andallofthesubjectspresentedwithskinnecrosisofthe
nose,regardlessofwhereinthemidfacethetreatmentsitehadbeen(Sunetal.,2015).In
thecollated12ocularcomplicationscasesfromallpapersconsidered,thenosewastreated
in42%ofcases,withnasolabialandglabella treatmentbothbeingresponsible for25%of
vision loss cases each. A staggering 59% of the collated 29 skin necrosis cases from all
consideredkeypapershadhadtreatmentofthenose,followedbytreatmentofnasolabial
foldswith34%ofcases.Thisclearlyoutlinesthenose,nasolabialandglabellaastopthree
sitestoresultinvascularcomplications.
Noseaugmentation is thereforeclassedas thehighest riskprocedure in themidface. It is
perhapsnocoincidencethatseveralofthepapershail fromChinaandKorea,wherenose
reshaping in order to westernise the patient’s appearance is a procedure that is on the
increase(S.W.Parketal.,2012).Whilstthenoseisregardedasatreatmentareathatonly
experiencedpractitionersshouldapproach,nasolabialfoldaugmentationisoneofthefirst
43
treatmentstaughtonmanyone-daycourses,especiallyintheUK–thisposesaseriousrisk
for patients and aworry to experienced practitioners receiving referrals for complication
management.
It has already been established that treatment of the nose is usually due to a desire to
reshape, rather than for rejuvenation purposes. Considering the incidence of ocular
complicationshowever,theredoesseemtobeaweightingtowardscomplicationsinyounger
individuals–allsubjectsdocumentedwithvisionlossasaresultofHAfillertreatmentwere
undertheageof45.Thismaynotbeacoincidence.Itisthoughtthatyoungerindividuals
may be at increased risk of ocular complications due to an increased number of patent
cutaneous arterial anastomoses. In a cadaver study by Tansatit et al, injection of facial
arterieswithdyeledtothedyebeingextrudedintotheglobe.Inmuchthesameway,filler
can causeocular complications. Large facial cutaneousarteries canbeeasily accidentally
cannulatedinfillertreatment.Whenpressureisexertedontheplunger,andthisissufficient
tocauseretrogradeflow,theophthalmicarterycanbereachedfromaugmentationsitesin
noseandnasolabialareas(Y.Chenetal.,2014;Tansatitetal.,2015).
Chenetalreportedthatinadditiontolossofvisioninonecase,thepatientexhibitedsigns
ofacerebralaccidentasaresultoffillertreatment,evidencedbymagneticresonanceimaging
(Y.Chenetal.,2014).Cerebralinfarctionhaspreviouslybeendocumentedinafurtherstudy
byHsiehetal,thoughttobecausedbypressureexceedingsystoliconinjectioncausingreflux
offillerembolustothemiddlecerebralartery.Thiscanhavetrulydevastatingconsequences
(Hsiehetal.,2015).Excesspressureoninjectionisthereforeasignificantriskfactorindermal
fillertreatment.
Thefirstcaseofpermanentvision lossasaresultofHAfillernon-surgicalrhinoplastywas
2011byKimetal(Y.J.Kimetal.,2011).By2013,thetotalnumberofcasesofvisionloss
associatedwithalltypesoffillertreatmentwasaround60.Therearearangeofwarningsigns
forvisionlossasaresultoffillertreatment,withocularpainandptosispresentinginalarge
number of cases. Associated loss of movement of the eye is due to occlusion of the
ophthalmicarteryandlackofperfusiontotheoculomotornerves(Y.Chenetal.,2014).In
thestudybyKwon,thepractitionerdocumenteda“bursting”sensationafterinitialinjection,
44
whichwasignored,shortlyafterwhichthepatientreportedlossofvision(Kwonetal.,2013).
Othersymptomsincludedtoothacheandpainononesideoftheface(Y.J.Kimetal.,2011).
Parketal linkedocclusionoftheophthalmicarterytoseverepain,whereascentralretinal
arteryocclusionandbranchretinalarteryocclusioncanpresentwithoutpain(S.W.Parket
al.,2012).Itappearsthatthesmallerthevesseloccluded,thelesspainresultsduetoless
tissuesupplied.ExaminationoftheretinainthestudiesbyKimetalandChenetalrevealed
cherry-redspotsassociatedwithophthalmicarteryocclusionandthinningofthefundus.The
outcomeforthesepatientswasnoimprovementinlightperception,withophthalmicartery
occlusionhavingtheworstprognosisandmostseveresymptoms(Y.Chenetal.,2014;Y.J.
Kimetal.,2011).
Interventions
Skinischaemia/necrosis
Earlyinterventionwithimpendingskinnecrosisislinkedtoafavourableoutcome;Sunetal
claim that when treatment is started within two days, the patient suffers a significantly
reduced incidenceof scar formation (Sunetal.,2015). Outof the29casesof impending
necrosisoftheskinanalysedfromtheliterature,17weretreatedintwodaysorless.Ofthese
17,two,or12%,sufferedadegreeofscarringafterhealing.Thiscanbecomparedtothose
thatreceivedtreatmentaftertwodays,afterwhich58%ofpatientssufferedscarringafter
vascularcompromise.Althoughtwodaysisanarbitrarytimespan,thisdoesappeartomark
the divide between excellent and mediocre outcome for resolution of necrosis of skin.
However,thecorrelationbetweentreatmenttimeandoutcomeneedstobemoreextensively
studiedtodemarcateoptimumoutcomeversustimetopresentation(Sunetal.,2015).
Astudyconductedin2007byHirschetalwasthefirstcaseinliteraturethatHyaluronidase
wasusedtosuccessfullymanagevascularcomplicationfromHAfillerandithasbeenusedto
routinelytreatvascularcomplicationsendangeringskinsince(Hirschetal.,2007).Ithasbeen
showneffectiveinthemanagementofimpendingskinnecrosis,butnotinthemanagement
ofocularcomplications.Outofthe29casesaffectingtheskin,22(76%)weretreatedwith
45
hyaluronidase,ofwhich17(59%)hadagoodoutcome.Theremainingfivepatientswhodid
not have a good outcome with necrosis and scarring a consequence of the vascular
complication,presentedat1day,2.5days,4days,5daysand7daysposttreatment.Sunet
aladvisedthatearlytreatmentisessential,anddescribethisaswithintwodaysoftreatment.
Considering the results from the collected studies, there is a significant improvement in
outcomewhentreatmentisstartedwithhyaluronidaseintheearlypresenter. Comparing
thiswiththeindividualsinthecasegroupsthatdidnotreceivehyaluronidase,onlyoneout
of7patientshadagoodoutcomewithoutscarring.Thelackofuseofhyaluronidaseseems
tobeinpartduetothelatepresentationofthepatientsof5dayspost-operativelyormore.
