ScienceDirect
www.sciencedirect.comMédecineetmaladiesinfectieuses49(2019)616–620
Short
communication
Non-HACEK
Gram-negative
bacillus
endocarditis
Endocardite
à
bacilles
à
Gram
négatif
non
HACEK
M.
Ertugrul
Mercan
a,
F.
Arslan
b,∗,
S.
Ozyavuz
Alp
c,
A.
Atilla
d,
D.
Seyman
e,
G.
Guliyeva
f,
B.
Kayaaslan
g,
S.
Sari
h,
B.
Mutay
Suntur
i,
B.
Isik
b,
A.
Mert
jaDepartmentofCardiology,FacultyofMedicine,IstanbulAcibademUniversity,Istanbul,Turkey
bDepartmentofInfectiousDiseasesandClinicalMicrobiology,FacultyofMedicine,IstanbulMedeniyetUniversity,Istanbul,Turkey cDepartmentofInfectiousDiseasesandClinicalMicrobiology,FacultyofMedicine,HacettepeUniversity,Ankara,Turkey dDepartmentofInfectiousDiseasesandClinicalMicrobiology,FacultyofMedicine,OndokuzMayisUniversity,Samsun,Turkey
eDepartmentofInfectiousDiseasesandClinicalMicrobiology,HealthSciencesUniversity,AntalyaEducationandTrainingHospital,Antalya,Turkey fDepartmentofInfectiousDiseasesandClinicalMicrobiology,FacultyofMedicine,EgeUniversity, ˙Izmir,Turkey
gDepartmentofInfectiousDiseasesandClinicalMicrobiology,FacultyofMedicine,YildirimBeyazitUniversity,Ankara,Turkey hDepartmentofIntensiveCareUnit,HealthSciencesUniversity,TurkiyeYüksek ˙IhtisasTainingandResearchHospital,Ankara,Turkey iDepartmentofInfectiousDiseaseandClinicalMicrobiology,HealthSciencesUniversity,AdanaNumuneTainingandResearchHospital,Adana,Turkey
jDepartmentofInfectiousDiseasesandClinicalMicrobiology,FacultyofMedicine,IstanbulMedipolUniversity,Istanbul,Turkey
Received18September2018;receivedinrevisedform16November2018;accepted12March2019 Availableonline30March2019
Abstract
Patientsandmethods.–Retrospectiveanalysisofclinicaldatausing26diagnosednon-HACEKGram-negativeinfectiveendocarditiscases fromninehospitalsinTurkey.
Results.–Meanageofpatientswas53(28–84)years,witha23%casefatality.Nineteen(73%)ofthe26patientshadatleastonepredisposing factor.Thepresenceofacentralvenouscatheterwasthemostcommonpredisposingfactor(7/26patients).Pseudomonasaeruginosa(7/26patients) andEscherichiacoli(7/26patients)werethemostcommonpathogens.Themediandurationoftheantibiotictherapywas42days(range3–84 days).Surgicalprocedureswereperformedin10patients.Thecasefatalitywassimilarinpatientswhodidordidnotundergosurgery(20%vs. 25%).
©2019PublishedbyElsevierMassonSAS.
Keywords: Endocarditis;Catheterization;Gram-negativebacteria
Résumé
Patientsetméthodes.–Analyserétrospectivedesdonnéescliniquessur26casd’endocarditeinfectieuseàbacillesàGramnégatifnonHACEK dansneufhôpitauxturcs.
Résultats.–L’âgemoyendelacohortedepatientsétaitde53ans(28–84ans).Lalétalitéétaitde23%.Parmiles26patients,19(73%)avaientau moinsunfacteurprédisposant.Laprésenced’uncathéterveineuxcentralétaitlefacteurprédisposantleplusfréquent(7/26patients).Pseudomonas aeruginosa(7/26patients)etEscherichiacoli(7/26patients)étaientlespathogèneslesplusfréquents.Laduréemédianedel’antibiothérapieétait de42jours(3à84jours).Desinterventionschirurgicalesontétéréaliséeschez10patients.Lalétalitéétaitsimilairechezlespatientsayantounon subiuneinterventionchirurgicale(20%vs25%).
©2019Publi´eparElsevierMassonSAS.
Motsclés:Endocardite;Cathétérisme;BactériesàGramnégatif
∗Correspondingauthor.Ist.MedeniyetUniv,GoztepeEAHastanesi,Dr.Erkincaddesi,34722KadıkoyIstanbul,Turkey.
