Review
Treatment
of
extensively
drug-resistant
Gram-negative
infections
in
critically
ill
patients:
Outcome
of
a
consensus
meeting
at
the
13th
Asia-Pacific
Congress
of
Clinical
Microbiology
and
Infection,
October
2012
Paul
Anantharajah
Tambyah
a,
Gabriel
Levy
Hara
b,
George
L.
Daikos
c,
Matthew
E.
Falagas
d,e,
Teresita
Mazzei
f,
Johan
W.
Mouton
g,
Andrea
Novelli
f,
Baiyi
Chen
h,
Minggui
Wang
i,
Wen-Chien
Ko
j,
Taisheng
Li
k,
Xinjian
Fan
l,
Ursula
Theuretzbacher
m,*
a
NationalUniversityofSingapore,1EKentRidgeRoad,Singapore119228,Singapore
b
InfectiousDiseasesUnit,HospitalCarlosG.Durand,BuenosAires,Argentina
c
FirstDepartmentofPropaedeuticMedicine,AthensUniversitySchoolofMedicine,Athens,Greece
d
AlfaInstituteofBiomedicalSciences(AIBS),Athens,Greece
e
TuftsUniversitySchoolofMedicine,Boston,MA,USA
fDepartmentofHealthSciences,ClinicalPharmacologyandOncologySection,UniversityofFlorence,Florence,Italy g
DepartmentofMedicalMicrobiology,RadboudUniversityNijmegenMedicalCentre,Nijmegen,TheNetherlands
h
TheFirstHospitalofChinaMedicalUniversity,Shenyang,China
i
HuashanHospital,FudanUniversity,12M.WulumuqiRoad,Shanghai200040,China
j
DepartmentofInternalMedicine,NationalChengKungUniversityMedicalCollegeandHospital,Taiwan
k
PekingUnionMedicalCollegeHospital,Beijing,China
lWestChinaSchoolofMedicine,WestChinaHospital,SichuanUniversity,Chengdu,China mCenterforAnti-InfectiveAgents,Eckpergasse13,1180Vienna,Austria
Contents 1. Introduction... 118 2. Carbapenems... 118 3. Aminoglycosides... 119 4. Colistin... 119 5. Tigecycline... 120 6. Fosfomycin... 120
7. Useofcombinationtherapy... 120
8. Novelantibioticsindevelopment... 121
9. Conclusion... 121
References... 121
ARTICLE INFO
Articlehistory:
Received8January2013
Receivedinrevisedform10April2013 Accepted11April2013 Keywords: Extensivelydrug-resistant Gram-negativebacteria Resistance Carbapenems Colistin Aminoglycoside Tigecycline ABSTRACT
Infections caused by multidrug-resistant and extensively drug-resistant Gram-negative bacilli are
increasingly challenging to manage in hospitals and long term-care facilities worldwide. As the
therapeuticoptionsarelimited,theInternationalSocietyofChemotherapyincollaborationwiththe
Asia-PacificSocietyofClinicalMicrobiologyandImmunologyorganisedaconsensusconferenceaspart
ofthe13thAsia-PacificCongressofClinicalMicrobiologyandInfection.Apanelofinternationalexperts
fromEurope,theAmericasandAsiawereconvenedtodiscusstheissuesoftherapeuticoptionsforthe
managementofthesedifficult-to-treatpathogens.
ß 2013InternationalSocietyforChemotherapyofInfectionandCancer.PublishedbyElsevierLtd.All
rightsreserved.
*Correspondingauthor.Tel.:+16504884890.
E-mailaddress:[email protected](U.Theuretzbacher).
ContentslistsavailableatSciVerseScienceDirect
Journal
of
Global
Antimicrobial
Resistance
j our na l ho me pa g e : ww w . e l se v i e r . com / l oca t e / j ga r
2213-7165/$–seefrontmatterß2013InternationalSocietyforChemotherapyofInfectionandCancer.PublishedbyElsevierLtd.Allrightsreserved. http://dx.doi.org/10.1016/j.jgar.2013.04.002
1. Introduction
Worldwide antibacterialresistanceisincreasingdramatically
[1].Theglobalmagnitudeoftheresistanceproblemisdifficultto
assess.Conflictingdefinitionsregardingcombinedresistancesthat
areusedbyepidemiologists,microbiologistsandclinicianshave
differentimplicationsforsurveillance,publichealthandclinical
casemanagement.Astheresistanceproblemcontinuestogrow,
harmonised definitions with which to describe and classify
bacteriathat are resistantto multipleantimicrobial agents are
needed.TheEuropeanCentreforDiseasePreventionandControl
(ECDC) and theUS Centers for Disease Control and Prevention
(CDC) published a standardised international terminology to
facilitate grading various antimicrobial resistance profiles and
reportingcomparabledata[2].Theydefine:
multidrug-resistant(MDR)asorganismsthatarenotsusceptible
toatleastoneagentinthreeormoreclassesofantibiotics to
whichtheyareusuallysusceptible(e.g.cephalosporinsand
b
-lactamase inhibitor combinations, fluoroquinolones and
ami-noglycosides);
extensivelydrug-resistant(XDR)asorganismsthatare
suscep-tibleonlytooneortwoclassesofantibiotics(mainlypolymyxins
andprobablytigecycline);and
pandrug-resistant(PDR)asorganismsthatarenotsusceptibleto
allknownorlicensedantibioticscurrentlyavailable.
Thescope ofthisdiscussionisconfinedmainlytoorganisms
thatwouldbethusclassifiedasXDR,forwhichtherearetrulyvery
fewtherapeutic options.Typical examplesof clinicallyrelevant
andmoreprevalentXDRGram-negativepathogensare
Pseudomo-nas aeruginosa, Acinetobacter baumannii and
carbapenemase-producingEnterobacteriaceae(CPE)(mainlyKlebsiellapneumoniae
and Escherichia coli). Commonly, such XDR strains carry an
extensive arsenal of unrelated resistance determinants suchas
multiple
b
-lactamase genes (classes A, B, C and D), 16S rRNAmethylase ArmA- and RmtB-encodinggenes (affecting all
ami-noglycosides) and dihydrofolate reductase genes (affecting
tri-methoprim),aswellaschangesineffluxpumps,chromosomally
mediated porin modifications (affecting
b
-lactams and othercompounds)andmodifiedtopoisomerases(affectingquinolones)
[3].The multidrugresistance pattern is theresult ofcombined
chromosome-andplasmid-encodedmechanisms.Selectionmay
becaused byawide rangeofunrelatedantibacterial drugsand
compounds.