Atthisstagetissuenecrosisismostlikelyatanadvancedstageandtheprognosisisseverely
compromised (Beleznayetal., 2014; Sunetal., 2015). It can thereforebeconcluded the
incidence of scarring and long term sequelae can be reduced when hyaluronidase is
administeredimmediatelyorattheveryleastwithintwodaysoftheproceduretoimprove
outcome.
Intheircasereportseriesinvestigatingimpendingskinnecrosis,Sunetalexploredtheuseof
Tanshinone and Papaverine alongside commonly used methods to improve vascular
compromise(Sunetal.,2015).TanshinoneisaChineseherbalremedyderivedfromSalvia
miltiorrhiza,whichhasbeenusedtotreatavarietyofcerebralandcardiovasculardisorders.
Tangetalinvestigateditseffectivenessinreperfusionofcerebralischaemiainrats,showing
thatithasaprotectiveeffectandimprovescerebralbloodflowintheanimalmodel(Tanget
al.,2014).However,itsuseinpreventionofnecrosisasaresultofHAfillerhasonlybeen
documentedinthestudybySunetal.Papaverineisanopiumalkaloidantispasmodicand
cerebralandcoronaryvasodilator(Fusi,Manetti,Durante,Sgaragli,&Saponara,2016),and
waspresumablybySunetaltoimproveperfusionoftheischaemictissue.Unfortunately,as
thevariableshavenotbeenseparatedandallinterventivemethodsincludinghyaluronidase,
papaverineandtanshinonewereallcombinedonthesamesubjects,itisdifficulttoascertain
whether the two novel interventions had any significant effect on the outcome for the
patient.Intheabsenceofdatatothecontrary,andgiventhattheoutcomesinthestudyby
Beleznay et al were similarly good with complete resolution of the skin lesions without
scarring, itmustbeassumedthattanshinoneandpapaverinedidnotsignificantly improve
46
outcomeforthepatientsaboveandbeyondwhatwouldhavebeenobtainedwiththeuseof
hyaluronidaseandstandardtreatmentmethodsalone(Beleznayetal.,2014;Sunetal.,2015).
Intheirsinglecasereport,Chenetalillustratedhowover-aggressivemanagementwithout
implementingcommonlyusedprotocolsforthemanagementofvascularcomplicationscan
leadtopooroutcome.Thepatientpresentedmoderatelylate,at2days,andinsteadofusing
hyaluronidase,massageoranyvasodilatingtreatments,surgicaldecompressionandsuction
drainagewasperformed.Itisdebateablewhetherthissurgicalinterventionmayhavecaused
furthertissueswelling,leadingtoexacerbationoftheareaofischaemia.Theoutcomeforthe
patientwasextensivescarringtothenoseandglabella.Chenetalconsidered,muchtothe
patient’sdetriment,thattheuseofhyaluronidaseatdaytwowouldnotbeofuse.Thisis
contrarytotheresultsachievedbyBeleznayetalwhereinthecaseoftwoindividualswitha
similarpatternofischaemiaat2daysand5days,completeresolutionwasachievedwiththe
useofhyaluronidaseandadefinedmanagementprotocol(Beleznayetal.,2014;Q.Chenet
al.,2016).
Incaseswhereischaemiahasledtonecrosisofskin,theoutcomecanbepositivelyinfluenced
byappropriatewoundmanagement.Siliconedressingshavebeenshowntoimprovehealing
of the necrotic tissue and wounds (Lansdown & Williams, 2007). Furthermore, where
accessible,hyperbaricoxygentreatmentcanimproveskinresolution.Itisoftenusedingraft
healing after cancer or trauma surgery and for the treatment of nonhealing wounds by
increasingendothelialcells,fibroblasts,keratinocytemigrationanddifferentiation,butcan
poseanexcesscost,riskandinconvenience(Grunebaumetal.,2009;Kassiretal.,2011).With
smallareasofnecrosisitisthoughtthatitposesnoadditionalbenefit.Indeed,thestudyby
Sun et al failed to provide any concrete evidence that recovery was better than the
straightforwardprotocolappliedbyBeleznayetalintermsofhealingwithoutscarring.Whilst
itmaysupplysomebenefitasanadjunctivetreatment,usinghyperbaricoxygenasafirstline
treatment in the case reportedbyChenetal showedworseoutcome. Itwould certainly
representasignificantfinancialinvolvement;lowcostinterventionsincludinghyaluronidase,
massageandaspirinhavebeenshowntohavemorefavourableoutcomeswhicharemore
easilyimplementable(Beleznayetal.,2014;Q.Chenetal.,2016;Sunetal.,2015).
47
Visionloss/disturbance
UrokinasewasadministeredtoonepatientinthecaseseriesbyParketalinordertoachieve
intra-arterialthrombolysisandreduceclotformationwithinthevesselwalls(S.W.Parketal.,
2012).Inmuchthesamewayaswhenadministeredforthetreatmentofischaemicstrokes
withinthefirstfewhoursafterthecardiovascularaccident,itwashopedthatreperfusionof
the vessel would be achieved, leading to the patient regaining some or all of their sight
(Wardlaw,Murray,Berge,&delZoppo,2014).However,therewasnoimprovementinlight
perception and no improvement in sight for this patient, leading to the conclusion that
urokinasewas ineffective. Indeed, itwouldbeunlikely to remedyaHA filler embolus as
urokinasewouldnotbeabletodegradehyaluronicacid.Inordertorulethisinterventionout
fully, however, further interventions with intra-arterial thrombolysis would need to be
documented.
NicergolinewasusedbyKwonetaltomanageacaseofbranchretinalarteryocclusionand
subsequent decreased sight resulting from HA filler treatment. Nicergoline was used in
conjunctionwithaspirin,IVsteroidsandhyaluronidasefortheskinlesion(Kwonetal.,2013).
Itisadrugthatcandecreasevascularresistanceandhasbeenusedtotreatcerebralinfarction
andacuteandchronicperipheralcirculationdisorders,amongstothers.Itactstopromote
cerebralmetabolicactivity,resultinginincreasedmetabolismofglucoseandoxygen,andis
antithromboticbyinhibitingplateletphospholipaseandinterferingwithplateletaggregation.
The patient treated with the combination of steroids, aspirin and nicergoline showed
improvement in sight and partial resolution of blepharoptosis and eyeball movement.
Unfortunately,however,theEuropeanmedicinesagencyhaverestrictedtheuseofallergot
derivatives including nicergoline due to concerns about the safety profile of these drugs.
Whilstthetreatmentcombinationmayhavebeeneffectiveforthispatient,afurtherlarger
case serieswouldneed toutilise this treatment;even if thisproved successful theuseof
nicergolinewouldberestrictedintheUK(Kwonetal.,2013;Saletu,Garg,&Shoeb,2014).
