Liver abscess in the tropics: experience in the University Hospital









LUNG ABSCESS AND BRONCHIAL CATHETERIZATION

On March 5 1951
pdf?md =d add f a e c b ed cafe e&pid= s . S main


More Analogies between Charlieu and Anzy-le-Duc

since the excavation of Cluny II by Kenneth John Conant with apses in echelon. The two churches apse. They were built by monasteries belonging to two.


Liver abscess in the tropics: experience in the University Hospital

Recently ultrasonography has been increasingly utilized. Table II shows the location of abscesses in the various subgroups. Of the amoebic abscesses 86.5% were 
.full


The Representation of Christ in the Apses of Early Christian Churches

FIGURE 2. Rome Santa Pudenziana





ABSTRACT Introduction : Le pronostic d'une réparation tendineuse

Le pronostic d'une réparation tendineuse en zone II après réparation des fléchisseurs profonds et superficiels reste médiocre . Nous avons comparé sur main 


Characterization of the APSES-family transcriptional regulators of

7 août 2018 Histoplasma capsulatum yeast cells were inoculated into 96-well plates (2 × 106 yeasts/mL) containing HMM and serial dilutions of each stress ...
foy


Contrôle sérologique des syphilis I et II après traitement à la pénicilline

Contrôle sérologique des syphilis I et II après traitement à la pénicilline. C. Truffat. N. Hunziker. W. Jadassohn. Travail de la Clinique universitaire de 


GENITOURINARY COMPLICATIONS OF GRANULOMATOUS

ii-No.z. 297. GENITOURINARY. COMPLICATIONS. OF. GRANULOMATOUS. BOWEL. DISEASE* formation between the intesti- nal tract and urinary tract. II. Abscess.





Surgical management of pyogenic liver abscess

We observed 5 pulmonary complications 5 wound in- fections and 2 perihepatic collections. The average hospital stay was 11.2 days. We observed only two deaths 


LE SENTIMENT AMÉRICAIN PENDANT LA GUERRE: II: APRÈS LA

II (D. APRÈS LA RUPTURE. Le président Wilson qui était


216808Liver abscess in the tropics: experience in the University Hospital

PostgraduateMedicalJournal(1987)63,551-554

TropicalMedicine

Liverabscessinthetropics:experienceintheUniversityHospital,KualaLumpur. K.L.Goh,N.W.Wong,M.Paramsothy,M.NojegandK.Somasundaram DepartmentofMedicine,FacultyofMedicine,UniversityofMalaya,59100KualaLumpur,Malaysia.

Summary:Wereviewed204casesofliverabscessseenbetween1970and1985.Ninetywerefoundtobeamoebic,24pyogenicandonetuberculous.Thecauseoftheabscessesintheremaining89patientswasnotestablished.Thepatientswerepredominantlymale,Indians,andinthe30-60agegroup.Themajorityofpatientspresentedwithfeverandrighthypochondrialpain.Themostcommonlaboratoryfindingswereleucocytosis,hypoalbnminaemiaand anelevatedserumalkalinephosphatase.Amoebicabscessesweremainlysolitarywhilepyogenicabscessesweremainlymultiple.Complicationswerefewinourpatientsandincludedruptureintothepleuralandperitonealcavitiesandsepticaemicshock.Anoverallmortalityof2.9%wasrecorded.Thedifficultyindiagnosingtheabscesstypeishighlighted.Thesinglemostimportanttestinhelpingusdiagnoseamoebicabscess,presumablythemostcommontypeofabscessinthetropics,istheEntamoebahistolyticaantibodyassay.Thistestshouldbeusedmorefrequentlyinthetropics.

Introduction

LiverabscessisafairlycommondiseaseinMalaysia.Balasegaram'reported442casesseenovera15-yearperiod.Whileitcanbesaidthathisexperienceisthatofareferralsurgicalcentre,manycasesofliverabscessarealsoseenandtreatedinsmallerhospitalsthrough-outthecountry.Thecommontypesofliverabscessseenareamoebicandpyogenicabscess.Althoughitisoftenpresumedthatthemajorityofabscessesseeninthetropicsareamoebicinorigin,theevidenceincriminatingEntamoebahistolyticaasthecausativeagenthasseldombeenproven.2Similarly,culturesforpyogenicorganismsareoftennegative.Wedescribeourexperiencewith204casesofliverabscessseenfrom1970-1985intheUniversityHosp-ital,KualaLumpur.