E-mailaddress:ferhatarslandr@hotmail.com(F.Arslan).
https://doi.org/10.1016/j.medmal.2019.03.013
1. Introduction
Infectiveendocarditis(IE)isanuncommonbutsevere infec-tionassociatedwithhighmorbidityandcasefatality[1].Despite improvementsinhealthcareandtreatmentoptions,thediagnosis andmanagement of IE remain challenging. Viridans strepto-cocciandStaphylococcusaureusarethemostcommonbacterial agentsofIE,followedbyenterococciandHACEKgroup bac-teria[2–5].Gram-negativebacteriararelycauseIEbecauseof their relatively lowendocardial affinity.The frequencyof IE due to non-HACEK Gram-negative bacteria was reported at approximately1%inamultinationalstudy[6].
Intravenous drug use has been considered for decades as themostcommonpredisposingfactorfornon-HACEK Gram-negativeIEuntilrecentstudiesandregistriesreporteddifferent results [6–8]. Increasing vascular catheterization and elderly population may be new risk factors for non-HACEK Gram-negativeIE.
Dataon non-HACEK Gram-negativeIE islimited to case reportsandcaseseries.Thereisaneedfor morestudies con-sideringtheincreasingandchangingepidemiologicalrisk.We aimedtocontributetothereportofclinicalcharacteristics, mana-gement,andoutcomeofnon-HACEKGram-negativeIE.
2. Methods
Weperformedaretrospective,multicenter,andobservational studybetween2007and2016.Adultpatientswithadiagnosis ofnon-HACEKGram-negativeIEwereincludedinthestudy. Individualpatientfilesweresearchedtoobtaindemographicand clinicaldata.Usingemails,weinvitedinfectiousdisease spe-cialiststotakepartinthestudy.Ninetertiaryhospitalsagreedto participate.Hospitalsweresentacaserecorddatafile(Microsoft Excelform)andwereaskedtofillitout.Thefollowingdatawas obtainedforallpatients presentingwithGram-negative endo-carditis: age, gender,known predisposing factors, symptoms, clinicalpresentation,echocardiogramfindings,microbiological testresults,treatment preferences,complications, andhistory of drug use. Participating hospitalswere asked toreport any otherendocarditiscaseobservedoverthepast10yearsandthe proportionofGram-negativeendocarditiscaseswascalculated. Followingobtentionofdatarecords,thediagnosisofpatients wasreviewedaccordingtothemodifiedDukecriteria[9].Ten patientswereexcludedfromthestudy:fourbecauseofalackof echocardiogramfindings,andsixbecauseof theidentification ofendemicpathogens(Brucella)andHACEKgroupbacteria.
Numericaldataispresentedasmean±SDormedian(IQR andrange).Categoricaldataispresentedasnumberand percent-age.
3. Results
3.1. Patients’characteristics
Twenty-sixpatients(53±14years[range:28–84],15[58%] females)presentingwithnon-HACEKGram-negativeIEwere assessed. Nineteen (73%) of them had at least one of the
predisposing factors considered in the study. None of them wereintravenous drugusers. Onepatienthadahistoryof IE. Rheumatic valve disease (n=4) and prosthetic valves(n=2) werethemostfrequentstructuralvalvediseases.Demographic, clinical, and treatment features of the study population are detailedinTable1.
All26caseswereconfirmedasdefiniteendocarditis accord-ingtothemodifiedDukecriteria:26(100%)wereconfirmedby bothmajorbloodculturecriteriaandmajorechocardiographic criteria.ThefrequencyofmodifiedDukecriteriaisdetailedin
Table2.
Intravascularcatheterizationwasobservedinsevenpatients: three patients had a short-term central venous line and four patients hada hemodialysiscatheter. All catheterizationsites weretheinternaljugularvein.Thepresenceofaheartmurmur was detectedin21patients (80%).Themost common symp-tomswerechillsandfever.Themediandurationoffeverbefore diagnosis was 11 days (IQR;5–14 days). None of ourstudy patientshadJanewaylesionsand/orOsler’snodes.Splinter hem-orrhages wereobservedinonepatientandRoth’sspotintwo patients. Erythrocytesedimentation rate(ESR) was measured in22patients.ThemeanESR was67±32and80%of ESRs were 50mm/h. Transthoracic echocardiography (TTE) was performed in all patients, and 20 of them were further eval-uatedwithtransesophagealechocardiography(TEE).Embolic eventswereobservedinsixpatients(23%).Themajorsitefor embolic events was the brain (n=5). One patient developed septic arthritis andonepatient experiencedpulmonary septic emboli.Hematuriawasdetectedin9/25(36%)patients.