Chenetalpresentedacaseseriesinwhichtreatmentwithnitroglycerin,digitalmassage,eye
drops,aspirinandprednisolonewasused.Inonecase,apatientshowedanimprovementin
48
bestcorrectedvisualacuity(BCVA).Asecondcaseinthisseriesimprovedinsightfromhand
movementto“improvedvisualacuity”.Itisnotspecifiedwhethernitroglycerinwasapplied
topically, as it isoftenwith skin ischaemia,or systemically. Nitroglycerin relaxesvascular
smoothmuscleleadingtovasodilation,whichisthoughttoassistinreperfusionofthetissues
affectedbyvascularcompromise(Kukovetz,Holzmann,&Romanin,1987).Theroleofaspirin
isantiplatelet,whereaseyemassageisthoughttohelpdislodgetheclotandprednisoloneto
reduceinflammation.Unfortunately,theimprovementmayalsobeacoincidence–thecase
wasdiagnosedasanteriorischaemicopticneuropathy,andthereforemuchlessseverethan
fullocclusionoftheophthalmicartery.Forthesecondcase,thedataondiagnosisismissing
andtheoutcomeisrathervague,soconclusionsabouttheeffectivenessoftheintervention
cannotbedrawn.TheremainingcasesinthesameseriesbyChenetalthatweretreated
with the sameprotocolafterdiagnosiswithophthalmicarteryocclusionor central retinal
artery occlusion showed no improvement in symptoms (Y. Chen et al., 2014). Indeed,
consideringthethreecasesofcentralretinalarteryocclusionandthreecasesofophthalmic
arteryocclusion fromthestudiesconsidered,noneof the interventionsbroughtaboutan
improvement.
Aninterventionbestperformedinahospitalsettingbyaspecialistophthalmologist,Parket
aldescribedtheuseofanteriorchamberparacentesistotreatvisionlossasaresultoffiller
treatment(S.W.Parketal.,2012).Anteriorchamberparacentesishassincebeenusedina
study by Rajabi et al to alleviate central retinal artery occlusion in a patient who had
undergoneorbitaltumourresection.However,inthisstudy,theinterventionwasperformed
immediately,inadditiontosystemicmannitolandintravenousacetazolamidetherapy,and
partial resolution to 1m counting finger was achieved (Rajabi, Naderan, Mohammadi, &
Rajabi,2015). Chenetalwereunable toachieve improvement for theirpatientwhohad
visionlossasaresultofdermalfillers.Thedifferencebetweenthetwocasesmaybedueto
thetimetotreatment–Rajabi’spatientwastreatedimmediately(indeedwasalreadyina
hospitalsetting),whereasChen’spatienthadafive-hourdelay,butalsoduetotheoccluding
material– fillerversusbloodclot. It isknownthat retinal ischaemiabecomes irreversible
after90minutes,soquickactiononbehalfofthereferringpractitionermayhaveimproved
theoutcome.Thisiscertainlyahospital-basedprocedureandnotonethatcanbesuggested
foruseinpractice,butitmayhavescopeyet.
49
Unfortunately,nointerventionshavebeenfoundthatwillreliablyinfluencetheoutcomeof
visualdisturbanceasaresultofHA(orany)fillertreatmentandcanbeusedinpractice.Even
largecentressuchasthosebyChenetalandParketalhaveshownthatthedevelopment
andimplementationisnotrelatedtofunding,butsimplybecausetheredoesnotcurrently
seemtobeareliableremedyforfiller-inducedblindness(Y.Chenetal.,2014;S.W.Parket
al.,2012).
Outcomes
In all cases considered, the exposure to the filler treatment preceded the vascular
complication,andnootherfactorscouldhavebroughtabouttheonsetofsymptoms.Itis
knownthatinjectionoffillercanleadtovisionlossortissueischaemia.
Outof29casesofskinischaemia,62%progressedtoafullrecovery.Consideringthetwelve
casesofocularcomplications,threecases(25%)showedslightimprovementinsightfroma
state of decreased sight, but no cases returned to normal and no cases with vision loss
showedanyimprovement.Acaseoflossofvisionisthereforesignificantlymoreconcerning
forthepatient’slongtermwellbeing.
Unfortunately, the three cases of decreased sight that showed improvement had a high
numberofvariablesincludinginterventionandfollowupandlittledocumentationaboutthe
dosesinvolvedintreatment.Also,thevesselsinvolvedweresmallerandthesymptomsless
severe(orsimplyunrecordedinonecase),whichcouldmeanthattheinterventionhadno
effectorthebodysimplydealtwiththeocclusion.
There has been no reproducibility in the outcome of vision loss as a result of HA filler
treatment.Therefore,ratherthanattempttochangeguidelinestonovelinterventionsthat
havebeensparselystudied,itmaybebettertocontinuewithexistinginterventionsthatare
notconsideredharmfultothepatient(butmayalsoresultinlittlebenefit).Impendingskin
necrosis treatment on the other hand has been studied and adequate, reproducible
50
interventions have been performed and documented by Beleznay et al and Sun et al
especially.PartsofthestudybySunetalwhichconsidernovelherbalremediesshouldbe
furtherinvestigatedforsafetyandefficacy(Beleznayetal.,2014;Sunetal.,2015).
BeleznayetalreportfivecasesofHAfillerinducedvascularcomplicationoveratenyearspan.
Whilst themanagement resulted in adequate resolution for all of their cases, the author
raisesthefollowingobservationsforthiscaseseries.Thefirstisthatthetreatingpractitioners
inthiscaseseriesareextremelyexperiencedandperformaverylargenumberofdermalfiller
treatments, yet they appear tohave ahigher thanaverage vascular complication rate, at
0.05%.Thiscouldbeduetohighernumbersofhighriskproceduresperformedbythiscentre,
althoughalloftheircaseswererelatedtotreatmentofthenasolabialfoldandnotthesite
considered tobe thehighest risk (thenose). Given theexperienceof thepractitioners in
question,itisunlikelytobeduetotechnique.Thisriskratemaybetheactual“honest”risk
rateforvascularcomplications;honestyinreportingisanissuethatisacauseforconcernin
theaestheticindustry,andmanysuspectthattheactualincidenceofvascularcomplicationis
muchhigherthanthe0.001%quotedbyCohenetal(Cohenetal.,2015). Theauthorhas
certainly never heard of non-healthcare providers of aesthetic treatments coming clean
willinglyabouttheircomplications.Afurtherobservationaboutthiscaseseriesisthetime
span over which the cases were reported. Technology in dermal fillers has come along
significantlyover the10-year timeperiod inwhich thecasesconsideredbyBeleznayetal
presented;unfortunately,thepresentationofthecasesalongthistimespanisnotrecorded.