Comparing worldwide epidemiological data is complicated
when applying susceptibility/resistance categories defined by
clinicalbreakpointvalues.Suchbreakpointsmayvarydepending
on the breakpoint-setting organisation. Although the major
organisations such as the Clinical and Laboratory Standards
Institute(CLSI) and theEuropean Committee on Antimicrobial
Susceptibility Testing (EUCAST) are trying to harmonise their
recommendations,somediscrepanciesmayremainpartlyrelated
to the different specific purposes of the guidelines [4]. Both
EUCASTandtheCLSIissuedbreakpointguidelinesin2011.These
take into account certain mechanisms of resistance, wild-type
population distributionsand
pharmacokinetic/pharmacodynam-ics(PK/PD)modelling[5,6].Adetaileddiscussionofthedifferences
betweentheEUCASTandCLSIbreakpointsisbeyondthescopeof
this review but should be considered when comparing and
definingmultidrugresistanceandextensivedrugresistance.
GrowingXDRratesamongclinicallyimportantGram-negative
bacterialedtoincreasingusageoflast-resortantibiotics,eitheras targetedtherapyaccordingtobacteriologicalsusceptibilitytesting orevenasempiricaltherapyincaseofknownhighXDRratesinan
institution. Giving a general and global recommendation for
changingempiricalchoicesinresponsetolocalXDRrateswould
notbefeasibleandwouldacceleratetheemergenceofresistanceto
last-resort antibiotics. Every institution should adapt their
empirical regimens according to their own frequentlyupdated
microbiologicalfindingsandshouldcarefullybalancethedecision
withtheconsequenceofincreasedselectionpressureandpotential
emergence of resistance without remaining therapy options.
Therefore, for the purposes of this manuscript we focus on
targeted therapy options whilst bearing in mind that these
recommendations can only be based on information garnered
from observational studies, given the absence of randomised
clinicaltrials.Theyneedtobefrequentlyadaptedwhenevernew
informationbecomesavailable.
2. Carbapenems
Ingeneral,theuseofcarbapenemsneedstobeweighedagainst
thefurtherselectionoforganismsthatareintrinsicallyresistantto
carbapenems such as Stenotrophomonas maltophilia or other
important nosocomial pathogens such as meticillin-resistant
Staphylococcusaureus(MRSA)orCandidaspp.Moreover,evenin
the case of CPE (as discussed below), inappropriate usage of
carbapenemscouldleadtotheselectionofincreasinglyresistant
organismsintheunitorhospital.Althoughthereisagreatdealof
reluctancetousecarbapenemsinthetreatmentofXDRorganisms
iftheminimuminhibitoryconcentrations(MICs)arebeyondthe
non-susceptiblecategory,theremaybearoleforcarbapenemsin
selected situations dependingon the MIC and the choice of a
combination partner.Thereis no clearevidencefor the
recom-mendation of carbapenems in such situations, but PK/PD
considerations and preliminary clinical results point to the
usefulnessofcarbapenemsdespitebeingin thenon-susceptible
category[7,8].Moreclinicaldataareexpectedtobecomeavailable inthenearfuture.
Using available antibiotics in the setting of MDR and XDR
infectionshighlightstheimportanceofknowingtheMICvalueasit
providesvitalinformationbeyondthebreakpointcategoriesabout
thePDfactor(MIC)inthePK/PDbalance,especiallyinseverelyill patientswithhighPKvariability.ConsideringtheMICratherthan
thebreakpointcategorysupportsthedecisionforusingaspecific
antibioticaswellasforchoosinganoptimiseddosingregimen.
Forcarbapenems,althoughtherearedifferentMICbreakpoints
asdefinedbyEUCASTandCLSI,itmaybepossibletoachieveatime abovetheMIC(T>MIC)forsignificantproportionsofthedosing
intervalfororganisms withMICsevenupto8
m
g/mL.Carbape-nemsdisplaytime-dependentbactericidalkilling,andsignificant
efficacyisobtainedwhenfreedrugconcentrationsremainabove
the MIC for 40–50% of the time between dosing intervals. An
optimiseddosingregimenofacarbapenem(e.g.meropenem2g
every8horimipenem1gevery6h)withaprolongedinfusion
timeofca.2–3heachwouldincreasetheprobabilityofattaining
the PK/PD target. Meropenem 2g every 8h with a prolonged
infusiontime of ca.2–3hdisplaysa relatively highprobability
(85%)forbactericidaltargetattainment(50%T>MIC)whenthe
MICfortheinfectingorganismisupto8
m
g/mL[9].Asthepharmacokineticsoftheantibioticscannotbereliably
predicted in critically illpatients, it would beideal toperform
therapeuticdrugmonitoring(TDM)inthesepatientswithXDRor
PDRinfectionstoensurethatthePDtargetswereindeedachieved
[10].Regrettably,inmostcountries,TDMisnotroutinelyavailable
for
b
-lactam agents such as carbapenems outside of researchsettings. For non-susceptible species with relatively low MICs,
whereacarbapenemisusedinhighdosesdespitethepresenceofa
carbapenemase,it may benecessaryto monitor concentrations
treatmentfailureincreaseswiththeMICevenwhencarbapenems
are used in combination, so they should not be used as no
substantialbenefitcanbeexpectedbasedonthecurrentevidence.
BeyondinvitrodataandPK/PDmodelling,thereisscarcebut
growingdataontheclinicaluseofcarbapenemsinthetreatmentof
XDRandPDRorganisms.Owingtothepressingclinicalproblem,
morestudiesarebeingundertakenandresultswillhopefullybe
available soon. There is no evidence fromdirect head-to-head
comparisonsthatmeropenem,imipenemordoripenemis
signifi-cantly better than any of the other comparators and thus no
recommendationcanbemadeforanyofthethreecarbapenemsin
preferencetotheothers.However,meropenemMICsaregenerally
lowerthanimipenemMICsforGram-negatives—inparticularforP.
aeruginosa(butnotnecessarilyA.baumannii)—andthis,together
withthe good safety profile even at higher doses, may favour
meropenem.Differencesinthepotentialforselectionofresistance
reflect subtle variations in the mode of penetration and other
characteristics.Theclinicalrelevanceofsuchdifferencesbetween
imipenem and meropenem is not clear. There is little, if any,
experiencewithdoripenemdosingregimensbeyond31gdaily.