Overall, the study included five (19%) patients presenting withprostheticvalveendocarditisand21(81%)patients present-ingwithnativevalveendocarditis.Ofthe26patientspresenting withnon-HACEKGram-negativeIE,eleven(42%)hadmitral vegetation,five(19%)hadtricuspidvegetation,andsix(23%) hadaorticvegetation.Aorticdehiscencewasobservedinthree patients (12%). Oneof the 26 (4%)patients hadintracardiac device-associated(ICD)vegetations.Thelargestvegetationsize was20mmandthemediansizewas11mm(5–20mm). 3.2. Microbiologicaldata
The microbiological spectrum and drug resistance pattern ofnon-HACEK Gram-negativeIEinthe studypopulationare detailedinTable3
.Atleasttwobloodculturesetsweredrawnineachofthe26 patients.Allpatientshadatleastonepositivebloodcultureset (themeannumberofdrawnbloodculturesineachpatientwith apreliminarydiagnosisofIEwas5,andanaverageoffourof themgavepositiveresults).
3.3. Treatmentandoutcomes
IEdiagnosiswasmadeafteramedianofsevendays(IQR: 3–11days)followinghospitalization.Themediandurationofthe antibiotictherapywas42days(range3–75days).Feverusually resolvedwithinsevendaysofaneffectiveantibiotictherapy.
Cardiovascular surgical procedures were performed in 10 patients during hospitalization (valve replacement in nine
M. Ertugrul Mer can et al. / Médecine et maladies infectieuses 49 (2019) 616–620 Table1
Demographical,clinical,andtreatmentfeaturesofGram-negativeendocarditiscases.
Paramètresdémographiques,cliniquesetthérapeutiquesdescasd’endocarditeinfectieuseàbacilleàGram-négatif.
Age,genderPredisposingconditions Pathogens Antibiotics Echocardiogramfindings SurgicalinterventionOutcome
36,M None Pseudomonasmendocinaa Ceftazidime+amikacin Mitralvegetation Yes Alive
41,M None P.aeruginosa Ceftazidime+amikacin Mitralvegetation Yes Alive
51,M Aorticvalvereplacement,mitralvalvereplacement Salmonellaenteritidis Penicilling+gentamicin Mitralvegetation,heartfailure Yes Alive 52,F Mitralvalvereplacement Escherichiacoli Ofloxacin Mitralvegetation No Alive 58,M Atrialseptaldefect,ventricularseptaldefect,prostheticvalve Enterobactercloacae Imipenem/cilastatin Aorticvalvedehiscence,heartfailure No Alive 68,M Centralvenouscatheter,hemodialysis,immunocompromised E.coli Piperacillin-tazobactam Aorticvegetation,heartfailure No Alive 48,F Immunocompromised,livertransplant S.enteritidis Ciprofloxacin Mitralvegetation Yes Alive 67,F Immunocompromised,prostheticvalve Klebsiellapneumoniae Meropenem Aorticvegetation,heartfailure No Death 51,M Centralvenouscatheter,intracardiacdevice E.cloacae Piperacillin/tazobactam+cefepime Vegetationonintracardiacdevice,
heartfailure
No Alive
44,F None E.coli Ceftriaxone Mitralvegetation,heartfailure No Alive
45,M Previousendocarditis Pasteurellamultocida Amoxicillin Valvedehiscence No Alive 58,F Centralvenouscatheter K.pneumoniae Imipenem/cilastatin+amikacin Tricuspidvegetation No Alive
67,F Immunocompromised E.coli Amoxicillin Mitralvegetation No Alive
66,F None E.coli Ceftriaxone Aorticvegetation,heartfailure Yes Alive
33,F None K.pneumoniae Ceftriaxone+gentamicin Mitralvegetation No Death
36,M Centralvenouscatheter,hemodialysis P.aeruginosa Imipenem/cilastatin Tricuspidvegetation Yes Alive 84,F Centralvenouscatheter,hemodialysis E.coli Imipenem/cilastatin Tricuspidvegetation No Death 56,F Centralvenouscatheter,immunocompromised K.pneumoniae Ceftazidime+tigecycline Tricuspidvegetation Yes Alive 42,F Aorticvalvereplacement,mitralvalvereplacement K.pneumoniae Ceftriaxone+gentamicin+rifampicinAorticvegetation No Alive 59,F Centralvenouscatheter,hemodialysis P.aeruginosa Meropenem Tricuspidvegetation No Alive 64,M Post-coronaryangiography P.aeruginosaa Ceftazidime+amikacin Mitralvegetation No Alive
61,M Post-coronaryangiography P.aeruginosaa Ceftazidime+amikacin Mitralvegetation,heartfailure Yes Death
28,F Post-coronaryangiography,aorticvalvereplacement P.aeruginosaa Meropenem+amikacin Aorticvalvedehiscence,heartfailure Yes Death
40,M Rheumaticheartvalvedisease S.enteritidis Ceftriaxone Aorticvegetation,heartfailure Yes Alive
47,F None E.coli Vancomycin+gentamicin Aorticvegetation No Alive
80,F Hemodialysis P.aeruginosa Notavailableb Mitralvegetation No Death
aPreviouslypublishedcasereports. b Thepatientdiedwithin3days.