Materialsandmethods

Adefinitivediagnosisofliverabscesswasmadebasedononeoranycombinationofthefollowing-surgicaldrainageoraspirationofpusfromtheliveroratautopsy,radionuclidescanning,ultrasonographyorcomputedtomographic(CT)scanning.Patientswereconsideredtohaveanamoebicliverabscesswhen(i)E.histolyticatrophozoitesweredemonstratedinabiopsyoftheabscesswall;(ii)E.histolyticatrophozoitesweredemonstratedinthepus;(iii)E.histolyticaantibodytitresweregreaterthan1:64;(iv)anchovysaucewasdrainedfromtheabscess;(v)E.histolyticatrophozoitesorcystswereisolatedinthestoolswithorwithoutahistoryofdysentery;and(vi)patientsrespondedtotreatmentwithanti-amoebicdrugsalone.Patientswereconsideredtohaveapyogenicabscesswhenpusand/orbloodculturesgrewpyogenicorgan-isms,providedtheabovecriteriaforamoebicabscesswerenegativeandtherewasnootherobvioussourceofinfection.Onepatientwasdiagnosedtohaveatuberculousabscesswhenbiopsyoftheabscesswalldemonstratedgranulomaswithacid-fastbacilli.Therewasnoevidenceofpulmonaryormiliarytuberculosis.Patientswereplacedinanindeterminategroupwhenthecriteriafordiagnosisofamoebicorpyogenicabscesswerenotpresentorknown.

A)TheFellowshipofPostgraduateMedicine,1987Correspondence:K.L.Goh,M.B.,B.S.(Mal),M.R.C.P.(UK)Accepted:29January1987copyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from

552K.L.GOHetal.

TableIClinicalpresentation

AmoebicPyogenicIndeterminateAllabscessesn=90n=24n=89n=204

Feverwithchillsandrigors80.079.288.882.7Righthypochondrialpain77.750.078.768.8Epigastricpain14.433.314.620.8Diarrhoeaordysentery19.98.315.914.7Jaundice24.416.717.019.4Tenderhepatomegaly90.087.590.989.8Rightbasaleffusionorcrackles/39.98.332.927.0rub

*Figuresareinpercentagesoftheabscesstype.

Results

Of204patientsreviewed,90(44.1%)werefoundtohaveamoebicabscess,24(11.8%)pyogenicabscess,

one(0.5%)tuberculousabscesswhile89(43.6%)wereconsideredindeterminate.Themajorityofpatientsweremale,withamaletofemaleratioof6:1.Indianswerethepredominantracialgroupaffectedandcomprised49.5%ofallpatientscomparedwith23.2%ofallhospitaladmis-sions(P<0.01).Theageofpatientsrangedfrom2

yearsto84years.Thepeakincidenceoccurredinthe30-60yearsage-group.Theclinicalpresentationofliverabscessisasshown(TableI).Thecommonpresentingclinicalfeaturesforallabscesstypeswerefeverwithchillsandrigors,righthypochondrialpainandatenderhepatomegaly.Only

asmallproportion(20%)ofpatientshadjaundice.Rightlowerchestsigns,chieflyarightbasaleffusion

wereseenin40%ofthepatientswithamoebicabscessandin32.9%ofpatientsintheindeterminategroup.Theincidenceis,however,muchlowerinthepyogenic

group(8.3%).Thehaemoglobinlevelwaslessthan11g/I00mlin53.5%ofthepatientsandthetotalwhitecellcount

waselevatedtoabove11x109/lin81%.Alowserumalbuminleveloflessthan35g/landanelevatedserumalkalinephosphataseofgreaterthan150IU/lwerefoundinabout80%ofpatients.Serumbilirubinandalaninetransaminaselevelswereraisedin29%and24%respectively.Therewasnosignificantdifferencebetweenthepatientgroups.Themostfrequentlyusedradiologicalmodalityindiagnosisandlocalizationofabscesswasradionuclidescanning,utilizedinthree-quartersofthepatients.Recently,ultrasonographyhasbeenincreasinglyutilized.TableIIshowsthelocationofabscessesinthevarioussubgroups.Oftheamoebicabscesses86.5%

weresingle,withthemajorityintherightlobe(77.7%ofthetotal).Ontheotherhand62.6%ofpyogenicabscessesweremultiple.Thefiguresfortheindeter-minategroupapproximatedthatofamoebicabsces-

ses.ThebacteriologyofpyogenicabscessisasshowninTableIII.EscherichiacoliandKlebsiellaspecieswerethemostcommonpyogenicoganismsandtogetheraccountedformorethanhalfoftheorganismsisolated.Predisposingfactorswerelookedforinthe24patientswithpyogenicabscess.In6,therewasevidenceofbiliarydiseasewhiletherewerenoobvious