Thus,despiteitsslightlyhigheractivityagainstP.aeruginosathere isnoclearpreferencetoitsuse.However,thereismoreexperience
withtheuseofmeropeneminhigh-doseandprolongedinfusion
regimens, as well as concerns regarding the increased risk of
seizureswithhighdosesofimipenem.Withrespecttotheuseof
prolonged or continuous infusion, there are some concerns
regarding the stability of imipenem [11,12]. For a duration of
administration of carbapenems beyond 3h the stability of
solutionsatroomtemperatureneedtobeconsideredandstrictly
controlled.
3. Aminoglycosides
As aminoglycoside use has declined in many settings,
aminoglycosideresistancehasdroppedandsometimes
aminogly-cosidesarestilleffectiveagainstsomeXDRstrains[13,14].Owing
tolackofevidencesupportingmonotherapyinthisspecialpatient
population,aminoglycosidesshould,in general,onlybeusedin
combinationbecauseoftheriskofthedevelopmentofresistance
ontherapyaswellasefficacyconcerns[13].Asaminoglycosides
reachhighconcentrationsintheurine,monotherapyappearstobe
an option in urinary tract infections (UTIs).Observationaldata
suggest that these drugs may be superior to polymyxin B or
tigecyclinewhenstrictlytargetingcarbapenem-resistantKlebsiella
sp. in the urine and have been successfully used in a limited
numberofpatientswithUTIs[15].
There are even less data on the appropriate dosage of
aminoglycosides to be used in the treatment of XDR and PDR
organisms.However,giventhefactthattheyare
concentration-dependent antibiotics and that septic patients might have
augmentedrenalclearanceofthesedrugs,theconsensuswasto
usedosagesonthehighsideofthedosingrange.Ontheotherhand,
thedurationofaminoglycosidetherapyshouldbeminimisedto
thesafestlevelappropriatetoreducetheriskoftoxicity.PK/PD
modellingsuggeststhat adaily10mg/kgdoseofgentamicinor
tobramycinprovidesan80%probabilityofresponse,evenforan
MICof4.0
m
g/mL,withanegligiblelikelihoodoftoxicity[16].As statedbefore,thedosageregimenfortheindividualpatientneedstoensureanadequatedrugexposureinthispatientinrelationto
theMICandshouldbere-evaluatedaccordingtoTDM[17].
4. Colistin
Therehasbeenmarkedrenewedinterestinthisagentas,very
often,XDRorganismsareonlysusceptibletothisagent.Colistinfell
outofusebecauseofconcernsaboutnephrotoxicitywhenbetter
toleratedantibioticsbecameavailable.Inrecentyears,itsusehas
increased markedly because of the growing incidence of XDR
Enterobacteriaceae, P.aeruginosaand Acinetobacterspp. Despite
therevivalofthisolddrugandusageasalast-resortantibiotic,
colistinhasneverbeensubjectedtocontemporarydrug
develop-mentprocedures,andmajorgapsinourknowledgehaveonlyjust
recentlybeenaddressed[18].Althoughthis agentis reportedly
effective in vitro against most strains of XDR Gram-negative
bacterial infections, the actual clinical outcomes are often
suboptimal,especiallywhengivenasa singledrug[19,20].This
is partlyduetothesevereunderlyingdiseases oftenassociated
withpatients infected withXDRbacteria but is alsocaused by
inadequatedosingregimensthathavebeenemployedempirically
formorethan50years. Therearemanyfactorsthatneedtobe
taken into account in understanding the PK/PD properties of
colistin,includingitscomplexpharmacokinetics[21].Colistin,a
mixture of several compounds,is administered as the inactive
prodrug colistimethate sodium (CMS). CMS and colistin have
mixedroutesofrenalandnon-renalelimination,ahalf-lifethat
varies between different patient populations, and a volume of
distributionthatincreaseswithcriticalillness.Inaddition,protein
binding is concentration-dependent. The complex
pharmacoki-netics of colistin have only recentlybegun tounfold based on
newly developed specific analytical methods. Based on the PK
characteristics,aloadingdosetoachieve activecolistin
concen-trationsquicklyandamaintenancedosingregimenthatisbased
onthepatient’skidneyfunctionhavebeenproposed.Thedosesof
CMSusedforsystemicinfectionsinadultsrangewidely.Basedon
recentPKstudies,activeconcentrationsatthesiteofinfectioncan
onlybeachievedwithhighdosages.Somerecentregimensused
4.5millionInternationalUnits(MIU)twiceadayor3MIUthree
timesadayafteraloadingdoseof9MIUasthismightprovidea
balancebetweengoodclinicaleffectandacceptabletoxicity.On
the other hand, such high doses revive the earlier concerns
regardingrenalandneurologicaltoxicity.Thereisclearlyaneedfor
randomisedclinicaltrialstodetermineoptimaldosingregimensof
colistin in order to optimise clinical outcome, minimise the
development of resistance, reduce toxicity and test optimum
combinations.
Multicentre randomised studies, one funded by the 7th
FrameworkProgramoftheEuropeanUnion(EU)
(AIDA-Preserv-ingOldAntibioticsfortheFuture)andoneclinicalstudyfundedby
the US National Institute of Allergy and Infectious Diseases
(NIAID), are underway to compare colistin treatment with a
combination regimen of colistin+carbapenem. Until more
information is available, CMS should be exclusively given to
patientswithinfectionscausedbyXDRGram-negativebacteria
susceptibleonlytocolistin.Thisdrugshouldnotbeusedtotreat
infectionscausedbybacteriasusceptibletootherantibioticsand
shouldnotbeusedasasingledrugalthoughclinicalevidenceis
currentlyscarce[22].
