Clin Infect Dis. 2010 Dorman S173 7

SUPPLEMENT ARTICLE

NewDiagnosticTestsforTuberculosis:Bench,Bedside,andBeyondSusan E. DormanJohnsHopkinsUniversityCenterforTuberculosisResearch,Baltimore,Maryland

Currenttoolsandstrategiesfordiagnosisoftuberculosis(TB)areinadequate,particularlyinsettingswithahighprevalenceofhumanimmunodeficiencyvirus(HIV)infection.Severalpromisingnewtoolsareatadvancedstagesofdevelopmentandevaluation.Thisreviewdescribessomeofthosepromisingnewtechnologiesandthekeybarrierstotheireffectiveimplementation.

Robert Koch’s recognition in 1882 of Mycobacteriumtuberculosisasthemicrobialcauseoftuberculosis(TB)led shortly thereafterto theidentificationofmethodstostainbacilliinclinicalspecimens,renderingtheor-ganismsvisiblewithuseoflightmicroscopy.Suchwasthe birth of TB diagnostics and of diagnostic micro-biologyingeneral.Tragically,developmentandimple-mentation of TB diagnostics kept pace neither withmedicaltechnologynorwiththecatastrophicexplosionofTB,includingdrug-resistantTB,inthewakeoftheglobal human immunodeficiency virus (HIV) pan-demic.Inadequatetoolsandweaksystemsforlabora-tory-baseddiagnosisofactiveTBhavecontributedto(1) underdiagnosis of disease, leading to individualmorbidity and mortality and to continued transmis-sion;(2)overdiagnosisofdisease,leadingtounneces-sarytreatmentwithattendantconsequencestothepa-tient and inappropriate resource utilization by thehealthcareprogram;and(3)delayeddiagnosisofdrugresistance,leadingtoacquisitionofadditionalresistanceandtomorbidityandtransmission.Besidesdeficienciesindiagnostictools,accesstoTBdiagnosticscontinuestobeamajorchallenge.However,notableadvancesinTBdiagnostictechnologieshavebeenmadeinthepastseveral years, and the potential exists for translating

thesedevelopmentsintomeaningfulimprovementsinglobalTBclinicalcareandcontrol.ThisreviewfocusesontypesoftechnologiescurrentlyintheTBdiagnosticspipeline and identifies areas of progress and gaps inknowledgethatarerelevanttomovingthefieldforward.The interferon-g release assays, which are principallyfordetectionofM.tuberculosisinfection,werereviewedrecentlyandarenotincludedhere[1].

DIAGNOSTICTOOLSUSEDROUTINELYFORDIAGNOSISOFACTIVETB

Acidfaststainingofclinicalmaterial,followedbysmearmicroscopy, remains themostfrequentlyusedmicro-biologicaltestfordetectionofTB.Themajordrawbackofsputumsmearmicroscopyisitspoorsensitivity,es-timatedtobe∼70%inarecentsystematicreview[2].However,thesensitivityofsputumsmearmicroscopyisclearlylessinmanyfieldsettingsandmaybeaslowas ∼35% in some settings with high rates of TB

andHIV coinfection [3]. Compounding the limitation ofpoortestsensitivityisinadequateorabsenttestqualityassurance in some resource-constrained settings, fur-therdrivingdowntheyieldofmicroscopy,drivingupthelaboratoryworkloadasmoresputumtestsperpa-tient are performed in an effort to reach a diagnosis,andincreasingdiagnosticdelayandpatientlosstofol-low-up.Drugsusceptibilitystatuscannotbeascertainedfromsmearmicroscopy.

Reprints or correspondence: Dr Susan E. Dorman, Johns Hopkins University

CenterforTuberculosisResearch,CRB2,1550OrleansSt,Rm1M-06,Baltimore,MD,21212(DSusan1@JHMI.edu).

The HIV pandemic has brought into focus the in-adequacy, from individual and public health perspec-tives,ofsputumsmearmicroscopyasthecornerstoneof TB diagnosis in low- and middle-income settings.

