Biomarkers and diagnostics for tuberculosis

8/7/2019 Biomarkers and diagnostics for tuberculosis

1/18

Series

www.thelancet.com 57

Published Online

May 19, 2010

DOI:10.1016/ S0140-

6736(10)60359-5

This is the ourth in a Series o

eight papers about tuberculosis

*These authors contributed

equally and are joint frst authors

Joint senior authors

Pfzer, New London, CT, USA

(R S Wallis MD);Department o

Epidemiology, Biostatistics

and Occupational Health,

McGill University, Montreal,

QC, Canada (M Pai MD,

D Menzies MD); Montreal Chest

Institute, Montreal, QC, Canada

(M Pai, D Menzies); Statens

Serum Institute, Copenhagen,

Denmark

(Pro T M Doherty PhD);

Stellenbosch University,

Stellenbosch, South Arica

(Pro G Walzl PhD); Foundationor Innovative New

Diagnostics, Geneva,

Switzerland (M D Perkins MD);

and Department o Inection,

University College London

Medical School, London, UK

(Pro A Zumla FRCP)

Correspondence to:

Pro Alimuddin Zumla, Centre

or Inectious Diseases and

International Health,

Department o Inection,

University College London

Medical School, Windeyer

Institute o Medical Sciences,

46 Cleveland Street, London

W1T 4JF, UK

[email protected]

or Dr Mark D Perkins, Foundation

or Innovative New Diagnostics,

Avenue de Bud 16,

1202 Geneva, Switzerland

mark.perkins@fnddiagnostics.

org

For more on Evidence-based

TB Diagnosis see http://www.

tbevidence.org

Tuberculosis 4

Biomarkers and diagnostics or tuberculosis: progress,

needs, and translation into practice

Robert S Wallis*, Madhukar Pai*, Dick Menzies, T Mark Doherty, Gerhard Walzl, Mark D Perkins, Alimuddin Zumla

Human inection with Mycobacterium tuberculosiscan progress to active disease, be contained as latent inection, or beeradicated by the host response. Tuberculosis diagnostics classiy a patient into one o these categories. These are notxed distinct states, but rather are continua along which patients can move, and are aected by HIV inection,immunosuppressive therapies, antituberculosis treatments, and other poorly understood actors. Tuberculosisbiomarkershost or pathogen-specicprovide prognostic inormation, either or individual patients or study cohorts,about these outcomes. Tuberculosis case detection remains difcult, partly because o inaccurate diagnostic methods.

Investments have yielded some progress in development o new diagnostics, although the existing pipeline is limitedor tests or sputum-smear-negative cases, childhood tuberculosis, and accurate prediction o reactivation o latenttuberculosis. Despite new, sensitive, automated molecular platorms or detection o tuberculosis and drug resistance,a simple, inexpensive point-o-care test is still not available. The eect o any new tests will depend on the method andextent o their introduction, the strength o the laboratories, and the degree to which access to appropriate therapyollows access to diagnosis. Translation o scientic progress in biomarkers and diagnostics into clinical and publichealth programmes is possiblewith political commitment, increased unding, and engagement o all stakeholders.

IntroductionTuberculosis, although a curable disease, continues to beone o the most important inectious causes o deathworldwide. Despite substantial investments and progressmade in expansion o the directly observed therapy, short

course (DOTS) strategy and improved treatment com pletion rates, inadequate case detection remains a majorobstacle to global control o tuberculosis. Eorts duringthe past decade to consistently diagnose and treat themost inectious cases have slowed the rate o diseaseincidence, but have not yielded substantial progress

towards elimination. This experience has reocusedattention on research and development or improveddiagnostics, therapeutics, and vaccinesareas inwhich progress has historically been slow. HumanMycobacterium tuberculosisinection is almost alwaysacquired by inhalation o inected aerosol droplets, whichare generated by people with active pulmonary disease

Key messages

Diagnosticsclassifypatientsatonepointintime,whereas

biomarkers can provide prognostic inormation about

uture health status and can advance knowledge o

disease pathogenesis.

Qualiedtuberculosisbiomarkersaremosturgently

needed as predictors o reactivation and cure, and

indicators o vaccine-induced protection. The biomarker

most closely approaching qualication is 2-month culture

conversion as a predictor o relapse risk.

Thetuberculosisdiagnosticspipelinehasrapidlygrown,

with development o several promising technologies.

Theexistingtuberculosisdiagnosticspipelinestilldoes

nothaveasimple,rapid,inexpensivepoint-of-caretest.

Accurate, rapid tests are also needed or smear-negative

and childhood tuberculosis, as are tests or latent

tuberculosis with increased predictive value or

reactivation.

Severaldiagnosticsanddiagnosticstrategieshavebeen

endorsed by WHO and are being introduced into clinical

use and national tuberculosis control programmes.