Colistinhasbeenusedinanebulisedformfortheprevention
and treatment of respiratory tract infections, although the
evidenceforthisindicationoutsidethecysticfibrosisareaisvery
limited [23]. Preliminary data in a small series of critically ill
patients withventilator-associated pneumonia(VAP)caused by
carbapenem-resistantAcinetobacterandK.pneumoniaecouldnot
show superiority of a regimen that included concomitant
inhalation of colistinin additiontointravenous(i.v.)treatment comparedwithi.v.treatmentalone[24],althoughitwaslimitedby
smallpatientnumbersandstudydesignissues[25].Theresultsof
small retrospective studies are conflicting, as demonstrated by
otherstudieswithimprovedsurvivalwithconcomitant
adminis-trationofaerosolisedandi.v. colistinin patientswithVAP[26]. Poolingthedataofrelevantclinicalstudies,adjunctiveaerosolised
comparedwithi.v.colistinalone[27].Moreconcreteclinicaldata
areneededtosupportsuchtreatmentdecisions.
Adjunct intrathecal colistin has been used for ventriculitis
causedbyXDRorganismsbuttheevidenceislargelyconfinedto
casereports[28].Arecentreviewof83episodesofMDR/XDRA.
baumanniiventriculitis/meningitisconcludedthatintraventricular
or intrathecal administration with concomitant i.v. colistin
representsanimportantmodeoflast-resorttherapy[29].
5. Tigecycline
There wassome optimism initially that tigecyclineretained
activity against a portion of XDR Enterobacteriaceae and A.
baumannii,butmorerecentstudiesshowedonlylimited
useful-ness of tigecycline in CPE (47% susceptible) with resistance
scattered amongst K. pneumoniae and Enterobacter spp. [30].
Similarly,contemporaryisolatesfromGreecewereonly
suscepti-blein64%ofcasesforXDRK.pneumoniaeand44%forA.baumannii [31]. Clinically, theuse of tigecycline is limited due to lack of
controlledstudies targetingXDRorganisms, concernsregarding
efficacyand safety incriticallyill patients, aswellastherapid
development of resistance. The emergence of resistance via
upregulationof effluxpumps during therapy is well described
andamatterofconcern[32].Theefficacyoftigecyclineinclinical trialsisrelatedtolowMICs,lowbaselinealbuminconcentration,
lowerseverityandabsenceofVAP[33],usuallynotrepresenting
thepatientpopulationwithXDRinfection.Poolingnon-inferiority studieswithtigecyclineshowedexcessmortalityincriticallyill
patientsinapprovedandnon-approvedindications,promptinga
USFoodandDrugAdministration(FDA)warningagainsttheusage
oftigecyclinewhenotheroptionsareavailable[34,35].Inaddition,
tigecyclineisalsohamperedbylowconcentrationsintheblood
andproteinbindingofca.70%thatlimitsitsuseinbloodstream
infections[36].Increasingthedoseoftigecyclineisnotgenerally
recommended as it is poorly tolerated with increasing risk of
nausea/vomiting related to higher drug exposure and the
consequences of this are not well described [37]. Although
tigecycline monotherapy results in higher failure rates [19,38]
thereis,however,someevidencethattigecyclineincombination
withcolistinandacarbapenemcouldbeeffectiveinthetreatment
ofseverelyillpatients withXDRGram-negativeinfections.This
remainstobeshowninlargerprospectivestudies.
6. Fosfomycin
FosfomycinisactiveinvitroagainstsomeXDRGram-negative
bacteria[39]andcouldbeacomponentofacombinationregimen
forthetreatmentoftheseinfections.Theconcentrationsrequired foractivityare,however,relativelyhighfornon-fermenters.There areonlyafewavailabledataonPK/PDcharacteristicsandclinical
outcomesoffosfomycin[40,41].BasedoninvitrodataandPK/PD
modelling, a high dose and prolonged infusion combination
strategy has been suggested [40]. Clinically, doses of 4g four
timesdailyand upto8gthreetimesdaily havebeenreported
[41,42],butappropriatedosageregimensinXDRsituationshave
notyetbeenestablished.
There are concernsaboutfosfomycin’slack ofefficacyfor A.
baumannii as well as rapid development of resistance in XDR
Gram-negativebacteria(especiallyP.aeruginosa)duringtherapy
evenwhenusedincombination[43,44].Fosfomycinshouldnever
beusedasmonotherapyforseveresystemicinfections.Asthei.v.
form of fosfomycin is not approved in many countries, most
experience is with oral preparations of the agent [42]. This
achieveshighconcentrationsintheurinarytractandhasbeenused
inthetreatmentofXDRandevenPDRUTIs.Fosfomycin’slimited
availabilityasani.v.formulationandmissingPK/PDdataaswellas clinicalinformationrestrictsitsuseinseverelyillpatients.
7. Useofcombinationtherapy
CombinationtherapyasastrategyagainstXDRandPDR
micro-organismsisincreasinglybeinguseddespitethepaucityofclinical evidence[45,46].Intermsofsusceptibility,itshouldberealised
thatsusceptibilitybreakpointsarealwaysbasedonmonotherapy
regimensandarethereforelesssuitabletouseasapredictorfor
clinical outcome than usually thought. Exploring combination
therapy,a recentretrospectivemulticentrestudyfocusingonK.
pneumoniae carbapenemase (KPC)-producing K. pneumoniae
bloodstream infections showed that carbapenems could be
effectiveincombinationwithotheragentsbutthisdidnotapply
equally toorganisms withanMIC>8
m
g/mL[47].The authorsconcludedthatcombinedtreatmentwithtwoormoredrugswith
invitroactivityagainsttheisolate,especiallythosealsoincludinga
carbapenem,maybemoreeffectivethanactivemonotherapy[47].
Similar results of a survival benefit with combination therapy
including a carbapenem were reported by Qureshi et al. [38].