ClinicalInfectiousDiseases 2010;50(S3):S173–S177ᮊ2010bytheInfectiousDiseasesSocietyofAmerica.Allrightsreserved.1058-4838/2010/5010S3-0014$15.00DOI:10.1086/651488

NewDiagnosticTestsforTB • CID 2010:50 (Suppl3) • S173

HIVinfectiondramaticallyincreasestheincidence,severity,andmortalityriskofactiveTB[4,5,6].Atthepublichealthlevel,failuretodiagnosediseaseinalargeproportionofHIV-infectedpatients with smear-negative TB maycontributetotransmis-sion [7] and further stresses health care and/or personal re-sources,becauseindividualsremaininthehealthcaresystemwithoutcorrectdiagnosisandtreatmentorexitthesystemandprobablydie.

genceofpublic-privatepartnershipsinvolvedinglobalhealth,engagementinTBdiagnosticsdevelopmentandevaluationhasincreasedamongpublichealthandacademicgroups,govern-ment funding agencies, and importantly, the industry sector.Theresulthasbeenanexpansioninthenumberofpromisingdiagnostictestsunderdevelopment,including2newtests(inadditiontoliquidculture)thathavebeenendorsedforusebytheWorldHealthOrganization.Table1listssomeofthemorepromisingnewtechnologiesandtestscurrentlyindemonstra-tionorlate-stageevaluationphase[11,12].

Culture of M. tuberculosis in clinical specimensissubstan-

tially more sensitive than smear microscopy. Culture can beperformed using solid media, such as Lowenstein-Jensen, orliquidmedia,suchasthatusedincommerciallyavailableau-tomatedsystems.Untiltherecentadventofmoleculartestsfordrugresistance(describedinthenextsection),isolationofM.tuberculosiswithuseofculturewasaprerequisiteforsubsequentphenotypicdrug-susceptibilitytesting.TheAchillesheelofcul-tureisthelongtimetoresults(10–14daysforliquidcultureand 3–4 weeks for solid culture), which is a consequence ofthelongdoublingtimeofM.tuberculosis.Currentlyavailableculture methods are technically demanding, require imple-mentationofbiosafetypracticesandequipmenttopreventin-advertentinfectionoflaboratorypersonnel,andhaverelativelyhighper-testprices.TheGlobalTuberculosisReport2008doc-umentsthestunninglackofculturefacilitiesinthegovernmenthealthsectorinmostdevelopingcountries,someofwhichhaveasinglelaboratorythatmayormaynotbewellresourcedorqualityassured[8].

BEYONDNEWTECHNOLOGY

Successfulimplementationofnewtoolswilldependonmorethantechnologicalinnovation(Figure1).Attheresearchlevel,rigorous implementation of well-designed, bias-minimizedstudies and complete and accurate reportingare essentialforappropriate decision making by the health care communitychargedwithimplementingtestsforindividualpatientevalu-ation or recommending tests for TB program use. The Stan-dards forReportingof DiagnosticAccuracyInitiativehasde-veloped standards and tools for improving the quality ofreportingofdiagnosticaccuracystudies,tobestallowthereaderto detect the potential for bias in a study and to assess thegeneralizabilityandapplicabilityofstudyresults[13].Assess-ment of test impactonrelevantclinicaloutcomesshouldbe-come a routine component of late-stage evaluation research.Diagnostic accuracy is arguably just a surrogate for patient-and public health–oriented outcomes. Economic assessmentsare important during the continuum of diagnostics research.Duringdevicedevelopment,costestimatescanguidetheneedfor device modifications to facilitate use in settings with thegreatestneed.Later,duringtheevaluationanddemonstrationphase,cost-effectivenessandcost-benefitanalysescanprovideinformation critical to policy development and implementa-tion.Operationsandhealthsystemsresearchisalsoneededtounderstandhowtoeffectivelyimplementnewtoolsinrelevantsettingswhereexistingaccesstoanddeliveryofhealthcareareweak.