Governmentsinallcountries,especiallyindustrialised

countries, have to increase unding or tuberculosis

research and control.

Search strategy and selection criteria

For the tuberculosis biomarker section, we searched

publications in PubMed and Google Scholar (19952009),

the Cochrane library (200109), and Embase (200106) with

the terms tuberculosis, Mycobacterium tuberculosis,

biomarkers, diagnostics, and clinical trials. For the

section on tuberculosis diagnostics, the search strategy was a

10-year review o diagnostic studies in PubMed and Embase.

This search was supplemented by searching the website

Evidence-based TB Diagnosis by the Stop TB Partnerships

New Diagnostics Working Group. We searched with the terms

tuberculosis, Mycobacterium tuberculosis, diagnosis,

diagnostics tests, and accuracy. For both sections, we

mainly selected publications in the past 10 years, but did not

excludecommonlyreferencedandhighlyregardedolder

publications. We also reviewed studies cited by articles

identied by this search strategy, and selected those that we

regarded as relevant. Review articles are preerentially cited to

provide readers with more details and reerences than this

overview can accommodate.


2/18


3/18

Series


Small studies have examined levels oM tuberculosis

antigen 85 and 85B RNA in sputum during treatment. Inone study,8 the magnitude and duration o increases inthis protein during the rst week o treatment predictedrelapse or ailure in our o 42 patients. A second study12noted that 85B RNA was cleared more rapidly romsputum during therapy than viable colony counts, butdid not predict subsequent relapse in one patient. Otheractors associated with mycobacterial dormancy(including proteins, RNA species, or lipids) could havegreater predictive value.13,14

An important shortcoming o all sputum biomarkers istheir limited role in paucibacillary and paediatrictuberculosis, and their lack o useulness in latentM tuberculosis inection. One study has reported the

presence o small ragments oM tuberculosis IS6110DNA in urine o 34 o 43 patients with tuberculosis butnot in healthy controls.15 None o the patients had overtrenal tuberculosis. The DNA ragments, termedtransrenal DNA (tr DNA), are thought to arise because oapoptosis o host cells. The investigators have reportedthat none o the 20 patients who were positive at diagnosisremained positive ater 2 months o therapy.16 Otherstudies o urinary mycobacterial DNA using dierentmethods have generally shown lower sensitivity.106109None has examined paediatric samples. A urinary testthat could serve as both diagnostic and prognostic wouldbe an important advance in paediatric tuberculosis. Otherstudies have examined lipoarabinomannan and othermycobacterial markers in urine, also with varying degreeso sensitivity.1726 Detection o volatile organic compoundsin the breath o patients with pulmonary tuberculosishas been reported.28 No studies have reported the changesin these markers during treatment or established arelation to clinical outcome or to another surrogateendpoint. Further study o non sputum microbialmarkers is an area o priority or tuberculosis research.

Measurement o bactericidal activity in whole bloodculture ater oral dosing o new tuberculosis drugs canhelp with selection o dose and dosing interval, and canidentiy compounds which, owing to their mechanism oaction and pharmacokinetic proles, can show additive

or synergistic eects when combined. Such eects mightnot be predicted well rom animal models because odierences in absorption and metabolism.110,111 Wholeblood bactericidal activity during tuberculosis treatmentcorrelates with the rate o decline in sputum colonycounts, is superior in patients whose sputum culturesconvert by the second month o treatment, and is superiorduring the intensive (our drug) phase o treatment. 31Two studies29,30 reported that regimens or drug sensitivetuberculosis were better than were those or MDRtuberculosis. These ndings suggest that the whole bloodmodel could also help in the identication o efcaciousmultidrug regimens.

Macrophages are activated by M tuberculosis via inter actions with toll like receptors. Several blood markers o

this activation might have roles as tuberculosis biomarkers.Neopterin, or example, is increased at diagnosis o diseasein proportion to extent o disease; it decreases during andater treatment.4145 In a small sample o HIV uninectedpatients matched or extent o disease at baseline,increased neopterin concentrations ater completion otreatment were associated with relapse.44Several othermarkers are also increased at baseline in proportion todisease extent and to decline with treatment, includingsoluble intercellular adhesion molecule (sICAM) 1,6063C reactive protein, 55,5759 soluble urokinase plasminogenactivator receptor,65 and procalcitonin.4653 In one study, a