Tzouvelekisetal.compiledtheresultsof34smallstudiesinvolving
atotalof298patients witheitherKPC-ormetallo-
b
-lactamase(MBL)-producingK. pneumoniae[19]. Theyfoundthat themost
effective regimens were those that included two active agents
including a carbapenem. Combinations of carbapenem with
colistin (5.5% of failures) or with an aminoglycoside (6.2%)
performedsignificantlybetterthanwhenthesedrugswereused
alone oras partof othercombinations, whilstcombinations of
tigecycline(24%of failures),colistin(32%) andaminoglycosides
(33.3%)inregimensnotincludingacarbapenemexhibitedhigher
failurerates.Ontheotherhand,theuseoftigecyclineorcolistinas
single active agents resultedin higher failure rates (35.7% and
47.2%, respectively).The superiority of combination therapy is
supportedbyanotherrecentreviewoncurrentclinicalexperience
withcarbapenemase-producingGram-negativebacteria[48].
Clinicalevidenceforthesuperiorityofcombinationtherapyin
XDR P.aeruginosa infections has not beenestablished yet. The
publishedmeta-analysisdidnotstratifyfortheresistancestatusas
itincludedmostlystudiesinthepre-XDRera[49–51].Also,the
studybyKumaretal.didnotfocusonXDRP.aeruginosaanddoes
notprovideanswerstothisclinicalquestion[52].Asfewtherapy
options exist for XDR P. aeruginosa [53]—usually colistin, less
commonlyfosfomycinandaminoglycosides—combination
thera-py appears to be preferable based on non-clinical data of the
abovementionedantibiotics.Clinicaldataarewarranted.
Acinetobacterspp.isincreasinglybecomingmoreXDRwithfew
therapeutic options remaining. Similarly, there are no well
designedclinicaltrials tocomparetreatment regimensforsuch
infections.Mostavailabledataarederivedfromuncontrolledcase series,animalmodelsorinvitrostudies.Investigationsevaluating
combinationsofrifampicin,sulbactam,aminoglycosides,colistin,
carbapenems and other drugs for the management of XDR
Acinetobacterinfectionshaveshownconflictingdata[54].Alack
ofcontrolledclinicaltrialsmakesitdifficulttoevaluatetheroleof
combinationtherapyforXDRAcinetobacterinfection.
Because of the limited activity and rapid emergence of
resistancetocolistin,thisdrugisoftenpropagatedasacomponent
ofacombinationasmentionedinthecarbapenemandtigecycline
sections. Based on the synergism between colistin and other
antibioticsinvitroandinanimalstudies,avarietyofcombinations havebeenusedintheclinicalsettingandincludedinuncontrolled studiesorsinglecaseseries.Besidesthecombinationsofcolistin
with carbapenems or tigecycline, combination regimens of
colistin+rifampicin have been widely discussed but the real
8. Novelantibioticsindevelopment
Most research and development (R&D) efforts focus on the
developmentofnewanaloguesofwellknownantibacterialclasses,
although cross-resistance is already apparent or anticipated
against every one of the new analogues currently in the
development pipeline. Therefore, such modified analogues of
existingantibioticclassesarelesslikelytobeusefulforXDRorPDR
pathogens.Similarly,theconceptofcombiningaknown
cephalo-sporinorcarbapenemanaloguewithanew
b
-lactamaseinhibitor(BLI)withextendedornarrow
b
-lactamasecoverageappearstobuysometimeintheresistancerace.Thelong-known
cephalo-sporinceftazidime aswellas thenewcephalosporinanalogues
ceftarolineand ceftolozaneare beingdeveloped incombination
withtheoldBLItazobactamorthenewBLIavibactam.Avibactam
expandstheinhibitory spectrum oftazobactam toinclude KPC
(classA),classCand someOXA type
b
-lactamasesbutnot therapidly spreading metallo-enzymes (e.g. NDM). XDR
Gram-negatives may have MBLs, upregulated efflux pumps, porin
deficiencies or hyperproduction of extended-spectrum
b
-lacta-mases/carbapenemases among their resistance traits.All
insur-mountablehurdlesforanyBLI!Inadditiontoavibactam,atleast
sixmoreBLIsareindevelopment,withconsiderablegapsintheir
activity that remain to be bridged. The phase 3-ready new
aminoglycoside analogue plazomicin overcomes many
wide-spreadaminoglycoside-specificresistancemechanisms[56].
Only R&D of new chemical entities unrelated to a known
antibacterial scaffold, not showing cross-resistance to existing
antibacterialdrugclassesandwithalowpropensityforemerging
resistance, would be a promising option against XDR strains.
However,veryfewsuchR&Dprogrammesarepubliclydisclosed
andtheyarenotenoughtofilltheR&Dpipelinestoincreasethe
likelihoodofnoveldrugseventually reachingthepatient.Novel
antibiotics without potential cross-resistance with established
drugsandpromisingactivityagainstXDRbacteriaareneitheron
thehorizonnoranticipatedforthenearfuture.Governmentsinthe USAandEUhaveinitiatedseveralhigh-levelinitiativesincludinga
widevarietyofpoliciesandincentivesdesignedtojumpstartthe
development of novel antibacterial drugs and technologies, as
recentlyreviewedbyTheuretzbacher[1,57].
9. Conclusion
Multiresistant Gram-negative bacilli are increasing threats
worldwide.Whilstprevention,antibioticstewardshipand
effec-tiveinfectioncontrolstrategieswillbeparamountintheeffortsto
controltheemergenceandspreadoftheseorganisms,novelagents
areurgentlyneededtotreatpatientsalreadyinfectedwithXDR
and PDR organisms. Right now, we are primarily limited to
combinationsofexistingagentsincludingpolymyxins,
aminogly-cosides, tigecycline, fosfomycin and carbapenems. The latter
agents appear to retain some activity against
carbapenemase-producingorganismswhenusedincombinationandfororganisms
with MICs up to 8
m
g/mL in the majority of patients whenoptimiseddosingregimensareused.Untilneweragentsmaybe
availableforclinicaluseinthefarfuture,theonlyoptionistouse
existing agents in a better way, for instance by optimising
exposuresandtheuseofcombinationtherapy[58].High-quality
randomisedclinicaltrialswouldprovideessentialinformationto
determinethebestcombinationofestablishedantibioticsthatcan
beused to treat XDRand PDR Gram-negativeinfections. More
researchandpractical guidancefor individualisingdosage
regi-mensbasedonTDMandclinicaldecisionsupportsystemsshould
beconsideredtoachieveoptimalexposureincriticallyillpatients.