Tuberculinskintestingusingpurifiedproteinderivativeandchest radiographs are used as adjuncts to smear microscopy(andculture,ifavailable)insomesettings;however,theformerhavepoorsensitivityandspecificityforactiveTB,andthelatterare often not available at the point of primary patient care.Trials of antibiotics directed against commonbacterialpneu-monia pathogens areoftenrecommendedinTBprogramdi-agnostic algorithms but are also fraught with problems andmayleadtolengthydiagnosticdelay.

NEWDIAGNOSTICTECHNOLOGIESANDTOOLS

Newprogrammaticapproaches,includingrevisedclinicalal-

gorithms for TB diagnosis, may be needed to maximize theimpactofnewtools.Forexample,shouldrapidmoleculartestsfordrugresistancebeperformedforallpersonswithsuspectedTBduringinitialevaluation,bereservedforuseintheinitialevaluationonlyofpersonswithsuspectedTBwithriskfactorsfor drug resistance, or be used in some other place in a di-agnostic algorithm? In populations with a high prevalence ofHIV infection, should urine-based antigen detection tests beused solely for evaluation of symptomatic persons with sus-pectedTB,orshouldtheyalsoplayaroleinroutinescreeningofHIV-infectedpersons[14]?Todate,mostTBdiagnostictestdevelopmenthasfocusedonmaximizingsensitivityandspec-

There is a clear need for development, introduction, and ef-fective implementation of cost-effective new tools that con-tributetoimprovementinpatient-centeredoutcomesandpub-lic health and that perform well for HIV-infected andHIV-uninfectedindividuals.TheStopTBPartnershipWorkingGrouponNewTBDiagnosticshasplacedpriorityonaccurate,simplenewtoolsforTBcasedetection,rapididentificationofdrug-resistantTB,andreliabledetectionoflatentTBinfection[9].

AgainstabackdropofincreasedneedfueledbyHIVinfectionanddrug-resistantTB,advancesinbiology(includingtheso-lutionoftheM.tuberculosisgenomein1998[10]),andemer-

S174 • CID 2010:50 (Suppl3) • Dorman

Figure 1. Components of the post–research-and-development process for promising new tuberculosis (TB) diagnostic technologies. QA, qualityassurance.

ificitytoruleinorconfirmaTBdiagnosis.Ontheotherhand,atestwithanexceedinglyhighnegativepredictivevaluemighthaveuseinrulingoutTBand,thereby,allowingefficienttriageof patients and resources; such a test would require carefulassessmenttodetermineitsoptimaluseinclinicalalgorithms.

Laboratory capacity needs to bestrengthened,especiallyin

resource-limitedsettings.Althoughsomeaspectsoflaboratorystrengtheningwillvary accordingtothecharacteristicsofthenewtests,thereare,nevertheless,generalunmetneeds,includ-ingthosefortrainingattechnologistandmanagementlevels,retentionoftrainedpersonnel,enhancementofquality-assur-ance systems, enhancement of results-reporting mechanisms,andreliablemechanismsforinstrumentmaintenanceandsup-plyprocurement.StrengtheningofHIV(andinsomeinstancesTB) laboratory capacity under the US President’s EmergencyPlan for AIDS Relief and related programs serves as a usefulmodel,asdoesthesuccessfulcollaborativeprogramundertakenby the Foundation for Innovative New Diagnostics, Partnersin Health, the World Health Organization, and the LesothoMinistry of Health and Social Welfare to strengthen the Na-tionalReferenceLaboratoryofLesotho[12].

Anumberofimportantbarriersexistwithrespecttothefullengagementofindustryandinvestorsindiagnosticsdevelop-ment. Barriers include uncertainty about the size and/or ac-cessibilityoftheTBdiagnosticsmarket,especiallyindevelopingcountries; complex regulatoryprocesses;unfavorableintellec-tualpropertyrightsprotections;andinsomeinstances,lackofknowledge about the types of tests thataremostneededandlikelytoberelevant.Arecentanalysisindicatesthat,worldwide,1US $1 billion is spent annually on TB diagnostics [15]. Ofinterest,approximatelythree-quartersofalldiagnostictestsare