Modernfillersmaynowperformbetterforthepurposeforwhichtheyareintended,andas
suggestedbyInglefieldetal,afactorinminimisingcomplicationsisselectingtheappropriate
filler–withsomanyonthemarket,fillerperformanceisimproving.
The larger centre studieswithmultiple cases in this reviewdemonstrated clearly defined
outcomes,whereasthesmallerstudiesincludedinthereviewareusuallyone-offcasereports
of adverseoutcomeswhichwere less reliabledue to the small amountofdata collected.
Noneofthestudieshadcomparisongroupsfortheintervention.Inordertoarriveatreliable
interventionstrategies,furtherinvestigationwithalargeramountofcaseswouldneedtobe
completedforvisionlossasaresultofHAfillertreatment.
51
Practitionerstatus
Unfortunately, the studies that were considered did not give adequate insight into
practitionersmostatriskofcausingvascularcomplications,asthiswasonlydocumentedin
9outofthe41casesincludedintheanalysis,andinthesethepractitionersweremedically
qualified, i.e.surgeonsorphysicians. Indeed,noresearchhasfocusedspecificallyonthis.
Chenetalbrieflytouchedonthequestionablelicensingstatusoftheperformingpractitioner
in their study. However, in this study it became evident that even thosewith specialist
medicaltrainingcanproveill-equippedtodealwiththecomplicationathand.Inthiscaseit
appears thatwhilstmedically qualified, the surgeonswere not trained to deal with filler
complications.Thepatientwaspassedfromoneinadequatepractitionertothenext,failing
to implement simple protocols including hyaluronidase in favour of a surgical approach,
resultinginapooroutcomeforthepatient
EspeciallyintheUK,whereKeoghdescribedfillersas“acrisiswaitingtohappen”(Keoghet
al.,2013),dermalfillertreatmentcanbeperformedbyanyone,whethermedicallytrainedor
not.Non-medicallyqualifiedindividualsoftenpracticewithnoclinicalsupportwhatsoever,
andindeedtheymaypracticeanywhere(Bruce&Jollie,2014).Unfortunately,non-medical
practitionersoftendonotfeelboundbythesamecodeofethicsthatmedicalprofessionals
must subscribe to, and clinical data on the outcomes of treatment by lesser licensed
individualsmaybedifficulttoachieve.
52
Evidencebasedguidelinesforthepreventionandmanagementofvascularcomplications
Vascular complications involving HA fillers can result in permanent disfigurement. The
following guidelines have been devised by assessing the relevant literature in relation to
favourableoutcomes.Theauthordoesnotconsideritreasonabletoexperimentwithnovel
techniques in practice setting, and any interventions have been devised in line with
techniquesandtreatmentsthathavebeendemonstratedaseffective.
Preventionprotocols
Preventionstrategiesshouldbeconstantlyreferredtobytheaestheticpractitioner.Table3
outlineshowsimplemeasurescanavertseriouscomplications.
53
Table3:PreventionprotocolsforvascularcompromisePreventionstrategiesandin-treatmentprecautionsforVascularComplicationsPatient • Detailedmedicalandpsychologicalhistory
• Aesthetictreatmenthistory–previousrhinoplastiesincreaserisk• Assessmentofanatomicalsites• Validinformedconsent• Documentationofrisks,benefitsanddiscussions• Pre-operativephotographs
Filler • Viscosity/propertiescorrectfortreatmentarea• Reversibilitywithhyaluronidase
HighRiskAreas • Knowyouranatomy• Nasalala:avoidifpossible,extremelyhighrisk• Nasolabialandnose:caution,relativerisk
Technique • Smallestpossibleneedlesizepreventsexcessproductinjection• Aspirationpriortoinjectiontocheckforintra-vascularposition• Retrogradeinjection• Lowinjectionforce–smallersyringesrequirelesspressure• Smallestpossiblevolumes• Stopandreassess ifneedlebecomesblockedor injectionrequires
higherpressurethananticipated• Cannulausedecreasestraumatovessel,especiallyinnose• Placementofproductatmidlineinnose,belowmusculoaponeurotic
layer• Superficialinjectiontechniqueinnasolabialfold• Compression of supply artery during procedure: upper nasolabial
fold,sideofnose,oralcommissure,• Compressionofanastomosis siteduringprocedure: superiornasal
corner
Pain • Investigatepainthatisdisproportionatewithtreatment• Counselpatienttoreportpain• Sharp,suddenpain:queryarterialocclusion• Dull,throbbingpain:queryvenousocclusion
PatientObservationandAftercare
• Blanchingorskinchangesmustbemonitoredtoresolution• Counselpatienttoreportskinorsensorychanges• Fullpost-operativeinstructionsgivenverballyandwritten• Outofhoursemergencynumberavailabletopatients
54
Managementprotocols
Unfortunately, fewstudieshavebeenpublishedthatfocusonthedevelopmentofnewor
improvedprotocolsratherthanjustreportingcases.Evenwiththecasesexaminedinthis
review,highnumbersofvariablesandlownumbersofsubjectsmakethedevelopmentofany
novelguidelinesverydifficult.Managementprotocolscanbeformulatedfromthosecases
thatshowfavourableoutcome,combinedwithhistoricallyeffectiveinterventions.
Lossofvision
Thisislikelytobeaterribledayforboththepatientandthepractitioner.
Retinalnecrosisbecomesirreversiblesome90minutesaftertreatmentandearlyrecognition
ofthecomplicationandrapidtreatmentinductionisessential.Thegoalistoregainsightand
reperfusionoftheretinaiscentraltothis;alleffortsmustbecentraltothis(S.Lazzerietal.,
2013).
Unfortunately,officebasedremediesthataestheticpractitionersarelikelytohaveattheir
disposal prove to be largely fruitless if occlusionof the ocular vessels has occurred. It is
impossible to remove the embolus as it is simply not accessible andhyaluronidase is not
suitable for lysis of emboli in vessels (Kwon et al., 2013). Whilst hyaluronidase is
recommendedimmediately,itwillmostlikelyonlyactontheareaofischaemiaaffectingthe
skinratherthanremedythelossofvision.
Reducingintraocularpressureisdifficultinpractice;eyemassage,alsoknownaseye-CPR,can
beperformedinordertoattempttodislodgetheclot(Y.Chenetal.,2014). Anybladeor
needle incisions to reduce intraocular pressure should be left for consultant
ophthalmologists.Itisperhapsbettertonotattemptanyheroicinterventionsinpractice–
practitionersshouldactwithintheircompetenceandscopeofpracticeandrecognietheneed
for early referral. It is important to have an immediate referral pathway to an
ophthalmologist or the nearest emergency room where treatment with IV
55
methylprednisolone1gfor3days,highdoseoralprednisoloneandaspirin100mgperdaycan
bestarted.Thisishoweverlikelytosimplyreducetheaccompanyingeffectsontheskin(Y.J.
