IJGE Issue 4Vol 1 2003Evaluation of Serological tests

OriginalArticle

Evaluation of Serological tests forthe diagnosis of

Helicobacterpylori infection

Haitham I Baqir; MB. ChB. PhD*; Maysaa HAl-Aubaidi; Bsc. MSc**; Saad Fakhri; FICMS***; Maiada MAl- Mousili Ph.D**; SalimAHamadi Ph.D**

Abstract

Atotalof58outpatientsreferredfor endoscopic evaluation of gastroduodenal symptoms were included in this study. Biopsy specimens were taken from the gastric antrum of each patient. Samples weretestedforthepresenceofH.pyloriby standard biopsy related tests (urease, histology,andculture)whichareconsidered as gold standard methods for H. pylori detection.Serafrom thesepatientswere testedforanti-H.pyloriantibodies by enzyme-linked-immunoassay, immuno- chromatogrphy,andlatexagglutinationtest fortheevaluationofperformanceindicesof thesetechniques.

Sensitivity,specificity,positiveand

Introduction

H.PyloriisaGram-negative,spiralshaped, microaerophilicbacillusthatresidesbeneathand within the mucouslayer of thegastric mucosa and produce multiple enzymes such as urease and mucolytic proteases that are important for its

survivalandforitspathogeniceffect(1).

negativepredictivevaluesandaccuracyof eachtestwerecalculatedrelativetooneor moreofthe“goldstandard”.

Atotalof45patientsgavepositive results for thepresence of H.pylori by two ormoreofthesetestsused.

Theother13samplesshowed negative results by all three tests used. Serological tests showsensitivities ranging from95.5%forELISAtechniqueto80%for latexagglutinationtest.Specificityranges from76.9%inELISAtechniqueto69.2% bylatexagglutinationmethod.

Serologicaltestscanprovidea reliablenoninvasivemethodsfordetection ofH.pyloriinfection.

cancerandgastriclymphoma(3).

Standarddiagnostictestreliesongastric biopsy.Ofthesetestsureasetestisveryreliable, sensitive, specific,inexpensiveandsimple.This testisdonebytransferringoneorpreferablytwo biopsies into urea containing test medium that detectsthepresenceofureasebyalkalinizationthat

Infectionisalmostacquiredinchildhood

resultsfrom cleavage of urea

(4,5).Histological

andthe main risk factor for infection is poor

socioeconomiccondition(2).

examinationof routinely stained gastric biopsy

couldhavesimilarsensitivityandspecificityby

Infection isalmostalwaysassociatedwith

experiencedpathologist

(1,4).Cultureisthemost

nonulcerdyspepsia,histologicchronic(typeB) gastritisandamajorriskfactorforthedevelopment of peptic ulceration, atrophic gastritis, gastric

laborious,tediousandexpensivedetectionmethod.

Even under most favorable conditions, the

sensitivityofcultureisbetween70-80%(4).Culture

*Dr.HaithamIBaqir;Centralpublichealthlaboratory.Baghdad,Iraq.

**Dr.MaysaaHAl-Aubaidi;Dr.MaiadaMAl-Mousili;SalimAHamadi;Collegeofpharmacy,Baghdad,Iraq.

***Dr.SaadFakhir;Al-Nahraincollegeofmedicine,Baghdad,Iraq.

41

HaithamIBaqir

IJGEIssue4Vol12003

shouldbeconservedforspecialcircumstancesaswhen antibioticresistanceissuspected(1,3,6).

H. pylori does not only lead to a strong

inflammatoryresponseofthegastricmucosabutalso inducesaprofoundspecifichumoralimmunereaction. ThepresenceofH.pyloriinfectioncanthusbereliably diagnosed by detecting IgG and IgA antibodies directedagainstspecificH.pyloriantigens.

Manyserodiagnostictestsareavailablebased onthe detectionofIgGclassantibodiesversusthis organism.Someofthesetestsareclaimedtobealmost equivalenttothoseof histologyandbiopsyurease

testing (7,8,9,10,11). Othersshowpoorcorrelationbetween thepresenceofH.pyloriinfectionandtheantibody response(12).

MaterialsandMethods

Fiftyeightpatientsattendingtheendoscopy unitofAl-Kademiateachinghospitalwithdifferent typesofgastriccomplaintswereenrolledinthisstudy.

Bloodsampleswerecollectedbeforeendo- scopy.Gastricantralbiopsyspecimensweretaken.

Patientsagedlessthan18years;patientswho had takenantibioticsorprotonpumpinhibitorsor bismuth preparationsin theprevious fourweeks were excludedfromthestudy.