performedoutsideestablishedmarketeconomies;however,thistestingburdenaccountsforonlyone-third(∼US$326million)of the current market, because the most common tests per-formed are sputum smear microscopy and chest radiograph,whichhaverelativelylowper-testcosts[15].Incountrieswithestablished market economies, tuberculin skin testing is themostfrequentlyusedTBtest,inaccordancewiththerelativelylowratesofTBandrelativelyhighprioritizationofdetectionandtreatmentoflatentTBinthosesettings.Theblood-basedinterferon-g release assays, including the QuantiFERON-TBGoldtests(Cellestis)andT-SPOT.TB(OxfordImmunotec),candetectTBwithahigherdegreeofspecificitythancanthetu-berculinskintestandarenowapprovedforuseintheUnitedStatesandanumberofothercountries.Todate,intheUnitedStates,useofthesetestsasreplacementsfororadjunctstothetuberculinskintesthasnotbeenwidespread,butmomentumappearstobegrowing.

Funding for TB diagnostics research by the top 40 TB re-

searchfundinginstitutionsin2007wasestimatedatUS$41.9million,lessthantheamountforTBdrugs(US$170million)andTBvaccines(US$71.2million)andmorethantheamountforTBoperationalresearch(US$36.8million)[16].ThisleveloffundingfallswoefullyshortoftheGlobalPlantoStopTB’srecommendationsofatleastUS$900millionperyearforre-searchanddevelopmentofnewtoolsforTBdiagnosis,treat-ment,andprevention[17].Supportfortechnicalassistancetonational TB programs as they implement and monitor newtoolscannotbeunderestimated.Fundingestimatesaside,itisclear that important advances in TB diagnosis have recentlybeenmade,andpotentiallyusefulnewtoolsareemerging;con-tinuedandaugmentedinvestmentwillberequiredtosuccess-S176 • CID 2010:50 (Suppl3) • Dorman

projectoftheInfectiousDiseasesSocietyofAmericaandtheHIVMedicineAssociation,throughagrantfromtheBill&MelindaGatesFoundation.

fullyimplementthemostpromisingofthesetoolsintheset-tings where they are most needed and to maintain a robustpipeline that will ultimately yield the tools that revolutionizeTBdiagnosis.

References

1. PaiM,ZwerlingA,MenziesD.Systematicreview:Tcell-basedassaysforthediagnosisoflatenttuberculosisinfection—anupdate.AnnIn-ternMed2008;149:177–184.

2. SteingartKR,HenryM,NgV,etal.Fluorescenceversusconventionalsmearmicroscopyfortuberculosis:asystematicreview.LancetInfectDis2006;6:570–581.

TOOLSINTHEPIPELINE:TRANSFORMATIVEORINCREMENTALGAINS?

Mostofthetoolsindemonstrationorlate-stageevaluationaresputum based and, thus, are likely to result—at best—in in-cremental gains in TB case detection. Their yield is expectedto be suboptimal for patients with TB who have only extra-pulmonaryTB,whohaverespiratorydiseaseinwhicharela-tivelylargeburdenoforganismsisnotincommunicationwiththe airways, and who cannot provide a respiratory specimenfortesting.Nevertheless,effectiveimplementationmight,overtime,haveasubstantialimpactonTBcontrolthroughdetectionofaveryhighproportionofindividualswithcapacitytotrans-mit infection to others (provided diagnosis is sufficientlypromptandtreatmentisavailable).Highlyaccurate,simpletoperform,point-of-caretestsamenabletotestingofreadilyavail-ableclinicalspecimens,suchasurineorblood,andwithabilitytodetectandpredictactiveTBanywhereinthebody,incom-binationwitheffectivepreventiveandtreatmentstrategies(asdescribed in other articles in this Supplement), are needed.Truly transformative change will require more than a perfectsputumtest,butareallygoodsputumtestwouldbeastepintherightdirection.

3. CorbettEL,WattCJ, WalkerN,et al.Thegrowingburdenoftuber-culosis: global trends and interactions with the HIV epidemic. ArchInternMed2003;163:1009–1021.