Kimetal.,2011).Thepractitionershouldaccompanythepatientandexplainwhathascaused
thevisionlossandhowthismayhavehappened.
Table 4 outlines the limited possible treatment options available to practitioners in the
practicesetting.
Table4:managementofvisionlossManagementofvisionloss
Identifythecomplication CheckforeyemovementCheckfordecreasedorlossofvision+/-pain
Immediatemanagement • Stoptreatment• Hyaluronidaseupto1000Ualonginjectionsite• Aspirin2x325mg• Eye-CPRtoattempttodislodgeclot
ImmediateReferral • ToA&E/Ophthalmology• Accompanypatientandassisttoexplainhowevent
occurredImpendingnecrosisofskin
Whenthevascularocclusionoccurs,treatmentshouldbeswiftandaggressive.Beleznayet
al produced effective and simple guidelines for treatment that can be used in aesthetic
practicewithoutbeingtoocostlyorconfusingforthepractitioner(Beleznayetal.,2014).
Theprotocolintable5andfollowingalgorithminfigure2isdevisedtoenableathreefold
approachtotissuerecovery,includingdissolutionoftheproduct,vasodilationandincreased
bloodflow.Theimmediateapplicationofheatwithawarmcompresswillaidbloodflowto
thearea.Whilsteffectiveforvasodilation,theuseofnitroglycerin(GTN)pasteiscontroversial
–itmaycauseorthostatichypotension,potentiationofvasodilationwithalcoholandother
vasodilatorssuchassildenafil,andcancauserashes,dizzinessandheadaches(Cohenetal.,
56
2015;Glaichetal.,2006).Nevertheless,theriskofthelikelyoutcomeofnecrosisandscarring
isthoughtbymanytobegreaterthanthetransientsideeffectsfromitsapplication(Edwards,
Wiholm,&Martinez,1996).
Hyaluronidase reverses HA filler and should be used without delay to dissolve filler and
decompressvessels.Anallergytestisnotthoughttobenecessaryinanemergencysituation.
Figure3outlinesthedilutionanduseofhyaluronidasewhenreconstituting1500iUpowder,
suchasHyalase(Wockhardt)intheUK.Someauthors,namelyCavallinietal,advocatethe
useofmuchhigherdosesinordertoachievedissolutionandimproveoutcome(Cavalliniet
al.,2016).
Aspirinblocksplateletaggregationandinananti-inflammatory.Itshouldbegivenwithout
delay.Adjunctivetreatmentmayalsoconsidertheuseofprednisolonetofurtherreducethe
inflammatory response. Other, lesswidelyused interventionsmay include theuseof the
vasodilatorsildenafiltofurtherincreasebloodflow,orlowmolecularweightheparininorder
topreventthrombosis(Beleznayetal.,2014)
Wherenecrosisislargeorwoundsareexhibitingsignsofhealingbysecondaryintent,referral
forhyperbaricoxygentreatmentmaybeconsidered(DeLorenzi,2013).
Similarlytothesituationofvisionloss,itisimportantforthepractitionertorecognisewhen
improvementisnotbeingachieved.Ifthepatientdoesnotseemtobegettingbetterwithin
30-60minutes,itisadvisabletofollowtheguidanceinfigure2andrefertoanemergency
room.Thepractitionershouldaccompanythepatienttoadviseandassistinsteadofsending
ananonymousreferral. It isadvisable toprior tosuchasituationoccurring,practitioners
exploretheiroptionsforreferral–preparationandcommunicationwiththenearesthospital
priortosuchasituationwillalleviatesignificantstressintherealevent.
57
Table5:Managementofskinischaemia/necrosis
ManagementofimpendingskinnecrosisIdentifythecomplication Checkforwarningsigns!
Blanchingandpain:possiblearterialocclusionViolaceousreticulated:possiblevenousocclusion
ImmediatemanagementThreefoldapproach:DissolutionVasodilationIncreasedBloodflow
• Stoptreatment• warmcompress• topical2%nitroglycerinGTNpasteorGTNpatch• massageaffectedareafirmly• hyaluronidase:injecteddeepdermalandsubcutaneous
alongpathofvesselandfillerinjectionarea• 2x325mgaspirin• ifnoimprovementin60minutes:repeathyaluronidase
andconsiderreferral(seefigure2below)• Highflowoxygen• Consideroralprednisolone20-40mg
Adjunctivetreatment • Highflowoxygen• Consideroralprednisolone20-40mg,3-5days• Considersildenafilforvasodilation
24hours • 75mgaspirindaily• reviewandreassure• continueprednisolone,ifstarted
Day1-7 • 75mgaspiringdaily• ApplyGTNeverydayuntilresolved• Dailyfollowupappointment• Woundcareifnecrotictissueandescharformation• Continueprednisoloneuntilday5,ifstarted
Woundcareifnecrosis • Siliconedressing• Minimizebacterialcontamination• Observe for secondary infection, treat if appropriate
withtopicalororalantibiotics• Consider Hyperbaric oxygen if slow healing or large
necrosis(ifaccessible)Day7-14
• Aspirin75mg• Dailyfollowupuntilresolution(maytakemorethan2
weeks)
58
Figure2:Algorithmforthemanagementskinischaemia/necrosis
59
Figure3:reconstitutionanddosageofhyaluronidase(Hyalase/Wockhardt)
60
TheEmergencykit
Inordertomanageacomplicationeffectively,anemergencybagforvascularocclusionshould
beavailableinthepractice,andsuggestedcomponentsareoutlinedinTable6below.
Table6:EmergencyKitforvascularcompromiseEmergencyKitBagforVascularCompromiseDrug ActionHyaluronidase DissolutionofHAfillerGTNpaste vasodilatorAspirin Anti-inflammatory,antiplateletPrednisolone20mgtablets Anti-inflammatoryOxygen (sufficient for 15-20 minutes highflow,recommendedCDcylindersize)
Improvedtissueperfusion
Sildenafil(optional) vasodilationAntihistamines IncaseofallergyAdrenaline Incaseofallergy
61
Furtherstudiessuggested
Mostofthestudiesrecordingvascularcomplicationsinvolveaverysmallsampleofpatients.
ThelargercasenumbersbyChenetal,Parketal,SunetalandBeleznayetalweremostly
referralcentresfordermalfillercomplications,whilstisolatedcentresdocumentsinglecase
reports.