TableIIAbscesslocation*

AmoebicPyogenicIndeterminate

SingleabscessRightlobe77.725.064.0Leftlobe8.812.513.5 MultipleabscessesRightlobe8.84.27.9Leftlobe1.14.23.4Bothlobes3.354.211.2

*Allfiguresinpercentagesoftheparticularabscesstype.copyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from

LIVERABSCESSINTHETROPICS553

TableIIIBacteriologyofpyogenicabscess

No.(%)

Escherichiacoli833.3Klebsiellasp.625.0Enterobactersp.312.5Streptococcusmilleri28.3Serratiamarcesens28.3Streptococcusfaecalis,Bac-1eachteroidessp.Citrobactersp.Pseudomonaspseudomallei

precipitatingfactorsintheremainder.Intheindeter-minategroup,abscessesfollowedaperforatedappen-dix,asepticabortionandin2patients,biliarydisease.Nostandardtreatmentprotocolwasadoptedinourpatients.Themajorityweretreatedwithbothantibioticsandanti-amoebicdrugsassoonasthediagnosisofliverabscesswasmade.Thepatientswereusuallyreferredearlyforasurgicalopinion.Theindicationsforneedleaspirationorsurgicaldrainageincludednon-resolutionofclinicalsignsandsymp-

tomswithadequatemedicaltreatment,signsofrup-

tureorimpendingruptureoftheabscessandalargesizeofabscessexceeding10cm,especiallywhenlocatedintheleftlobe.Intheamoebicabscessgroup,44.4%ofpatientshadopendrainageand17.7%hadneedleaspirationoftheirabscessescarriedout.Inthepyogenicgroup,70.8%ofabscessesweresurgicallydrainedand8.3%needleaspirated.However,inthelargegroupwithindeterminateabscess27.0%ofpatientshadopendrainageand13.5%,needleaspirationperformed.Complicationswereuncommoninourgroupofpatients;7patientshadperforationsintotheperiton-ealcavity,4hadperforationsintothepleuralcavity.Twopatientswithamoebicabscessdevelopedseverefulminatingcolitisandsepticaemicshockandsuccum-bedtotheillness;bothpatientswereaborigines.Threepatientsinthepyogenicgroupandoneintheindeter-minategroupdiedfollowingsepticaemicshock.Theoverallmortalityratewas2.9%.

Discussion

Whilethediagnosisofliverabscessisreliablymadeonclinicalgroundsandeasilyconfirmedbyradionuclidescanning,ultrasonographyorCTscanning,thediag-nosisofabscesstypeisdifficult.Thisisreflectedbythelargenumberofpatientsplacedintheindeterminate

group.Thereareseveralreasonsforthisproblem.Notallabscessesaredrainedoraspiratedyieldingpusforexamination.Evenifpuswereavailable,amoebacanonlybelocatedwithdifficulty.Biopsyoftheabscesswallisnotalwaysdoneatopendrainage,andagain,ifdone,amoebamaynotbefoundbecausetheydisin-tegrateanddisappearquicklywithprecedinganti-amoebictreatment.3Pusculturesforpyogenicorgan-ismsareaffectedbypreviousantibiotictherapy-oftenthecasewithourpatientswhohavesoughttreatmentbeforecomingtohospital.Furthermore,pusculturesforanaerobicorganismsarenotroutinelydone.Thisispertinentastherearereportsthatanaerobesmayaccountforupto38%ofallpyogenicabscesses.4Associatedorprecedingdysenteryinpatientswithamoebicliverabscessisuncommonwithourpatientsandsodoesnothelpinmakingadiagnosisofamoebicabscess.Anchovysaucewhenaspiratedordrainedishighlycharacteristicofamoebicabscess.However,itmaynotbefoundinallamoebicabscesses.Entamoebahistolyticaantibodyassayisusefulinconfirmingthediagnosisofamoebicabscessastissueinvasionbytheamoebainducesastrongantibodyresponse.5Theindirectimmunofluorescentmethodisusedinourlaboratoryandhasbeenfoundtobehighlysensitive,withareactivityof96%amongconfirmedcasesofliverabscess.6Unfortunatelythistestisstillnotreadilyavailablelocally.Only4seraweretestedintheindeterminategroup;2werenon-reactivewhile2hadtitresof1in16whichwereconsiderednotsignificant.Itislikelythereforethatmanycasesofamoebicabscesswereleftundiagnosedintheindeter-minategroup.Thesexandracedistributionofourgroupofpatientswasverysimilartootherreportsfromthisregion.8'9Malepreponderanceisnotedinourseries.Indianswerethepredominantracialgroupaffectedoverall.ThisisnotsurprisingasthemajorityofIndianpatientsseekingtreatmentinthehospitalcomefromoutlyingareasaroundKualaLumpur,wherehygieneispoor.Clinicalsignsandsymptomswerefairlyreliableandconsistentinhelpingusdiagnoseliverabscess.However,theclinicalfeaturesofamoebicandpyogenicabscesseswereonthewholeremarkablysimilaranddidnothelpusdistinguishbetweenthetwo.Anexceptionistheoccurrenceofrightbasallungsignsin40%ofpatientswithamoebicabscesswhichwasonlypresentin8.3%ofpatientswithpyogenicabscess.Thiscanperhapsbeexplainedbythefactthatthemajorityofamoebicabscesseswereintherightlobeoftheliverwhereitwouldbemostlikelytocauserightbasallungsignsespeciallywhenlocatedinthediaphragmaticsurface.Thehigherproportionofabscessesintherightlobealsoexplainsthehigherratioofrighthypochondrialpaintoepigastricpaininamoebicabscess.Ontheotherhand,inpyogenicabscess,whereasignificantproportionofabscesseswereintheleftlobeorinbothlobes,epigastricpainwaspresentinanincreasedproportionofpatients.Wehaveobservedthatthenumberandlocationofcopyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from