Presently,evidence-basedrecommendationsonhowtooptimally
treat critically ill patients with XDR or PDR Gram-negative
organismsarenotpossibleasclinicaldataarescarceormissing
altogether.Generalprinciplesoftreatment,suchaschoosingthe
most appropriate antibiotics based on MIC data as well as
consideringtheindividualdrugexposure basedonTDM, would
help to optimise therapy but may not be available in all
institutions. Funding
None.
Competinginterests
PAT received research support from GSK, Sanofi-Pasteur,
Inviragen, Fab’entech, Baxter and Adamas, and honoraria from
MSD,Johnsonand Johnson,3MandAstraZeneca(last5 years);
GLH received consultancy/lecture fees from Janssen, MSD and
Phoenix ArgentinaLaboratories; GLDhasservedas anadvisory
boardmemberforPfizerHellas,Novartis,MSDandAchaogenand
has also received a research grant from Pfizer Hellas; MEF
participatedinadvisoryboardsofAchaogen,Astellas,AstraZeneca, Bayer,BasileaandPfizer,receivedlecturehonorariafromAngelini,
Astellas, AstraZeneca,Cipla,Glenmark, MerckandNovartis, and
receivedresearchsupportfromAngelini,AstellasandRokitan(last
10years);TMreceivedgrantsandsupportfromPfizer,Angelini,
AstellasPharma,EliLilly,Janssen-Cilag,Novartis,Sanofi-Aventis,
Schering-Plough,Valeas,WyethLederleandZambon(last5years);
ANreceivedgrantsandsupportfromAstra-Zeneca,Gilead,Janssen,
MSD and Pfizer (last 2 years). All other authors declare no
competinginterests.
Ethicalapproval
Notrequired. References
[1]TheuretzbacherU.Globalantibacterialresistance:thenever-endingstory. JournalofGlobalAntimicrobialResistance2013.http://dx.doi.org/10.1016/ j.jgar.2013.03.010.inpress.
[2]MagiorakosAP,SrinivasanA,CareyRB,CarmeliY,FalagasME,GiskeCG,etal. Multidrug-resistant,extensivelydrug-resistantandpandrug-resistant bacte-ria:an internationalexpertproposalforinterim standarddefinitions for acquiredresistance.ClinicalMicrobiologyandInfection2012;18:268–81. [3]PoirelL,BonninRA,NordmannP.Analysisoftheresistomeofa
multidrug-resistantNDM-1-producingEscherichiacolistrainbyhigh-throughputgenome sequencing.AntimicrobialAgentsandChemotherapy2011;55:4224–9. [4]HombachM,BloembergGV,Bo¨ttgerEC.Effectsofclinicalbreakpointchanges
in CLSI guidelines2010/2011and EUCAST guidelines2011 onantibiotic susceptibilitytestreportingofGram-negativebacilli.JournalofAntimicrobial Chemotherapy2012;67:622–32.
[5]KahlmeterG,BrownDF,GoldsteinFW,MacGowanAP,MoutonJW,Osterlund A,etal.EuropeanharmonizationofMICbreakpointsforantimicrobial sus-ceptibility testing of bacteria. Journal of Antimicrobial Chemotherapy 2003;52:145–8.
[6]MoutonJW,BrownDF,ApfalterP,Canto´nR,GiskeCG,IvanovaM,etal.Therole ofpharmacokinetics/pharmacodynamicsinsettingclinicalMICbreakpoints: theEUCASTapproach.ClinicalMicrobiologyandInfection2012;18:E37–45. [7]Esterly JS,Wagner J, McLaughlinMM, PostelnickMJ,Qi C, Scheetz MH.
Evaluationofclinicaloutcomesinpatientswithbloodstreaminfectionsdue toGram-negativebacteriaaccordingtocarbapenemMICstratification. Anti-microbialAgentsandChemotherapy2012;56:4885–90.
[8]DaikosGL,MarkogiannakisA.Carbapenemase-producingKlebsiella pneumo-niae:(when)mightwestillconsidertreatingwithcarbapenems?Clinical MicrobiologyandInfection2011;17:1135–41.
[9]KutiJL,DandekarPK,NightingaleCH,NicolauDP.UseofMonteCarlo simula-tiontodesignanoptimizedpharmacodynamicdosingstrategyfor merope-nem.JournalofClinicalPharmacology2003;43:1116–23.
[10]HayashiY,LipmanJ,UdyAA,NgM,McWhinneyB,UngererJ,etal.b-Lactam therapeuticdrugmonitoringinthecriticallyill:optimisingdrugexposurein patientswithfluctuatingrenalfunctionandhypoalbuminaemia.International JournalofAntimicrobialAgents2013;41:162–6.
[11]KeelRA, SutherlandCA, CrandonJL, NicolauDP. Stabilityof doripenem, imipenemand meropenemat elevatedroom temperatures.International JournalofAntimicrobialAgents2011;37:184–5.
[12]MoutonJW,VinksAA.Continuousinfusionofb-lactams.CurrentOpinionin CriticalCare2007;13:598–606.
[13]Durante-MangoniE,GrammatikosA,UtiliR,FalagasME.Dowestillneedthe aminoglycosides? International Journal of Antimicrobial Agents 2009;33:201–5.
[14]MarakiS,MavrosMN,KofteridisDP,SamonisG,FalagasME.Epidemiologyand antimicrobialsensitivitiesof536multi-drug-resistantGram-negativebacilli isolatedfrompatientstreatedonsurgicalwards.SurgicalInfections(Larchmt) 2012;13:326–31.
[15]SatlinMJ,KubinCJ,BlumenthalJS,CohenAB,FuruyaEY,WilsonSJ,etal. Comparativeeffectivenessofaminoglycosides,polymyxinB,andtigecycline forclearanceofcarbapenem-resistantKlebsiellapneumoniaefromurine. An-timicrobialAgentsandChemotherapy2011;55:5893–9.
[16]DrusanoGL,LouieA.Optimizationofaminoglycosidetherapy.Antimicrobial AgentsandChemotherapy2011;55:2528–33.