Unfortunately, all studies that have been analysed for this review have been conducted
outsidetheUKandmaynotbeanaccuraterepresentationoftheprotocolsthatUK-based
practitionersareawareof. Inaddition to this, theUKhas someof the laxest regulations
governing theuseofdermal fillersandmanypractitionersare completelyunlicensedand
haveverylittleformalmedicaloranyothertraining,makingtheriskrateintheUKpotentially
higher(Keoghetal.,2013).Thismakesgatheringanyusefulinformationforstudypurposes
extremelydifficult.
In order for UK-based research into the prevention and management of vascular
complicationsaffectingvisionandskintobecompleted,thefollowingaresuggested:
1. A UK-wide dermal filler complications referral centre, consisting of a range of
appropriatelytrainedandexperiencedprofessionalsfromdental,medicalandnursing
backgrounds. A“centre”doesnotnecessarilyhavetobeonelocation,butmorea
body who will accept referrals, treat to a prescribed consistent guideline, share
information, and from this information develop and enhance interventions for
complications.Recommendationsandguidelinescouldbeformulatedfromthis.
2. AUK-widecomplicationsreportingmethod,whichrequirescompletionofareporting
cardwhichcanbesenttoadatacollectioncentre(thismaybethereferralcentre).A
similarsystemhasbeendevelopedbytheMHRA,knownastheyellowcardscheme,
toreportadversereactionstomedicines(MHRA,2016).Astructuredform,suchas
theexampledevelopedbytheauthorseeninfigure4,couldbeusedtoreportthe
incidentorreferthepatienttothereferralcentre,shouldthisbenecessary.Detailed
formssuchasthiswouldfocusthereferringpractitionertoreporttherelevantdata
andassistinanyonwardtreatment. Practitionersshouldbeencouragedtoreport;
theoptionofananonymisedformmayimprovepractitionercompliance.Awareness
62
ofsuchaschemecouldbepromotedbyusingorganisationssuchasSafeFace,IHAS,
cosmeticinsuranceprovidersandpharmaceuticalcompanies.Itdoesnot,however,
commitunscrupulouspractitionerstoreportingadverseeventsandthecooperation
fromnon-healthcarepractitionersshouldbeencouragedatthepointofproductsale.
Figure4:vascularcomplicationreportingform
63
VascularComplicationReportingFormAboutthePatient
Name(Onlyfortreatmentreferral;anonymiseifusingthisformtoreportdata)
Age Sex
Aboutthetreatment
DermalFillerused
BatchNumber
Pleaseusediagrambelowtodetailinjectionsitesandquantities
Injectionsite
Injectionmethod(needle/cannula)
Injectionvolume
Aboutthecomplication
Symptoms
Diagnosis
Timetosymptoms/presentation Treatmentalreadyadministeredandoutcome
Aboutthepractitioner
Name
Profession/Qualification
Referralsfortreatment:sendpatientwithform,accompanywherepossibleReferralsfordatacollection:omitnamesofpractitionerandpatient
64
Conclusions
Itistimetofacetherealissue:fillerblindnessintheUKisnotamatterofifbutwhen.Cases
ofskinnecrosisarealreadyincreasinginincidence.Practitionersmustdobetterinorderto
preventadevastatingoutcomeforthepatient.
Asaestheticmedicineadvances,sodoesthecomplexityoftreatments;fillerrhinoplastiesare
notonlyblurringthelinebetweensurgicalandnon-surgical,butarefastbecomingthemost
dangerousprocedureinthemidface,analreadyhighriskarea.
Theauthorhasrecognisedsignificantshortcomingsinapproachestovascularcomplication
managementintheaestheticpracticeasaresultofthisreview.Progressisslow,withlittle
usefulinputsincethedevelopmentofthefirstprotocolbyHischetalin2007(Hirschetal.,
2007). Nevertheless,effectiveprotocols for themanagementof impendingnecrosishave
beendeveloped,andfurtherdocumentedinthisreview.Earlyinterventionwithintwodays
oftheeventwithuseofhyaluronidaseisaprovenwaytoreducethelongtermsequelaefor
thepatient.Thetreatmentofvisionlossasaresultoffiller,however,isaproverbialstabin
thedark.Therearenoprovenandeffectivemethodstocombatfillerblindnessthatcanbe
effectivelyusedintheaestheticpracticesetting.Theprognosisisoftenhopelessandthere
appeartobenoremediesforthisonthehorizonasnovelconceptshavenotbeenproven
enoughtointegrateintopractice.Hospitalinterventionsareequallyuntested,andtheonus
is on the referring practitioner to convey the patient to an appropriately qualified
ophthalmologist with the right experience before irreversible ocular damage sets in.
Recognisingthecomplicationearlyisthereforeacentralpillartomanagement.
Guidelineshavelargelybeendevelopedinaresponsivefashion,withtheincreasingnumber
ofdermalfillercasualtiespromptingamovetowardsdevelopmentofpreventionprotocols.
Hyaluronicacidfillermaybereversible,butitisstilldangerous,andpractitionersshoulddo
everythingintheirpowertomitigatetherisktothepatient.
ThestateofUKlegislationisamatterofconcern.Practitionersembarkontreatmentwith
littleorno formal training,anddue to lax lawssurrounding theuseofdermal fillersnon-
65
healthcare providers with little accountability are a further risk to patients in an already
potentiallydangerousfield. Ignoranceisbliss,andmanypractitionersattempttreatments
thataresimplyoutsideoftheircompetenciesandwithoutproperpreventionormanagement
strategiesinplace.Itisforthisreasonthattheevidencebasedguidelinesandchartsinthis
reviewweredevisedassimple,easytounderstandprotocolstouseinaestheticpractice.
As we embrace aesthetic treatments in the field of medicine, we must be prepared to
encounterthevascularcomplicationsthatwillundoubtedlycomewithit.AUK-widesupport
orreferralsystemwouldserveasaneffectivehelpforthepatientswhoaremostatneed,
whilst providing data for the development of strategies to improve outcome for those
affectedbyvascularevents.
Thereisadifficultroadaheadwiththetrueincidenceofvascularcomplicationsunknownbut
undoubtedlyrising.Itishopedthatthisreviewwillclarifyandfacilitatetheirtreatment,and
take away some of the stress and anguish felt by practitioner and patient by providing
structureandguidanceinatimeofchaos.
66
References
Abduljabbar,M.H.,&Basendwh,M.A.(2016).Complicationsofhyaluronicacidfillersandtheirmanagements.JournalofDermatology&DermatologicSurgery.doi:10.1016/j.jdds.2016.01.001
Alijotas-Reig,J.,Fernandez-Figueras,M.T.,&Puig,L.(2013).Inflammatory,immune-mediatedadversereactionsrelatedtosofttissuedermalfillers.SeminArthritisRheum,43(2),241-258.doi:10.1016/j.semarthrit.2013.02.001
Beleznay,K.,Humphrey,S.,Carruthers,J.,&Carruthers,A.(2014).VascularCompromisefromSoftTissueAugmentation.JournalofClinicalandAestheticDermatology,7(9),6.