Thebloodspecimenscollectedwereallowedto clotandtheserawereseparated.Theserawerefrozen andstoredat-20oCuntilrequired.

Antralbiopsyspecimenswerecollectedfor cultureofH.pylori,histologyandureaseproduction.

Culture

Biopsyspecimensforcultureweretransported to bacteriological laboratory in sterile brain heart

infusionbrothandwerekeptinacoolbagor4oCuntil cultured. The specimens were processed within a limited time of not more than four hours.Antral biopsies were crushed on sterile glass slides, homogenizedwithsterileneedlesandthenculturedon brainheartinfusionagarcontaining7% horseblood,

0.25% yeast extract and Campylobacter selective supplement (skirrow-Oxoid SR 69) containing vancomycin,polymyxinandtrimethoprim.ThepH was adjusted to 6.8-6.9.Plateswereincubatedin microaerophilicenvironmentgeneratedbygaspack

(GenerbagMicroaer,BioMerieux45531)at37oCfor uptosevendays.SuspectedcoloniesofH.pyloriwere identifiedbyGramsstaining,catalaseandoxidasetest. ConfirmationoftheisolatewasdonebyAPIcampy system(BioMerieux).Subculturingwasdoneinbrain heartinfusionbroth-filledcontainers,incubatedfor3 daysundermicroaerophilicconditions(13,14).

Microaer,BioMerieux45531)at37oCforuptoseven days.SuspectedcoloniesofH.pyloriwereidentified by Grams-staining, catalase and oxidase test. ConfirmationoftheisolatewasdonebyAPIcampy system(BioMerieux).Subculturingwasdoneinbrain heartinfusionbroth-filledcontainers,incubatedfor3 daysundermicroaerophilicconditions(13,14).

Histology

Hematoxylinand Eosin stain was used by pathologistsforidentificationofthebacteriainthe

biopsyspecimens(13,14).

Ureasetest

PresumptiveevidenceofthepresenceofH. pylori inbiopsymaterialwasobtainedbyplacinga portionofthecrushedtissuebiopsymaterialdirectly into urea containing agar which was prepared as follows:4.6gmoftheureaagarbasesuspendedin190

mLdistilledwater,autoclavedat115oCfor20minutes, thencooledto50oCbeforeasepticallyadding10mLof

40%w/vureasolution,mixedwell,distributedinto

sterilecontainersandallowedtosetatslopes.

Apositivetestmanifestedbycolorchanges

(yellowtopink)duetoalkalinizationofmediais

consideredindicativeoftheorganismpresence(13,14).

Serology

Serumspecimensweretestedforanti-H.pylori antibodies using commercially available kits. Techniques included were latex agglutination, immunochromatography, andenzymelinked immunosorbentassay.Thesensitivities,specificities, positive andnegativepridictivevaluesofthosekits wereevaluated.ThedetectionofH.pyloriinantral biopsy specimens by culture, histology, urease productionoranycombinationof thosetestswere consideredasthe“goldstandard”.

Latex test:thePyloriDryLatextest(Orion Diagnostics)containslatexparticlessensitizedwithH. pyloriantigen.H.pyloriantibodiesifpresentinthe serumsamples willreactwiththesensitizedlatex resultinginvisuallydetectableclumps.

Immunochromatography(BiosignH.pylori WB)isaonestepimmunochromatographictestforthe detectionofantibodiestoH.pyloriinhumanserum. Themethod employsacombinationofanti-human immunoglobulindye conjugate(colloidalgold)and highlypurifiedH.pyloriproteins.Asthesampleflows through the absorbent device, the anti-human immunoglobulindyedconjugatebindtothe human IgG antibodiesforming anantigen antibodycomplex. ThiscomplexbindstoH.pyloriproteinsfixedinthe zone(B)andproducesacoloredbandIntheabsenceof

42

IJGEIssue4Vol12003

EvaluationofSerologicaltests

(Colloidalgold)andhighlypurifiedH.pyloriproteins. Asthesampleflowsthroughtheabsorbentdevice,the anti-humanimmunoglobulindyedconjugatebindto thehumanIgGantibodiesforminganantigenantibody complex.Thiscomplex bindstoH.pyloriproteins fixedinthezone(B)andproducesacoloredbandInthe absence of anti- H. pylori antibodies, there is no coloredbandinthetestzone(B).Thereactionmixture continuesflowingthroughtheabsorbentdevicetothe controlzone(C).Unboundconjugatebindstothe reagent fixedinthecontrolzone(C),producinga coloredband,indicatingtheproperperformanceofthe test.