554K.L.GOH,etal.

abscesseswereusefulinhelpingtodistinguishbetweenamoebicandpyogenicabscess.Asinglerightlobeabscesswasmorelikelytobeamoebicwhilemultipleabscessesinbothlobesweremorelikelytobepyogenic.Thesefindingsareconsistentwiththeobservationsofothers.8'9Notunexpectedly,theGram-negativebacilli,E.coliandKlebsiellawerethemostcommonpyogenicorganismsisolated.Bacteroideswasfoundinonlyonepatientbutasnotedbefore,thedetectionofanaerobicorganismsishamperedbycollectionandinadequateanaerobicculturingtechnique.Onepatienthadsys-temicmelioidosiswithdevelopmentofmultipleliverabscesses;Pseudomonaspseudomalleiwasculturedfrombothbloodandpus.Whileassociatedpredisposingfactorswerefoundinonlysixpatientswithpyogenicabscess,wefoundfourpatientswithdefiniteevidenceofbiliarydiseaseorintra-abdominalsepsisintheindeterminategroup.Thesepatientsarelikelytohavehadpyogenicabsces-ses.Theoverallmorbidityandmortalityratesinourpatientswerelowwhichcouldperhapsbeattributedtoapromptandaggressiveapproachtotreatment.Perforationsintotheperitonealandpleuralcavitiesbutnotintothepericardialsacweredocumented.Itisinterestingtoreporttwocasesoffulminatingamoebiccolitis,anuncommoncomplicationofamoebiasis,occurringwithliverabscess.Deathsamongourpatientsusuallyfollowedoverwhelmingsepticaemia.Finally,wehavealsonotedtheremarkablesimilarityintheepidemiologicaldata,clinicalfeatures,andlaboratoryandradiologicalresultsoftheamoebicandindeterminategroups.Itwouldbetemptingtopresumethatthemajorityofthepatientsintheindeterminategroupwereamoebic.Asamoebicserologicaltestshavebeenshowntobehighlysen-sitive,amoreconclusivediagnosisofamoebicliverabscesscouldhavebeenmadeifthesetestshadbeenperformedonmorepatients.Thiswouldhavebeenextremelyhelpfulinplanningmorepreciseanddefin-itivetreatmentofthedisease,therebycuttingdownoncostsandundesirablesideeffectsoftherapy.

Acknowledgements

WewouldliketothankMrLowTing,DepartmentofMedicine,UniversityofMalaya,forhelpinguspreparethemanuscript.