Bruce,C.,&Jollie,C.(2014).Reviewofqualificationsrequiredforthedeliveryofnon-surgcalcosmetic..HealthEducationEngland.
Carruthers,J.D.,Fagien,S.,Rohrich,R.J.,Weinkle,S.,&Carruthers,A.(2014).Blindnesscausedbycosmeticfillerinjection:areviewofcauseandtherapy.PlastReconstrSurg,134(6),1197-1201.doi:10.1097/PRS.0000000000000754
Cavallini,M.,Gazzola,R.,Metalla,M.,&Vaienti,L.(2013).Theroleofhyaluronidaseinthetreatmentofcomplicationsfromhyaluronicaciddermalfillers.AesthetSurgJ,33(8),1167-1174.doi:10.1177/1090820X13511970
Cavallini,M.,Molinary,P.,Al-Niami,F.,Basso,M.,Bettelli,G.,Bezer,L.,...Tarantino,P.(2016).ManagingErrorsandComplicationsinAestheticMedicine.OfficinaEditorialeOltrarno.
Chen,Q.,Liu,Y.,&Fan,D.(2016).SeriousVascularComplicationsafterNonsurgicalRhinoplasty:ACaseReport.PlastReconstrSurgGlobOpen,4(4),e683.doi:10.1097/GOX.0000000000000668
Chen,Y.,Wang,W.,Li,J.,Yu,Y.,Li,L.,&Lu,N.(2014).Fundusarteryocclusioncausedbycosmeticfacialinjections.ChinMedJ,127,1434-1437.doi:10.3760/cma.j.issn.0366-6999.20133254
Chinnawong,D.,Tansatit,T.,Phanchart,P.,&Rachkaew,N.(2015).Surfacelandmarkofthefacialveinforthedermalfillersinjection.JournalofOralandMaxillofacialSurgery,Medicine,andPathology,27(5),645-649.doi:10.1016/j.ajoms.2015.01.015
Cohen,J.L.(2008).Understanding,avoiding,andmanagingdermalfillercomplications.DermatolSurg,34Suppl1,S92-99.doi:10.1111/j.1524-4725.2008.34249.x
Cohen,J.L.,Biesman,B.S.,Dayan,S.H.,DeLorenzi,C.,Lambros,V.S.,Nestor,M.S.,...Sykes,J.(2015).TreatmentofHyaluronicAcidFiller-InducedImpendingNecrosisWithHyaluronidase:ConsensusRecommendations.AesthetSurgJ,35(7),844-849.doi:10.1093/asj/sjv018
Cotofana,S.,Schenck,T.L.,Trevidic,P.,Sykes,J.,Massry,G.G.,Liew,S.,...Remington,B.K.(2015).Midface:ClinicalAnatomyandRegionalApproacheswithInjectableFillers.PlastReconstrSurg,136(5Suppl),219S-234S.doi:10.1097/PRS.0000000000001837
Daines,S.M.,&Williams,E.F.(2013).Complicationsassociatedwithinjectablesoft-tissuefillers:a5-yearretrospectivereview.JAMAFacialPlastSurg,15(3),226-231.doi:10.1001/jamafacial.2013.798
DeLorenzi,C.(2013).Complicationsofinjectablefillers,partI.AesthetSurgJ,33(4),561-575.doi:10.1177/1090820X13484492
67
DeLorenzi,C.(2014).Complicationsofinjectablefillers,part2:vascularcomplications.AesthetSurgJ,34(4),584-600.doi:10.1177/1090820X14525035
Edwards,R.,Wiholm,B.E.,&Martinez,C.(1996).Conceptsinrisk-benefitassessment.Asimplemeritanalysisofamedicine?DrugSaf.,15(1),1-7.
Few,J.,Cox,S.E.,Paradkar-Mitragotri,D.,&Murphy,D.K.(2015).AMulticenter,Single-BlindRandomized,ControlledStudyofaVolumizingHyaluronicAcidFillerforMidfaceVolumeDeficit:Patient-ReportedOutcomesat2Years.AesthetSurgJ,35(5),589-599.doi:10.1093/asj/sjv050
Funt,D.,&Pavicic,T.(2013).Dermalfillersinaesthetics:anoverviewofadverseeventsandtreatmentapproaches.ClinCosmetInvestigDermatol,6,295-316.doi:10.2147/CCID.S50546
Fusi,F.,Manetti,F.,Durante,M.,Sgaragli,G.,&Saponara,S.(2016).ThevasodilatorpapaverinestimulatesL-typeCa2+currentinrattailarterymyocytesviaaPKA-dependentmechanism.VasculPharmacol,76,53-61.
Glaich,A.S.,Cohen,J.L.,&Goldberg,L.H.(2006).InjectionNecrosisoftheGlabella:ProtocolforPreventionandTreatmentAfterUseofDermalFillers.DermatologicSurgery,32(2),276-281.doi:10.1111/j.1524-4725.2006.32052.x
Grunebaum,L.D.,BogdanAllemann,I.,Dayan,S.,Mandy,S.,&Baumann,L.(2009).Theriskofalarnecrosisassociatedwithdermalfillerinjection.DermatolSurg,35Suppl2,1635-1640.doi:10.1111/j.1524-4725.2009.01342.x
Hilton,S.,Schrumpf,H.,Buhren,B.A.,Bolke,E.,&Gerber,P.A.(2014).Hyaluronidaseinjectionforthetreatmentofeyelidedema:aretrospectiveanalysisof20patients.EuropeanJournalofMedicalResearch,19(30),1-5.
Hirsch,R.J.,Cohen,J.L.,&Carruthers,J.D.(2007).Successfulmanagementofanunusualpresentationofimpendingnecrosisfollowingahyaluronicacidinjectionembolusandaproposedalgorithmformanagementwithhyaluronidase.DermatolSurg,33(3),357-360.doi:10.1111/j.1524-4725.2007.33073.x
Hsieh,Y.-H.,Lin,C.-W.,Huang,J.-S.,&Yeh,P.-T.(2015).Severeocularcomplicationsfollowingfacialcalciumhydroxylapatiteinjections:Twocasereports.TaiwanJournalofOphthalmology,5(1),36-39.doi:10.1016/j.tjo.2014.03.009
Humphrey,C.,Arkins,J.,&Dayan,S.(2009).SoftTissueFillersintheNose.AesthetSurgJ,29(6),477-484.
Inglefield,C.,Collins,F.,Duckett,M.,Goldie,K.,Huss,G.,Paun,S.,&Williams,S.(2014).Expertconsensusoncomplicationsofbotulinumtoxinanddermalfillertreatment.MerzAesthetics/AestheticMedicineExpertGroup,2.