Enzymelinkedimmunosorbentassay(Bio-Hit, Finland) The test is based on sandwich enzyme immunoassay technique with purified H. pylori bacterial antigen adsorbed on microwell plate and detection antibody labeled with horse radish peroxidase.

Results

Fiftyeightpatientsparticipatedinthisstudy. Theiragesrangedfrom18-62yearswithanaverageof

34.7years.

Fortyfive ofthe 58patients werepositive for H.pyloribyoneormoreofthe“goldstandard”tests (culture, histology and direct urease test). The remaining13werenegativeforH.pyloribyallthe threetests.Thepatternoftheseresultsareshownin (table1).

Immunodiganostictestweredoneonseraof

threetests.Thepatternoftheseresultsareshownin

(table1).

Immunodiganostictestsweredoneonseraof those patients. Rapid latex test could confirm the infectionin36casesofthosewhowerepositivebythe standardinvasivetechniques.Itmissedthediagnosisin

9casesandgaveafalsepositivereactionin4casesthus givingasensitivityof80%andaspecificityof69.2%.

Immunochromatographic technique could detect40casesoftheprovedcases.Otherindicesare shownintable2.

ValueoftheELISA systemwascalculatedas

Enzymeimmunounits(EIU).

EIUwascalculatedastheabsorptionat450 dividedbytheabsorptionofapositivecontrol.Values exceeding30EIUwereconsideredaspositive(Cut-off valueof30wasusedandaccordingtomanufacturer instructions).Withthiscut-offvalue43casesoutofthe

45positivecasesbytheinvasivetechniquegavea positivereaction.Ontheotherhandseraofthreeofthe patientswhoshowednegativeresultsbythestandard proceduresyieldedpositiveserologicaltest.Thusthe serumIgGELISAhada sensitivityof95.5%anda specificityof76.9%.

Detailsofperformanceindicesofthedifferent techniquesareshownintable2.

Discussion

Ulcerdiseaseisaninfectiousdisease(15,16).Ifthe infectionisdiagnosedandtreated,ulcerdiseasecanbe cured.AndasthepathogenicroleofH.pyloriinulcer

Table 1:The results of biopsy related tests for the detection of H. pylori infection in

gastricantral biopsy.

43

MaysaaHAl-AubaidiIJGEIssue4Vol12003

Table 2: Evaluation of the performance of serological tests in comparison with gold

standardbiopsy related tests.

Gold

Sensitivity

Specificity

PPV

NPV

Overall

Serologicaltestsstandard%%

+-

%%accuracy %

(ELISA)+43395.576.9

-210

+404

-5988.869.2

93.5

91

83.3

64.3

91.3

84.4

Latextest+3648069.2

-99

9050

77.5

Table 3: Changes in performance characteristics of Elisa with different cut-off values.

Cut-offSensitivitySpecificityPPVNPVOverall

value%%%

%accuracy %

101001580.3

2095.553.887.7

3095.576.993.5

4077.784.694.5

10081

77.786.2

83.391.3

52.379.3

5062.292.396.541.3689

6055.510010039.365.5

44

IJGEIssue4Vol12003

EvaluationofSerological tests

120-

100-

80-

60-

40-

-120

-100

-80

-60

-40

20-

0-

Sensitivity

Specificity

-20

-0

102030405060

Cut-offvalues(U/ml)

Fig.l

Changesin the sensitivity and specificity with different cut- off values for the ELISAassay.

Discussion

Ulcerdiseaseisaninfectiousdisease(15,16).If theinfectionisdiagnosedandtreated,ulcerdisease canbecured.AndasthepathogenicroleofH.pylori in ulcer disease and other upper gastro-intestinal conditions isestablished,testing oftheorganism gainswideracceptance.

Though endoscopy provides means of obtaining the organism for culture, screening for refluxeosophigitisandpossiblestomachcancer,itisa costy and unpleasant for the patient. Moreover,

is acostyandunpleasantforthepatient.Moreover, cultureoftheorganismisdifficulttoperformandwas notevenevaluatedinmanylargestudies.

Inthis study, out of the 45 total positive biopsies,only10werepositivebyculture,i.e22.2%. Itshouldbekeptinmindthatnegativeresultsdoesno exclude the presence ofH. pylori infection, although isolationbyofthemicroorganismbyculturecertainly indicates it's presence. Many factors could have contributed to thisreduced sensitivity ofthis method inourstudy.

45

SaadFakhriIJGE Issue 4Vol 1 2003

thisreducedsensitivityofthismethodinourstudy.