References

1.Balasegaram,M.Managementofhepaticabscess.CurrProblSurg1981,18:282-340.2.Vijendran,M.Thediagnosisandcurrenttreatmentofliverabscess.MedJMalaysia1977,32:133-138.3.Manson-Bahr,P.E.C.&Apted,F.I.C.Amoebiasis,giar-diasisandbalantidiasis.InManson'sTropicalDiseases.BailliereTindall,Edinburgh,London,1982,pp.121-145.4.Perera,M.R.Presentation,diagnosisandmanagementofliverabscess.Lancet1980,ii:629-632.5.Krupp,M.I.Antibodyresponseinintestinalandextrain-testinalamoebiasis.AmJTropMedHyg1970,19:57-62.6.Thomas,V.,Sinniah,B.&Yap,P.L.Assessmentofthesensitivity,specificity,andreproducibilityoftheindirectimmunofluorescenttechniqueforthediagnosisofamebiasis.AmJTropMedHyg1980,30(1):57-62.7.Fock,K.M.,Ong,N.T.&Chua,K.L.Liverabscess:atwo-yearstudy.SingaporeMedJ1985,26:350-353.8.Teh,L.B.,Ng,H.S.,Kwok,K.C.etal.Liverabscess-aclinicalstudy.AnnAcadMedSingapore1986,15(2):176-000.9.May,R.P.Difficultiesindifferentiatingamoebicfrompyogenicliverabscess.ArchIntMed1967,119:69-74.copyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from

PostgraduateMedicalJournal(1987)63,551-554

TropicalMedicine

Liverabscessinthetropics:experienceintheUniversityHospital,KualaLumpur. K.L.Goh,N.W.Wong,M.Paramsothy,M.NojegandK.Somasundaram DepartmentofMedicine,FacultyofMedicine,UniversityofMalaya,59100KualaLumpur,Malaysia.

Summary:Wereviewed204casesofliverabscessseenbetween1970and1985.Ninetywerefoundtobeamoebic,24pyogenicandonetuberculous.Thecauseoftheabscessesintheremaining89patientswasnotestablished.Thepatientswerepredominantlymale,Indians,andinthe30-60agegroup.Themajorityofpatientspresentedwithfeverandrighthypochondrialpain.Themostcommonlaboratoryfindingswereleucocytosis,hypoalbnminaemiaand anelevatedserumalkalinephosphatase.Amoebicabscessesweremainlysolitarywhilepyogenicabscessesweremainlymultiple.Complicationswerefewinourpatientsandincludedruptureintothepleuralandperitonealcavitiesandsepticaemicshock.Anoverallmortalityof2.9%wasrecorded.Thedifficultyindiagnosingtheabscesstypeishighlighted.Thesinglemostimportanttestinhelpingusdiagnoseamoebicabscess,presumablythemostcommontypeofabscessinthetropics,istheEntamoebahistolyticaantibodyassay.Thistestshouldbeusedmorefrequentlyinthetropics.

Introduction

LiverabscessisafairlycommondiseaseinMalaysia.Balasegaram'reported442casesseenovera15-yearperiod.Whileitcanbesaidthathisexperienceisthatofareferralsurgicalcentre,manycasesofliverabscessarealsoseenandtreatedinsmallerhospitalsthrough-outthecountry.Thecommontypesofliverabscessseenareamoebicandpyogenicabscess.Althoughitisoftenpresumedthatthemajorityofabscessesseeninthetropicsareamoebicinorigin,theevidenceincriminatingEntamoebahistolyticaasthecausativeagenthasseldombeenproven.2Similarly,culturesforpyogenicorganismsareoftennegative.Wedescribeourexperiencewith204casesofliverabscessseenfrom1970-1985intheUniversityHosp-ital,KualaLumpur.

Materialsandmethods

Adefinitivediagnosisofliverabscesswasmadebasedononeoranycombinationofthefollowing-surgicaldrainageoraspirationofpusfromtheliveroratautopsy,radionuclidescanning,ultrasonographyorcomputedtomographic(CT)scanning.Patientswereconsideredtohaveanamoebicliverabscesswhen(i)E.histolyticatrophozoitesweredemonstratedinabiopsyoftheabscesswall;(ii)E.histolyticatrophozoitesweredemonstratedinthepus;(iii)E.histolyticaantibodytitresweregreaterthan1:64;(iv)anchovysaucewasdrainedfromtheabscess;(v)E.histolyticatrophozoitesorcystswereisolatedinthestoolswithorwithoutahistoryofdysentery;and(vi)patientsrespondedtotreatmentwithanti-amoebicdrugsalone.Patientswereconsideredtohaveapyogenicabscesswhenpusand/orbloodculturesgrewpyogenicorgan-isms,providedtheabovecriteriaforamoebicabscesswerenegativeandtherewasnootherobvioussourceofinfection.Onepatientwasdiagnosedtohaveatuberculousabscesswhenbiopsyoftheabscesswalldemonstratedgranulomaswithacid-fastbacilli.Therewasnoevidenceofpulmonaryormiliarytuberculosis.Patientswereplacedinanindeterminategroupwhenthecriteriafordiagnosisofamoebicorpyogenicabscesswerenotpresentorknown.