[17]vanLent-EversNA,MathoˆtRA,GeusWP,vanHoutBA,VinksAA.Impactof goal-orientedandmodel-basedclinicalpharmacokineticdosingof aminogly-cosidesonclinicaloutcome:acost-effectivenessanalysis.TherapeuticDrug Monitoring1999;21:63–73.
[18]BergenPJ, LandersdorferCB, LeeHJ,Li J, NationRL. ‘Old’antibioticsfor emergingmultidrug-resistantbacteria.CurrentOpinioninInfectiousDiseases 2012;25:626–33.
[19]Tzouvelekis LS, Markogiannakis A, Psichogiou M,Tassios PT, Daikos GL. CarbapenemasesinKlebsiellapneumoniaeandotherEnterobacteriaceae:an evolving crisis of global dimensions. Clinical Microbiology Reviews 2012;25:682–707.
[20]PaulM, BisharaJ, LevcovichA, ChowersM,GoldbergE, SingerP,et al. Effectivenessandsafetyofcolistin:prospectivecomparativecohortstudy. JournalofAntimicrobialChemotherapy2010;65:1019–27.
[21]CouetW,Gre´goireN,MarchandS,MimozO.Colistinpharmacokinetics:thefog islifting.ClinicalMicrobiologyandInfection2012;18:30–9.
[22]DerisZZ,YuHH,DavisK,SoonRL,JacobJ,KuCK,etal.Thecombinationof colistinanddoripenemissynergisticagainstKlebsiellapneumoniaeatmultiple inoculaandsuppressescolistinresistanceinaninvitropharmacokinetic/ pharmacodynamic model. Antimicrobial Agents and Chemotherapy 2012;56:5103–12.
[23]Michalopoulos A,PapadakisE.Inhaledanti-infectiveagents:emphasis on colistin.Infection2010;38:81–8.
[24]KofteridisDP,AlexopoulouC,ValachisA,MarakiS,DimopoulouD, Georgo-poulosD, etal.Aerosolized plus intravenouscolistinversus intravenous colistin alone for the treatment of ventilator-associated pneumonia: a matched-controlstudy.ClinicalInfectiousDiseases2010;51:1238–44. [25]KwaAL,FalagasME,MichalopoulosA,TamVH.Benefitsofaerosolizedcolistin
forventilator-associatedpneumonia:absenceofproofversusproofof ab-sence?ClinicalInfectiousDiseases2011;52:1278–9.
[26]ArnoldHM,SawyerAM,KollefMH.Useofadjunctiveaerosolizedantimicrobial therapyinthetreatmentofPseudomonasaeruginosaandAcinetobacter bauman-niiventilator-associatedpneumonia.RespiratoryCare2012;57:1226–33. [27]FlorescuDF,QiuF,McCartanMA,MindruC,FeyPD,KalilAC.Whatisthe
efficacy and safety of colistinfor thetreatment of ventilator-associated pneumonia?Asystematicreviewandmeta-regression. ClinicalInfectious Diseases2012;54:670–80.
[28]KhawcharoenpornT,ApisarnthanarakA,MundyLM.Intrathecalcolistinfor drug-resistantAcinetobacterbaumanniicentralnervoussysteminfection:a case series and systematic review. Clinical Microbiology and Infection 2010;16:888–94.
[29]KaraiskosI,GalaniL,BaziakaF,GiamarellouH.Intraventricularandintrathecal colistinasthelasttherapeuticresortforthetreatmentofmultidrug-resistant andextensivelydrug-resistantAcinetobacterbaumanniiventriculitisand men-ingitis: a literature review. InternationalJournal of AntimicrobialAgents 2013.http://dx.doi.org/10.1016/j.ijantimicag.2013.02.006.pii:S0924-8579(13) 00054-X.
[30]LivermoreDM,WarnerM,MushtaqS,DoumithM,ZhangJ,WoodfordN.What remains against carbapenem-resistant Enterobacteriaceae? Evaluation of chloramphenicol,ciprofloxacin,colistin,fosfomycin,minocycline, nitrofuran-toin,temocillinandtigecycline.InternationalJournalofAntimicrobialAgents 2011;37:415–9.
[31]ZarkotouO,PournarasS,AltouvasG,PitirigaV,TzirakiM,MamaliV,etal. Comparative evaluation of tigecycline susceptibility testing methods for expanded-spectrumcephalosporin-andcarbapenem-resistant Gram-nega-tivepathogens.JournalofClinicalMicrobiology2012;50:3747–50. [32]SpanuT,DeAngelisG,CiprianiM,PedruzziB,D’InzeoT,CataldoMA,etal.In
vivo emergence oftigecycline resistance inmultidrug-resistant Klebsiella pneumoniae andEscherichiacoli.AntimicrobialAgentsand Chemotherapy 2012;56:4516–8.
[33]BhavnaniSM,RubinoCM,HammelJP,ForrestA,DartoisN,CooperCA,etal. Pharmacologicalandpatient-specificresponsedeterminantsinpatientswith hospital-acquiredpneumoniatreatedwithtigecycline.AntimicrobialAgents andChemotherapy2012;56:1065–72.
[34]Prasad P,Sun J, DannerRL, Natanson C. Excessdeaths associated with tigecyclineafterapprovalbasedonnoninferioritytrials.ClinicalInfectious Diseases2012;54:1699–709.
[35]USFoodandDrugAdministration(FDA).FDAdrugsafetycommunication: increasedriskofdeathwithTygacil(tigecycline)comparedtootherantibiotics used to treat similar infections. http://www.fda.gov/Drugs/DrugSafety/ ucm224370.htm[accessedMarch2013].
[36]VanWartSA,OwenJS,LudwigEA,MeagherAK,Korth-BradleyJM,Cirincione BB.Populationpharmacokineticsoftigecyclineinpatientswithcomplicated intra-abdominalorskinandskinstructureinfections.AntimicrobialAgents andChemotherapy2006;50:3701–7.
[37]RubinoCM,BhavnaniSM,ForrestA,DukartG,DartoisN,CooperA,etal. Pharmacokinetics–pharmacodynamicsoftigecyclineinpatientswith com-munity-acquired pneumonia. Antimicrobial Agents and Chemotherapy 2012;56:130–6.