1.Onlyonebiopsyfromeachpatientwastakenfor culture. Using more than one biopsy from differentgastricsitescouldhaveraisedthenumber ofpositiveresults.

2.Itiswellknownthatthebacteriumisslowgrowing andfastidiousanditispossiblethatsomeH.pylori strainswillnotformcoloniesonsomecurrently

availablemedia,liketheonewehaveused(17).

3. Patients' ingestion of topical anesthetic, semithicone,priortreatmentwithantibiotics,H2- receptorantagonistsorprotonpumpinhibitorscan

reducetheviabilityofbacteria(18).

4.The useof abundantamounts ofgluteraldehyde in sterilization of the endoscope might have deleteriouseffectonthebacterium.

Histological examination gave acceptable results.Thisispossiblyduetoexaminationofantral biopsieswherebytheantrumismoreaffectedthanthe

body(19).

Unlikebiopsyrelatedtests,serologicaltests can detectsystemicimmunologicalresponsetoH. pyloriinfectionwhicheffectivelysamplethewhole stomach.

Manyserologicaltestswereintroducedasnon invasivealternative.Thesestudiesgaveconflicting resultsregardingdifferentserologicaltestsavailable forH.pyloridiagnosis.

We attempted to evaluatethree commercially availablekitswithdifferenttechniques,namelylatex test, immunochromato-graphic test and enzyme- linkedimmunosorbentassay.

Noninvasive serological testsare as accurate indicatorsofH.pyloristatusastheinvasivetest. However, latex agglutination technique and immunochromatographyarelessaccurateandless specific than the ELISAtest but its ease of use, convenience, lower cost, more rapid results and availability inprimary caremake ituseful forpatient screening.

Sera of dyspeptic patients with negative reactionbythegoldstandardcriteriashowedpositive reactionbyserologicaltestinvariablepercentages.

6%, 10%, 11% were positive by ELISA, immunochromatographyandlatextestrespectively. ThismeansthatserologicalevidenceofH.pyloriwas greaterthan the prevalenceofinfectionbybiopsy relatedtests.Suchanobservationcouldbeduetothe followingpossiblecauses.

1.H.pyloriinfectioninthestomachmaybepatchy possibly due to metaplasiaor regrowth after failed

ofH. pylori was greater than the prevalence of infectionbybiopsyrelatedtests.Suchanobservation couldbeduetothefollowingpossiblecauses:-

1.H.pyloriinfectioninthestomachmaybepatchy possibly due to metaplasiaor regrowth after failed eradication.Suchconditionscouldbedetectedby

serologicaltestsmoreproperly(2).

2.Biopsyspecimensampleonlyaverysmallpartof thestomachwhereasantibodydetectionmethods effectivelysamplethewholestomach.

3.AntibodyagainstH.pyloriremaindetectablefor manymonthsaftereradication.

Sofalsepositiveresultsbyimmunological testsmaybefalsenegativeresultsbythegoldstandard criteria.

Latexagglutinationtestshoweda performancecharacteristicsthatwerelowerthanthe other two tests which could be attributed to the

detectionlimit of this test (0.006-0.06 ug/ml)(20). Though immunochromatography had better performance than latex test, itis still less than that of theELISAtechniqueandthisispossibleduetothe lacking of amplification effect of enzyme immunoassay(detectionlimitofELISAis<0.0001-

0.01ug/ml)(21).

One of the problems we faced was the calculationofthecut-offvalue,sincethelattermust bedeterminedforeachassaybasedontheprevalence ofthemicroorganisminthepopulation.Tillnowthere are no epidemiological studies concerning the seroprevalenceofH.pyloriantibodiesamongIraqi people,thereforethecut-offvaluesuggestedbythe manufacturer was used. Moreover, changes in performancecharacteristicsofELISA withdifferent cut-offvalueswasstudied.Maximumaccuracywas obtainedatacut-offvalueof30.Table(3)andfigure (1)showtherelationofcut-offvalueandperformance indices.

Positive predictive value and negative predictivevaluearedependantontheprevalenceof

theorganismwithinaparticularpopulation(22).

Conclusions

Fromthisstudy,onecanconcludethatnon- invasiveserologicaltestsareconvenientfordiagnosis ofH. pyloriinfectionduetoitsgoodperformance characteristicsandsimplicityofthetechniques.These tests vary in performance indices, with ELISA techniquehavingthebestover

allaccuracy, followed by immunochromatography and latex test.