A)TheFellowshipofPostgraduateMedicine,1987Correspondence:K.L.Goh,M.B.,B.S.(Mal),M.R.C.P.(UK)Accepted:29January1987copyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from

552K.L.GOHetal.

TableIClinicalpresentation

AmoebicPyogenicIndeterminateAllabscessesn=90n=24n=89n=204

Feverwithchillsandrigors80.079.288.882.7Righthypochondrialpain77.750.078.768.8Epigastricpain14.433.314.620.8Diarrhoeaordysentery19.98.315.914.7Jaundice24.416.717.019.4Tenderhepatomegaly90.087.590.989.8Rightbasaleffusionorcrackles/39.98.332.927.0rub

*Figuresareinpercentagesoftheabscesstype.

Results

Of204patientsreviewed,90(44.1%)werefoundtohaveamoebicabscess,24(11.8%)pyogenicabscess,

one(0.5%)tuberculousabscesswhile89(43.6%)wereconsideredindeterminate.Themajorityofpatientsweremale,withamaletofemaleratioof6:1.Indianswerethepredominantracialgroupaffectedandcomprised49.5%ofallpatientscomparedwith23.2%ofallhospitaladmis-sions(P<0.01).Theageofpatientsrangedfrom2

yearsto84years.Thepeakincidenceoccurredinthe30-60yearsage-group.Theclinicalpresentationofliverabscessisasshown(TableI).Thecommonpresentingclinicalfeaturesforallabscesstypeswerefeverwithchillsandrigors,righthypochondrialpainandatenderhepatomegaly.Only

asmallproportion(20%)ofpatientshadjaundice.Rightlowerchestsigns,chieflyarightbasaleffusion

wereseenin40%ofthepatientswithamoebicabscessandin32.9%ofpatientsintheindeterminategroup.Theincidenceis,however,muchlowerinthepyogenic

group(8.3%).Thehaemoglobinlevelwaslessthan11g/I00mlin53.5%ofthepatientsandthetotalwhitecellcount

waselevatedtoabove11x109/lin81%.Alowserumalbuminleveloflessthan35g/landanelevatedserumalkalinephosphataseofgreaterthan150IU/lwerefoundinabout80%ofpatients.Serumbilirubinandalaninetransaminaselevelswereraisedin29%and24%respectively.Therewasnosignificantdifferencebetweenthepatientgroups.Themostfrequentlyusedradiologicalmodalityindiagnosisandlocalizationofabscesswasradionuclidescanning,utilizedinthree-quartersofthepatients.Recently,ultrasonographyhasbeenincreasinglyutilized.TableIIshowsthelocationofabscessesinthevarioussubgroups.Oftheamoebicabscesses86.5%

weresingle,withthemajorityintherightlobe(77.7%ofthetotal).Ontheotherhand62.6%ofpyogenicabscessesweremultiple.Thefiguresfortheindeter-minategroupapproximatedthatofamoebicabsces-

ses.ThebacteriologyofpyogenicabscessisasshowninTableIII.EscherichiacoliandKlebsiellaspecieswerethemostcommonpyogenicoganismsandtogetheraccountedformorethanhalfoftheorganismsisolated.Predisposingfactorswerelookedforinthe24patientswithpyogenicabscess.In6,therewasevidenceofbiliarydiseasewhiletherewerenoobvious

TableIIAbscesslocation*

AmoebicPyogenicIndeterminate

SingleabscessRightlobe77.725.064.0Leftlobe8.812.513.5 MultipleabscessesRightlobe8.84.27.9Leftlobe1.14.23.4Bothlobes3.354.211.2

*Allfiguresinpercentagesoftheparticularabscesstype.copyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from

LIVERABSCESSINTHETROPICS553

TableIIIBacteriologyofpyogenicabscess

No.(%)