Table2
FrequencyofDukeCriteriaamong26patientspresentingwithnon-HACEK Gram-negativebacillusendocarditis.
FréquencedescritèresdeDukechez26patientsprésentantuneendocarditeà bacilleàGram-négatifnon-HACEK.
Criteria Frequencyn/n(%)
Majorbloodculturecriteria 26/26(100%)
Majorechocardiographiccriteria 26/26(100%)
Minorcriteria Predisposingfactor 19/26(73%) Fever 20/26(77%) Vascularsigns 6/26(23%) Immunologicalsigns 2/26(8%) Microbiologicalcriteria 0/26(0%) Table3
Microbiologicalspectrumanddrugresistancepatternofnon-HACEK Gram-negativeinfectiveendocarditisinthestudypopulation.
Spectred’activitémicrobiologiqueetrésistancedesendocarditesinfectieusesà bacilleàGram-négatifnon-HACEKdelapopulationàl’étude.
Bacteria Frequency,n(%) Drugresistancepattern
E.coli 7(26) 6/7pan-susceptible,1/7MDR P.aeruginosa 7(27) 7/8pan-susceptible,1/8MDR P.mendocina 1(3) K.pneumoniae 5(19) 2pan-susceptible,2MDR,1XDR S.enteritidis 3(10) Pan-susceptible E.cloacae 2(8) 2MDR P.multocida 1(4) Pan-susceptible MDR:multidrug-resistant;XDR:extensivelydrug-resistant.
patients andvalverepairinonepatient). Themedian timeto surgeryafterinitialdiagnosis of IEwas 24days (range5–54 days).Five(50%)of10patientswhounderwentsurgical proce-dureshadpositivevalve(vegetation)cultures(P.aeruginosain twopatients).Sixpatients(23%)diedduringthehospitalization period.Casefatality(2 of 10patients withmedical treatment onlyvs.4of16patientswithsurgicalandmedicaltreatment) didnotdiffersignificantly.
Sixhospitalswereabletoprovidethenumberofendocarditis cases.Amongallendocarditiscases,the proportionof Gram-negativeendocarditiswas1%.
4. Discussion
Endocarditisisachronicinfectionresultingfromthe inter-action between bacterial virulence (adherence, biofilm) and endocardial endothelial surface [10]. Non-HACEK Gram-negativebacteriadonotusuallycauseendocarditisduetotheir limitedabilitytoformbiofilmsandlowaffinitytoendocardial endothelium, except for Salmonella spp. [11]. A small num-berofstrainsofnon-HACEKGram-negativeagentscapableof formingbiofilms mightbethe causativeagents of endocardi-tiswith thehelp of predisposing factors(valveabnormalities andprostheticmaterials)that can facilitatethe persistence of the microorganism within the vegetation [12]. Pseudomonas speciescancausecommunity-acquiredandnosocomial endo-carditis [13]. In our study three P. aeruginosa endocarditis casesdevelopedafter coronaryangiographicintervention due
to contamination of the contrast medium in a tertiary hos-pital. Salmonella species are also rare pathogens responsible forinfectiveendocarditis[14].Whileenvironmentalpathogens arepan-susceptibletoantimicrobials,Klebsiellaendocarditisis associatedwithahigherriskofdrugresistance.
The most prominent risk factor for non-HACEK Gram-negative IE observed in our study was determined by the presence of an intravascular catheter, as recently reported
[5–7]. Permanent central venous catheters (e.g., hemodialy-sis catheters)are associated with ahigher risk of developing Gram-negativeendocarditis,especiallyinimmunocompromised patients.Previousstudiesreportedintravenousdruguseasthe most prevalentriskfactorfor non-HACEK Gram-negativeIE
[7,8].Intravenousdrugusewasnotobservedinourstudy pop-ulation. Onthe basisof underlyingvalvediseases(rheumatic valvedisease,degenerativevalvedisease,andthepresenceof prosthetic valve), the predisposing condition distribution was similartotheresultsofpreviousstudies[5–7].