Panel 1: Candidate Mycobacterium tuberculosis and host

tuberculosis biomarkers

Predication o durable (non-relapsing) tuberculosis cure

Microbial markers in sputum

2-monthcultureconversion24

Serialcolonycountsortimetoculturepositivity59

Othermicrobialmarkers8,1014

Other microbial markers

UrineM tuberculosis DNA,15,16 lipoarabinomannan1726

Volatileorganiccompounds27,28

Mycobactericidal activity

Wholebloodculture2931

Tuberculosis-specifc T-cell unction

Interferon,3237 interleukin 42 splice variant3840

Macrophage activation markers

Neopterin,4145 procalcitonin,4653 C-reactive protein,5459

soluble intercellular adhesion molecule 1,6064 soluble

urokinase plasminogen activator receptor,64,65 monocyte

CDllc66

Multiple host markers

Proteomics67,68

Transcriptomics69,70

Indication o reactivation risk and prediction o

eradication o latent inection


Interferon7177

Interferon-inducedprotein107880

Interleukin42splicevariant3840,81,82

Skintest83,84

Macrophage activation

Neopterin85

Procalcitonin51

Prediction o vaccine ecacy


Interferon86

PolyfunctionalTcells8790

Mycobactericidal activity

Wholebloodculture

9199

Mononuclearcells86,98,100,101


4/18

Series

60 www.thelancet.com

mathematical model including change in sICAM 1 duringthe rst week o therapy predicted 2 month sputumculture conversion.64As a group, these assays are simple,

inexpensive, widely available, and can be done on rozenplasma samples; as a result, they can be readilyincorporated into clinical trials or treatment protocols.They seem to have greatest prognostic value whenmeasured at or near the completion o therapy. Furtherresearch is needed to establish the sensitivity o these teststo predict tuberculosis reactivation or relapse, and theextent to which their lack o specicity or M tuberculosisinection conounds their interpretation.

Multiple biomarkers, when combined, can dosubstantially better than can any one marker. Forexample, a panel consisting o leptin, prolactin,osteopontin, insulin like growth actor II, macrophageinhibitory actor, and CA 125 yielded 953% sensitivityand 994% specicity or diagnosis o ovarian cancer, or

which measurement o CA 125 alone detects only

3040% o early cases.

112

Increased specicity and highpredictive value may similarly be achieved or otherwisenon specic tuberculosis biomarkers by measuringmultiple parameters by proteomics, transcriptomics,and metabolomics.10,68 Tuberculosis can be dierentiatedrom other inectious and inammatory diseases on thebasis o proteomic ngerprinting study o serum bysurace enhanced laser desorption/ionisation time o ight (SELDI ToF) mass spectrometry. 67 Serum amyloid Aand transthyretin were among the candidate biomarkersidentied. This analytical method can detect a very largenumber o peptides, although it is airly insensitive.A related approach reduces the potential number ocandidate molecules to a small set o small molecules

termed the metabolome, representing metabolicintermediates, hormones, other signalling molecules,and secondary metabolites. Its main disadvantage is thatseveral analytical methods are necessary to completetheir characterisation.

Two reports suggest the easibility o distinguishingvarious stages oM tuberculosis inection by gene expressionmicroarray. One study70 recorded many candidate genesthat were dierentially expressed by mononuclear cells opatients with tuberculosis, people with latent inection,and uninected people. However, a small subset o thesegeneslactoerrin, CD64, and the Ras associated GTPase33Awas sufcient or classication o the three groups.A second report identied signature proles o nine genesin blood that could distinguish our groups: patients withactive tuberculosis, those with latent inection, curedpatients, and cured patients with several previous episodeso tuberculosis.69 Similar studies undertaken during or atcompletion o therapy could identiy proles associatedwith durable cure or relapse. However, the genomics andproteomics platorms might be susceptible to biasesindicating regional dierences in host and microbialgenetics. These ndings have yet to be veried acrossseveral clinical populations.

T cell based assays o intereron release or diagnosiso latent tuberculosis inection tend to show high levels inpeople with active disease at diagnosis that decrease ater

completion o treatment, but this pattern does not occurconsistently.32 A small study using a non commercial assayto monitor responses at earlier timepoints noted that o18 active cases o tuberculosis with positive T cell responsesat baseline, only ve who did not show a microbiologicalor clinical response remained positive ater 3 months otreatment.33 Subsequent studies using commercial assayshave not yielded such denitive results. Although moststudies have concurred in nding sustained positive T cellresponses in tuberculosis treatment ailures, most haveound reversion o T cell responses in responders to be tooincomplete and delayed to be useul as a biomarker.3437,113The diagnosis o active tuberculosis can also be establishedon the basis o T cell requencies at the site o diseaserather than in blood.114116 However, the requirement or an

Figure 2: Relation between eects o new tuberculosis treatments on rates o relapse and 2-month sputum

culture conversion in randomised controlled trials

Axesindicatenaturallograteratios(experimental/control),withdottedlinesindicatingequality(noeect).