4. PageKR,Godfrey-FaussettP,ChaissonRE.Tuberculosis-HIVcoinfec-tion: epidemiology, clinical aspects,andinterventions.In:RaviglioneM, ed. Reichman and Hershfeld’s tuberculosis: a comprehensive in-ternationalapproach.NewYork:InformaHealthcare;2006:371–416.

5. BurgessAL,FitzgeraldDW,SevereP,etal.IntegrationoftuberculosisscreeningatanHIVvoluntarycounsellingandtestingcentreinHaiti.AIDS2001;15:1875–1879.

6. HargreavesNJ,KadzakumanjaO,WhittyCJ,SalaniponiFM,HarriesAD, Squire S. “Smear-negative” pulmonary tuberculosis in a DOTSprogramme: poor outcomes in an area of high HIV seroprevalence.IntJTubercLungDis2001;5:847–854.

7. BehrMA,WarrenSA,SalamonH,etal.TransmissionofMycobacteriumtuberculosis from patients smear-negative for acid-fast bacilli. Lancet1999;353:4444–4449;erratum:Lancet1999;353:1714.

8. WorldHealthOrganization.Globaltuberculosiscontrol:surveillance,planning, financing: WHO report 2008. WHO/HTM/TB/2008.393.http://www.who.int/tb/en.Accessed30June2009.

9. StopTBNewDiagnosticsWorkingGroup.StrategicPlan2006–2015.http://www.stoptb.org.Accessed30June2009.

10. Cole ST, Brosch R, Parkhill J, et al. Deciphering the biology ofMy-cobacteriumtuberculosisfromthecompletegenomesequence.Nature1998;393:537–544.

11. PerkinsMD,CunninghamJ.Facingthecrisis:improvingthediagnosisoftuberculosisintheHIVera.JInfectDis2007;196(Suppl1);S15–S27.

12. WorldHealthOrganizationandStopTBPartnership.Newlaboratorydiagnostictoolsfortuberculosiscontrol.http://www.apps.who.int/tdr.Accessed30June2009.

13. Bossuyt PM, Reitsma JB, Bruns DE. Towards completeandaccurate

reportingofstudiesofdiagnosticaccuracy:theSTARDinitiative.AnnInternMed2003;138:40–44.

14. Lawn SD, Edwards DJ, Kranzer K, et al. Urine lipoarabinomannanassayfortuberculosisscreeningbeforeantiretroviraltherapydiagnosticyield and association with immune reconstitution disease. AIDS2009 ;23:1875–1880.

CONCLUSIONS

What’s new in TB diagnostics? A lot, but not enough. Thefutureisbrighterasseveralpromisingnewtoolsenterthedem-onstrationandlateevaluationstages.Buttheneedisgreat,andimportantbarriersremainintranslatingtechnicaladvancesintomeaningful and sustainable improvements in individual andpublichealthinsettingshardesthitbyTB.

15. WorldHealthOrganizationSpecialProgrammeforResearchandTrain-

ing in Tropical Diseases, and Foundation for Innovative New Diag-nostics.Diagnosticsfortuberclosis:globaldemandandmarketpoten-tial.2006.http://apps.who.int/tdr.Accessed30June2009.

16. Agarwal N, Syed J, Harrington H. Tuberculosis research and devel-

opment: a critical analysis of funding trends, 2005–2007 an update.http://www.treatmentactiongroup.org.Accessed30June2009.

17. StopTBPartnershipandWorldHealthOrganization.GlobalPlantoStopTB2006–2015.Geneva,WorldHealthOrganization,2006.

Acknowledgments

Potentialconflictsofinterest. S.E.D.:noconflicts.Financial support. National Institutes of Health (HHSN2722009

00050C).Supplementsponsorship. Thisarticleispartofasupplemententitled

“SynergisticPandemics:ConfrontingtheGlobalHIVandTuberculosisEp-idemics,”whichwassponsoredbytheCenterforGlobalHealthPolicy,a

NewDiagnosticTestsforTB • CID 2010:50 (Suppl3) • S177