ISAPS.(2016).ISAPSInternationalSurveyonAesthetic/CosmeticProceduresPerformedin2014.http://www.isaps.org/Media/Default/global-statistics/2015ISAPSResults.pdf,accessed21.06.2016.
Kassir,R.,Kolluru,A.,&Kassir,M.(2011).Extensivenecrosisafterinjectionofhyaluronicacidfiller:casereportandreviewoftheliterature.JournalofCosmeticDermatology,10,224-231.
Keogh,B.,Halpin,T.,Kennedy,I.,Kydd,C.,Leonard,R.,Parry,V.,...Withey,S.(2013).ReviewoftheRegulationofCosmeticInterventions.DepartmentofHealthPublication.
Kim,D.W.,Yoon,E.S.,Ji,Y.H.,Park,S.H.,Lee,B.I.,&Dhong,E.S.(2011).Vascularcomplicationsofhyaluronicacidfillersandtheroleofhyaluronidasein
68
management.JPlastReconstrAesthetSurg,64(12),1590-1595.doi:10.1016/j.bjps.2011.07.013
Kim,S.N.,Byun,D.S.,Park,J.H.,Han,S.W.,Baik,J.S.,Kim,J.Y.,&Park,J.H.(2014).Panophthalmoplegiaandvisionlossaftercosmeticnasaldorsuminjection.JClinNeurosci,21(4),678-680.doi:10.1016/j.jocn.2013.05.018
Kim,Y.J.,Kim,S.S.,Song,W.K.,Lee,S.Y.,&Yoon,J.S.(2011).Ocularischaemiawithhypotonyafterinjectionofhyaluronicacidgel.OphthalPlastReconstrSurg,27(6),e152-155.doi:10.1097/IOP.0b013e3182078dff
King,M.(2014).Theuseofhyaluronidaseinaestheticpractice.AestheticComplicationsExpertGroup,http://www.visage-aesthetics.com/wp-content/uploads/2016/02/ACEGUIDESHYALASE.pdf,accessed23.06.2016.
Kukovetz,W.R.,Holzmann,S.,&Romanin,C.(1987).Mechanismofvasodilationbynitrates:roleofcyclicGMP.Cardiology,74(Suppl1),12-19.
Kwon,S.G.,Hong,J.W.,Roh,T.S.,Kim,Y.S.,Rah,D.K.,&Kim,S.S.(2013).Ischemicoculomotornervepalsyandskinnecrosiscausedbyvascularembolizationafterhyaluronicacidfillerinjection:acasereport.AnnPlastSurg,71(4),333-334.doi:10.1097/SAP.0b013e31824f21da
Lansdown,A.B.,&Williams,A.(2007).Aprospectiveanalysisoftheroleofsiliconinwoundcare.JWoundCare,16(9),404-407.
Lazzeri,D.,Agostini,T.,Figus,M.,Nardi,M.,Pantaloni,M.,&Lazzeri,S.(2012).Blindnessfollowingcosmeticinjectionsoftheface.PlastReconstrSurg,129(4),995-1012.doi:10.1097/PRS.0b013e3182442363
Lazzeri,S.,Figus,M.,Nardi,M.,Lazzeri,D.,Agostini,T.,&Zhang,Y.X.(2013).Iatrogenicretinalarteryocclusioncausedbycosmeticfacialfillerinjections.AmJOphthalmol,155(2),407-408.doi:10.1016/j.ajo.2012.10.012
MHRA.(2016).Yellowcardscheme.https://yellowcard.mhra.gov.uk/yellowcards/reportmediator/.
Montes,J.R.(2012).Volumetricconsiderationsforlowereyelidandmidfacerejuvenation.CurrOpinOphthalmol,23(5),443-449.doi:10.1097/ICU.0b013e3283560ab5
Mosby.(2013).Mosby'sDentalDictionary,3rdEdition.ElsevierPublishing.Ozturk,C.N.,Li,Y.,Tung,R.,Parker,L.,Piliang,M.P.,&Zins,J.E.(2013).Complications
followinginjectionofsoft-tissuefillers.AesthetSurgJ,33(6),862-877.doi:10.1177/1090820X13493638
Park,S.W.,Woo,S.J.,Park,K.H.,Huh,J.W.,Jung,C.,&Kwon,O.K.(2012).Iatrogenicretinalarteryocclusioncausedbycosmeticfacialfillerinjections.AmJOphthalmol,154(4),653-662e651.doi:10.1016/j.ajo.2012.04.019
Park,T.H.,Seo,S.W.,Kim,J.K.,&Chang,C.H.(2011).Clinicalexperiencewithhyaluronicacid-fillercomplications.JPlastReconstrAesthetSurg,64(7),892-896.doi:10.1016/j.bjps.2011.01.008
Rajabi,M.T.,Naderan,M.,Mohammadi,S.Z.,&Rajabi,M.B.(2015).Centralretinalarteryocclusionfollowingorbitaltumourresection:israpidinterventioneffective?IndianJOphthalmol,63(8),678-680.
Saletu,B.,Garg,A.,&Shoeb,A.(2014).Safetyofnicergolineasanagentformanagementofcognitivefunctiondisorders.BiomedResInt,2014,Epub610103.
Sun,Z.S.,Zhu,G.Z.,Wang,H.B.,Xu,X.,Cai,B.,Zeng,L.,...Luo,S.K.(2015).ClinicalOutcomesofImpendingNasalSkinNecrosisRelatedtoNoseandNasolabialFold
69
AugmentationwithHyaluronicAcidFillers.PlastReconstrSurg,136(4),434e-441e.doi:10.1097/PRS.0000000000001579
Tang,Q.,Han,R.,Xiao,H.,Li,J.,Shen,J.,&Luo,Q.(2014).ProtectiveeffectoftanshinoneIIaonthebrainanditstherapeutictimewindowinratmodelsofcerebralischaemia-reperfusion.ExpTherMed,8(5),1616-1622.
Tansatit,T.,Moon,H.J.,Apinuntrum,P.,&Phetudom,T.(2015).VerificationofEmbolicChannelCausingBlindnessFollowingFillerInjection.AestheticPlastSurg,39(1),154-161.doi:10.1007/s00266-014-0426-z
USFDA,U.S.F.a.D.A.(2016).SoftTissueFillersApprovedbytheCenterforDevicesandRadiologicalHealth.http://www.fda.gov/MedicalDevices/ProductsandMedicalProcedures/CosmeticDevices/WrinkleFillers/ucm227749.htm,accessed19.06.2016.
Wardlaw,J.M.,Murray,V.,Berge,E.,&delZoppo,G.J.(2014).Thrombolysisforacuteischaemicstroke.CochraneDatabaseSystRev,29(7),CD000213.