References

46

IJGE Issue 4Vol 1 2003

Evaluation of Serological tests

indices,withELISAtechniquehavingthebestover allaccuracy,followedbyimmunochromatography andlatextest.

References

1.Udaya BS, Prakash MD, Thomas MH, and

AbermanMD.Myointernalmedicineboard

review.4th ed.Lippincott Williamsand Wilkins

Publications.NewYork.2000.

12.DharR,MustafaAS,DharPM,KhanMS,Al- RashidiFJ,Al-ShamaliAA,AliFH.Evaluation and comparison of two immunodiagnostic assaysforHelicobacterpyloriantibodieswith culture results. Diagn-Microbiol-Infect-Dis.

1998;30:1-6.

13.MurrayPR,Baron,EJO,PfallerMA, Tenoves

FC and Yolken RH. Manual of clinical

2.GoodmanKJ,andCorreaP.Transmissionof

Helicobacter pylori among sibling. Lancet.

microbiology. 7th

WashingtonD.C.

ed. 1999. ASM press.

2000;355:358-62.

3. Greenwood D, Slack RCB and Peutherer.

14.ForbesBA,SahmDF,WeissfeldAS.Bailyand

Scott'sdiagnosticmicrobiology.10th ed.1998.

Medical microbiology. 5th

ed. ELST with

Mosby(London).

chrchilllivingstonepublication(London).1997.

4.SuerbaumS.Helicobacterpylori,microbiology, virulence factorsandclinicalmanifestations. BiotestBulletin.1995;5:115-26.

5.GoodwinCS,MersallMM,andNorthfield TC.

Helicobacter pylori infection. Lancet.

1997;349:256-69.

6.GlupozynskiY.Microbiologicalandserologioal diagnostic tests for Helicobacter pylori: an overview.Br-Med-Bull.1998:54:175-86.

7.Luthra GK,DiNuzzoAR,Gourley WK, Crowe SB. Comparison of biopsy and serological methods of diagnosis of Helicobacter pylori infectionandthe potentialroleofantibiotics. Am-J-Gastroenterol.1998;93:1291-6.

8.0ksanenA,VeijolaL,SipponenP,Schauman KO,RautelinH.EvaluationofPylorisetScreen, a rapid whole-blood diagnostic test for Helicobacterpyloriinfection.J-Clin-Microbiol.

1998;36:955-7.

9.AndersenLP,KiileriokS,PedersenG,Thoreson

15.PeuraDA.Helicobacterpylori:adiagnostic

dilemma and a dilemma of diagnosis. Gastroenterology.1995;109:313-5.

16.GrahamDY.TherapyofHelicobacterpylori:

current status and issues. Gastroenterology.

2000.118;S2-S8.

17.MegraudF,BonnetF,GamierM,Lamouliatte H.Characterization of campylobaoteri- opyloridisby culture,enzymaticprofileand proteincontent.JClin Microbiol.1985:22:

1007-10.

18.UtierAF,HavstadS,MaCK,BlaserMJ,Perez- perez GI, and Schubert TT. Accuracy of invasive and non invasive tests to diagnose Helicobacter pylori infection. Gastroenter- ology.1995;109:136-41.

19.DixonMF,GentaRM,YardleyJHandCorreaP

. Classification and grading of gastritis. The updatedSydneysystem.AmJSugPathol;1996;

201:1161-81.

20.Goldsby RA, Kinat TJ, and Osborne BA.

AC,JorgensenF,RathJ,LarsenNE,Borup0,

KUBY immunology, 4th

ed. 2000. W.H.

KrogteltK,ScheibelJ,RuneS.Ananalysisof seven different methods to diagnose Helicobacter pylori infections. Scand-J- Gastroenterol.1998;33:24-30.

10.FaigelDO,MagaretN,CorlessC,Lieberman DA,FennertyMB.Evaluationofrapidantibody testsforthe diagnosisofHelicobacterpylori infection.Am-J-Gastroenterol.2000;95:72-7.

11.van-DerEnde A,van-DerHulstRW,RoordaP, Tytgat GN, Dankert J. Evaluation of three commercial serological tests with different methodologies to assess Helicobacter pylori infection.J-Clin-Microbiol.1999;37:4150-2.

Freemanandcompany(NewYork).

21. Jackson TM, and Ekins, RP. Theoretical limitations on immunoassay sensitivity. J ImmunolMethods.19986;87:13-20.

22.KangG,RajanDP,PatraS,ChackoA,Mathan MM.Useofserology,ureasetest,andhistology indiagnosisofHelicobacter pylori infection insymptomaticandasymptomaticIndianJMed Res.1999;110:86-90.

47