Escherichiacoli833.3Klebsiellasp.625.0Enterobactersp.312.5Streptococcusmilleri28.3Serratiamarcesens28.3Streptococcusfaecalis,Bac-1eachteroidessp.Citrobactersp.Pseudomonaspseudomallei

precipitatingfactorsintheremainder.Intheindeter-minategroup,abscessesfollowedaperforatedappen-dix,asepticabortionandin2patients,biliarydisease.Nostandardtreatmentprotocolwasadoptedinourpatients.Themajorityweretreatedwithbothantibioticsandanti-amoebicdrugsassoonasthediagnosisofliverabscesswasmade.Thepatientswereusuallyreferredearlyforasurgicalopinion.Theindicationsforneedleaspirationorsurgicaldrainageincludednon-resolutionofclinicalsignsandsymp-

tomswithadequatemedicaltreatment,signsofrup-

tureorimpendingruptureoftheabscessandalargesizeofabscessexceeding10cm,especiallywhenlocatedintheleftlobe.Intheamoebicabscessgroup,44.4%ofpatientshadopendrainageand17.7%hadneedleaspirationoftheirabscessescarriedout.Inthepyogenicgroup,70.8%ofabscessesweresurgicallydrainedand8.3%needleaspirated.However,inthelargegroupwithindeterminateabscess27.0%ofpatientshadopendrainageand13.5%,needleaspirationperformed.Complicationswereuncommoninourgroupofpatients;7patientshadperforationsintotheperiton-ealcavity,4hadperforationsintothepleuralcavity.Twopatientswithamoebicabscessdevelopedseverefulminatingcolitisandsepticaemicshockandsuccum-bedtotheillness;bothpatientswereaborigines.Threepatientsinthepyogenicgroupandoneintheindeter-minategroupdiedfollowingsepticaemicshock.Theoverallmortalityratewas2.9%.

Discussion

Whilethediagnosisofliverabscessisreliablymadeonclinicalgroundsandeasilyconfirmedbyradionuclidescanning,ultrasonographyorCTscanning,thediag-nosisofabscesstypeisdifficult.Thisisreflectedbythelargenumberofpatientsplacedintheindeterminate

group.Thereareseveralreasonsforthisproblem.Notallabscessesaredrainedoraspiratedyieldingpusforexamination.Evenifpuswereavailable,amoebacanonlybelocatedwithdifficulty.Biopsyoftheabscesswallisnotalwaysdoneatopendrainage,andagain,ifdone,amoebamaynotbefoundbecausetheydisin-tegrateanddisappearquicklywithprecedinganti-amoebictreatment.3Pusculturesforpyogenicorgan-ismsareaffectedbypreviousantibiotictherapy-oftenthecasewithourpatientswhohavesoughttreatmentbeforecomingtohospital.Furthermore,pusculturesforanaerobicorganismsarenotroutinelydone.Thisispertinentastherearereportsthatanaerobesmayaccountforupto38%ofallpyogenicabscesses.4Associatedorprecedingdysenteryinpatientswithamoebicliverabscessisuncommonwithourpatientsandsodoesnothelpinmakingadiagnosisofamoebicabscess.Anchovysaucewhenaspiratedordrainedishighlycharacteristicofamoebicabscess.However,itmaynotbefoundinallamoebicabscesses.Entamoebahistolyticaantibodyassayisusefulinconfirmingthediagnosisofamoebicabscessastissueinvasionbytheamoebainducesastrongantibodyresponse.5Theindirectimmunofluorescentmethodisusedinourlaboratoryandhasbeenfoundtobehighlysensitive,withareactivityof96%amongconfirmedcasesofliverabscess.6Unfortunatelythistestisstillnotreadilyavailablelocally.Only4seraweretestedintheindeterminategroup;2werenon-reactivewhile2hadtitresof1in16whichwereconsiderednotsignificant.Itislikelythereforethatmanycasesofamoebicabscesswereleftundiagnosedintheindeter-minategroup.Thesexandracedistributionofourgroupofpatientswasverysimilartootherreportsfromthisregion.8'9Malepreponderanceisnotedinourseries.Indianswerethepredominantracialgroupaffectedoverall.ThisisnotsurprisingasthemajorityofIndianpatientsseekingtreatmentinthehospitalcomefromoutlyingareasaroundKualaLumpur,wherehygieneispoor.Clinicalsignsandsymptomswerefairlyreliableandconsistentinhelpingusdiagnoseliverabscess.However,theclinicalfeaturesofamoebicandpyogenicabscesseswereonthewholeremarkablysimilaranddidnothelpusdistinguishbetweenthetwo.Anexceptionistheoccurrenceofrightbasallungsignsin40%ofpatientswithamoebicabscesswhichwasonlypresentin8.3%ofpatientswithpyogenicabscess.Thiscanperhapsbeexplainedbythefactthatthemajorityofamoebicabscesseswereintherightlobeoftheliverwhereitwouldbemostlikelytocauserightbasallungsignsespeciallywhenlocatedinthediaphragmaticsurface.Thehigherproportionofabscessesintherightlobealsoexplainsthehigherratioofrighthypochondrialpaintoepigastricpaininamoebicabscess.Ontheotherhand,inpyogenicabscess,whereasignificantproportionofabscesseswereintheleftlobeorinbothlobes,epigastricpainwaspresentinanincreasedproportionofpatients.Wehaveobservedthatthenumberandlocationofcopyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from