[38]QureshiZA,PatersonDL,PotoskiBA,KilaykoMC,SandovskyG,SordilloE,etal. TreatmentoutcomeofbacteremiaduetoKPC-producingKlebsiella pneumo-niae:superiorityofcombinationantimicrobialregimens.AntimicrobialAgents andChemotherapy2012;56:2108–13.
[39]EndimianiA,PatelG,HujerKM,SwaminathanM,PerezF,RiceLB,etal.Invitro activityoffosfomycinagainstblaKPC-containingKlebsiellapneumoniaeisolates, includingthosenonsusceptibletotigecyclineand/orcolistin.Antimicrobial AgentsandChemotherapy2010;54:526–9.
[40]AsuphonO,MontakantikulP,HoungsaitongJ,SonthisombatP,KiratisinP. OptimizingfosfomycindosingfortreatmentofPseudomonasaeruginosa infec-tionsincritically-illpatientsbasedonpharmacokinetics/pharmacodynamics (PK/PD)target.In:52ndInterscienceConferenceonAntimicrobialAgentsand Chemotherapy(ICAAC).ASMPress;2012.p.A-1279.
[41]MichalopoulosA,VirtziliS,RafailidisP,ChalevelakisG,DamalaM,FalagasME. Intravenousfosfomycinforthetreatmentofnosocomialinfectionscausedby carbapenem-resistantKlebsiellapneumoniaeincriticallyillpatients:a pro-spectiveevaluation.ClinicalMicrobiologyandInfection2010;16:184–6. [42]FalagasME,KastorisAC,KapaskelisAM,KarageorgopoulosDE.Fosfomycinfor
thetreatmentofmultidrug-resistant,includingextended-spectrumb -lacta-maseproducing,Enterobacteriaceaeinfections:asystematicreview.Lancet InfectiousDiseases2010;10:43–50.
[43]KarageorgopoulosDE,MiriagouV,TzouvelekisLS,SpyridopoulouK,DaikosGL. EmergenceofresistancetofosfomycinusedasadjuncttherapyinKPC Klebsi-ellapneumoniaebacteraemia:reportofthreecases.JournalofAntimicrobial Chemotherapy2012;67:2777–9.
[44]KarageorgopoulosDE,WangR,YuXH,FalagasME.Fosfomycin:evaluationof thepublishedevidenceontheemergenceofantimicrobialresistancein Gram-negative pathogens. Journal of Antimicrobial Chemotherapy 2012;67: 255–68.
[45]AbadCL,KumarA,SafdarN.Antimicrobialtherapyofsepsisandsepticshock— whenaretwodrugsbetterthanone?CriticalCareClinics2011;27:e1–27. [46]PaulM,LeiboviciL.Combinationantimicrobialtreatmentversus
monother-apy:thecontributionofmeta-analyses.InfectiousDiseaseClinicsofNorth America2009;23:277–93.
[47]TumbarelloM,VialeP,ViscoliC,TrecarichiEM,TumiettoF,MarcheseA,etal. PredictorsofmortalityinbloodstreaminfectionscausedbyKlebsiella pneu-moniaecarbapenemase-producingK.pneumoniae:importanceofcombination therapy.ClinicalInfectiousDiseases2012;55:943–50.
[48]AkovaM,DaikosGL,TzouvelekisL,CarmeliY.Interventionalstrategiesand currentclinicalexperiencewithcarbapenemase-producingGram-negative bacteria.ClinicalMicrobiologyandInfection2012;18:439–48.
[49]SafdarN,HandelsmanJ,MakiDG.Doescombinationantimicrobialtherapy reducemortalityinGram-negativebacteraemia?Ameta-analysis. Lancet InfectiousDiseases2004;4:519–27.
[50]VardakasKZ,TansarliGS,BliziotisIA,FalagasME.b-Lactamplus aminoglyco-sideorfluoroquinolonecombinationversusb-lactammonotherapyfor Pseu-domonas aeruginosa infections:a meta-analysis. International Journal of AntimicrobialAgents2013;41:301–10.
[51]PaulM,Benuri-SilbigerI,Soares-WeiserK,LeiboviciL.blactammonotherapy versusblactam-aminoglycosidecombinationtherapyforsepsisin immuno-competent patients:systematicreview andmeta-analysis ofrandomised trials.BritishMedicalJournal2004;328:668.
[52]KumarA,ZarychanskiR,LightB,ParrilloJ,MakiD,SimonD,etal.Cooperative AntimicrobialTherapyofSepticShock(CATSS)DatabaseResearchGroup. Earlycombinationantibiotictherapyyieldsimprovedsurvivalcomparedwith monotherapyinsepticshock:apropensity-matchedanalysis.CriticalCare Medicine2010;38:1773–85.
[53]TamVH,ChangKT,AbdelraoufK,BriosoCG,AmekaM,McCaskeyLA,etal. Prevalence,resistancemechanisms,andsusceptibilityofmultidrug-resistant bloodstreamisolatesofPseudomonasaeruginosa.AntimicrobialAgentsand Chemotherapy2010;54:1160–4.
[54]MaragakisLL,PerlTM.Acinetobacterbaumannii:epidemiology,antimicrobial resistance, and treatment options. Clinical Infectious Diseases 2008;46: 1254–63.
[55]DrapeauCM,GrilliE,PetrosilloN.Rifampicincombinedregimensfor Gram-negativeinfections:datafromtheliterature.InternationalJournalof Antimi-crobialAgents2010;35:39–44.
[56]ZhanelGG,LawsonCD,ZelenitskyS,FindlayB,SchweizerF,AdamH,etal. Comparisonofthenext-generationaminoglycosideplazomicintogentamicin, tobramycin and amikacin. Expert Review of Anti-infective Therapy 2012;10:459–73.
[57]Theuretzbacher U.Accelerating resistance, inadequate antibacterial drug pipelinesandinternationalresponses.InternationalJournalofAntimicrobial Agents2012;39:295–9.
[58]MoutonJW,AmbrosePG,CantonR,DrusanoGL,HarbarthS,MacGowanA, etal.Conservingantibioticsforthefuture:newwaystouseoldandnewdrugs fromapharmacokineticandpharmacodynamicperspective.DrugResistance Updates2011;14:107–17.