The dynamic epidemiological changes in both host- and pathogen-related factorsarenewchallenges forclinicians. IE patientsarenowolderandhavemorecomorbidconditions[15]. AccordingtotheInternationalCollaborationon Endocarditis-ProspectiveCohortStudy(ICE-PCS)results,healthcarecontact andimplantationofendovasculardevicesareprimaryrisk fac-torsfornon-HACEKGram-negativeIE[5,6].Seventeenpatients in our study had underlying diseases that related to at least one surgical (valve replacement, catheterization) or medical (chemotherapy,hemodialysis)intervention.
Thecasefatalityobservedinourstudywassimilarbetween patients who did or did not undergo surgical therapy (20% vs.25%).Thisresultsuggeststhatsurgicaltreatmentmaynot alwaysbeanabsoluteindicationasmentionedinprevious stud-ies[5,7].Ourin-hospitalcasefatalitywassimilartotheresults ofpreviouslarge-scalestudies[5,7].
5. Conclusion
Cliniciansmayencounternon-HACEKGram-negative infec-tiveendocarditis,mostlyinmedicalwardsasmanyintravascular interventionsareperformedandaselderlypatientsaremanaged inthosewards.
Disclosureofinterest
Theauthorsdeclarethattheyhavenocompetinginterest.
References
[1]MoreillonP,QueY-A.Infectiveendocarditis.Lancet2004;363:139–49.
[2]FederspielJJ.IncreasingUS ratesofendocarditiswithStaphylococcus aureus:1999–2008.ArchInternMed2012;172:363.
[3]Cahill TJ, Prendergast BD. Infective endocarditis. Lancet 2016;387:882–93.
[4]PericásJ,ZboromyrskaY,CerveraC,etal.Enterococcalendocarditis revisited.FutureMicrobiol2015;10:1215–40.
[5]MorpethS,MurdochD,CabellCH,etal.Non-HACEKgram-negative bacillusendocarditis.AnnInternMed2007;147:829–35.
[6]MurdochDR,ClinicalPresentation.Etiology,andoutcomeofinfective endocarditisinthe21stcentury:theInternationalCollaborationon Endo-carditis–ProspectiveCohortStudy.ArchInternMed2009;169:463.
[7]LoubetP,LescureF-X,LepageL,etal.Endocarditisduetogram-negative bacilliataFrenchteachinghospitalovera6-yearperiod:clinical charac-teristicsandoutcome.InfectDisLondEngl2015;47:889–95.
[8]BaddourLM.Infectiveendocarditis:diagnosis,antimicrobialtherapy,and managementofcomplications:astatementforhealthcareprofessionals fromtheCommitteeonRheumaticFever,Endocarditis,andKawasaki Dis-ease,CouncilonCardiovascularDiseaseintheYoung,andtheCouncilson ClinicalCardiology,Stroke,andCardiovascularSurgeryandAnesthesia, AmericanHeartAssociation:EndorsedbytheInfectiousDiseasesSociety ofAmerica.Circulation2005;111:e394–434.
[9]LiJS,SextonDJ,MickN,etal.ProposedmodificationstotheDukeCriteria forthediagnosisofinfectiveendocarditis.ClinInfectDis2000;30:633–8.
[10]ParsekMR,SinghPK. Bacterialbiofilms:anemerginglinkto disease pathogenesis.AnnuRevMicrobiol2003;57:677–701.
[11]SanchezCJ,MendeK,BeckiusML,etal.Biofilmformationby clini-calisolatesandtheimplicationsinchronicinfections.BMCInfectDis 2013;13:47.
[12]Keren I, Shah D, Spoering A, et al. Specialized persister cells and themechanismof multidrugtolerancein Escherichiacoli.JBacteriol 2004;186:8172–80.
[13]ReyesMP,AliA,MendesRE,etal.ResurgenceofPseudomonas Endo-carditisinDetroit,2006–2008.Medicine2009;88:294–301.
[14]ChengW-L,LiC-W,LiM-C,etal.Salmonellainfectiveendocarditis.J MicrobiolImmunolInfect2016;49:313–20.
[15]CahillTJ,BaddourLM,HabibG,etal.Challengesininfectiveendocarditis. JAmCollCardiol2017;69:325–44.