554K.L.GOH,etal.

abscesseswereusefulinhelpingtodistinguishbetweenamoebicandpyogenicabscess.Asinglerightlobeabscesswasmorelikelytobeamoebicwhilemultipleabscessesinbothlobesweremorelikelytobepyogenic.Thesefindingsareconsistentwiththeobservationsofothers.8'9Notunexpectedly,theGram-negativebacilli,E.coliandKlebsiellawerethemostcommonpyogenicorganismsisolated.Bacteroideswasfoundinonlyonepatientbutasnotedbefore,thedetectionofanaerobicorganismsishamperedbycollectionandinadequateanaerobicculturingtechnique.Onepatienthadsys-temicmelioidosiswithdevelopmentofmultipleliverabscesses;Pseudomonaspseudomalleiwasculturedfrombothbloodandpus.Whileassociatedpredisposingfactorswerefoundinonlysixpatientswithpyogenicabscess,wefoundfourpatientswithdefiniteevidenceofbiliarydiseaseorintra-abdominalsepsisintheindeterminategroup.Thesepatientsarelikelytohavehadpyogenicabsces-ses.Theoverallmorbidityandmortalityratesinourpatientswerelowwhichcouldperhapsbeattributedtoapromptandaggressiveapproachtotreatment.Perforationsintotheperitonealandpleuralcavitiesbutnotintothepericardialsacweredocumented.Itisinterestingtoreporttwocasesoffulminatingamoebiccolitis,anuncommoncomplicationofamoebiasis,occurringwithliverabscess.Deathsamongourpatientsusuallyfollowedoverwhelmingsepticaemia.Finally,wehavealsonotedtheremarkablesimilarityintheepidemiologicaldata,clinicalfeatures,andlaboratoryandradiologicalresultsoftheamoebicandindeterminategroups.Itwouldbetemptingtopresumethatthemajorityofthepatientsintheindeterminategroupwereamoebic.Asamoebicserologicaltestshavebeenshowntobehighlysen-sitive,amoreconclusivediagnosisofamoebicliverabscesscouldhavebeenmadeifthesetestshadbeenperformedonmorepatients.Thiswouldhavebeenextremelyhelpfulinplanningmorepreciseanddefin-itivetreatmentofthedisease,therebycuttingdownoncostsandundesirablesideeffectsoftherapy.

Acknowledgements

WewouldliketothankMrLowTing,DepartmentofMedicine,UniversityofMalaya,forhelpinguspreparethemanuscript.

References

1.Balasegaram,M.Managementofhepaticabscess.CurrProblSurg1981,18:282-340.2.Vijendran,M.Thediagnosisandcurrenttreatmentofliverabscess.MedJMalaysia1977,32:133-138.3.Manson-Bahr,P.E.C.&Apted,F.I.C.Amoebiasis,giar-diasisandbalantidiasis.InManson'sTropicalDiseases.BailliereTindall,Edinburgh,London,1982,pp.121-145.4.Perera,M.R.Presentation,diagnosisandmanagementofliverabscess.Lancet1980,ii:629-632.5.Krupp,M.I.Antibodyresponseinintestinalandextrain-testinalamoebiasis.AmJTropMedHyg1970,19:57-62.6.Thomas,V.,Sinniah,B.&Yap,P.L.Assessmentofthesensitivity,specificity,andreproducibilityoftheindirectimmunofluorescenttechniqueforthediagnosisofamebiasis.AmJTropMedHyg1980,30(1):57-62.7.Fock,K.M.,Ong,N.T.&Chua,K.L.Liverabscess:atwo-yearstudy.SingaporeMedJ1985,26:350-353.8.Teh,L.B.,Ng,H.S.,Kwok,K.C.etal.Liverabscess-aclinicalstudy.AnnAcadMedSingapore1986,15(2):176-000.9.May,R.P.Difficultiesindifferentiatingamoebicfrompyogenicliverabscess.ArchIntMed1967,119:69-74.copyright. on May 19, 2023 by guest. Protected byhttp://pmj.bmj.com/Postgrad Med J: first published as 10.1136/pgmj.63.741.551 on 1 July 198

7. Downloaded from