Symbols indicate pairs o study groups, diering only in the intensive phase (n=16) or throughout treatment

(n=14). Symbol sizes vary according to precision. The solid line indicates the results o meta-regression analysis

(p


5/18

Series


invasive procedure restricts the easibility o this approach

or diagnosis and treatment monitoring. Lastly, antibodyconcentrations to some mycobacterial antigens are raisedat diagnosis and might be modulated by treatment;however, perormance characteristics seem inadequate ora prognostic role.117,118 Immunological memory seems tohamper the ability to quickly detect treatment eects, as isthe case with T cell assays.

Biomarkers indicating reactivation riskSeveral natural history studies o household contacts oactive cases o tuberculosis suggest that in HIV uninectedpeople, particularly high or increasing concentrationso tuberculosis specic intereron production mightpredict overt tuberculosis, although the numbers o

tuberculosis cases in these studies are small.7577 Positiveresponses to intereron release assays (IGRAs) otherwiseseem to coner only a small risk o reactivation (1020 per1000 person years), which is similar to that o a positivetuberculin skin test. Some studies have suggested that re lative mRNA concentrations o intereron , interleukin 4,and interleukin 42 (a splice variant o interleukin 4)might be better predictors than intereron alone, sinceratios o intereron or interleukins 4 and 42 all ashealthy contacts develop tuberculosis, and increase aspatients with tuberculosis are cured.3840,81 The ratio ointerleukin 4 to 42 is also increased in longstanding latenttuberculosis inection, presumably suggesting low risk oreactivation.82 One study85 reported nding intermediateconcentrations o neopterin in health care workers whowere heavily exposed to tuberculosis, potentially indicatingrisk o reactivation o latent M tuberculosis inection. Nostudies have yet examined macrophage and T cell markerstogether in this context.

Findings rom studies in experimentally inectedguineapigs suggest that prognostication o tuberculosiswith the tuberculosis specic antigens ESAT 6 andCFP 10 as skin tests might also be possible.83 A similarthough less pronounced occurrence had been describedin patients with responses to a tuberculin skin test.84Sucha skin test might show better specicity or latentM tuberculosis inection and increased positive predictive

value or tuberculosis than might the tuberculin skintest. Conrmation o these ndings in human beingsand validating their prognostic signicance are priorityareas o research.

How the loss o CD4 T cells due to HIV inection willaect prognostication o tuberculosis with use o T cell based assays is unclear. One study o acute HIV inectionnoted that tuberculosis specic T cell intereron responses were lost rapidly in our o ve patients, all owhom remained well.119 In the th, tuberculosis specicresponses increased progressively ater HIV inectionwas acquired, culminating in the diagnosis o activetuberculosis. Increasing counts might correlate withantigen burden and presage tuberculosis reactivation inHIV positive and HIV negative people, but studies in

HIV inected patients with specic ranges o CD4 T cells

will be needed to conrm this observation.Several studies o people recently exposed totuberculosis showed that T cell requencies decrease atercompletion o isoniazid preventive therapy, although theyinrequently reverted to negative.120123 However, in otherstudies,124126 responses were unaected. Factors aectingthe likelihood o IGRA reversion due to isoniazidpreventive therapy could include the duration o inection,the type o assay, the magnitude o the response, or therisk o reinection.127 No studies have specically examinedthe prognostic signicance o reversion. IGRAs areunlikely to be adequate indicators o successul isoniazidpreventive therapy. Multiplex assays assessing both T celland macrophage actors could prove useul.

Biomarkers predicting vaccine ecacyThere are no qualied biomarkers to indicate protectionby new vaccines against tuberculosis. Although bothnatural inection and vaccination with Mycobacteriumbovis BCG result in the acquisition o delayed typehypersensitivity and expansion o antigen specicintereron producing T cell populations, the linkbetween these responses and protection rom disease isweak. In the case o BCG, or example, intereron producing T cell requencies poorly predict the protectiveefcacy o various BCG strains in animals.128 Protectionmight instead correlate better with the presence opolyunctional antigen specic T cells that secrete severalcytokines, as has been described in leishmaniasis.129However, data or the use o this biomarker or vaccine induced protection in tuberculosis are scarce. Thepotential eect o this insufcient knowledge is shown intwo studies o the eect o route o BCG administrationon its efcacy. The rst study89 noted superiorimmunogenicity in inants (intereron , tumournecrosis actor, and interleukin 2 responses in bothCD4 and CD8 T cells) when the vaccine was administeredpercutaneously rather than intradermally. However, asubsequent study130 in this population showed the twomethods o administration did not dier in protectionagainst tuberculosis. These ndings suggest that

assessment o T cell responses alone might be in sufcient to predict protection rom tuberculosis byvaccination.

For all other licensed vaccines, bactericidal or viralneutralisation assays have supplemented standardmeasurements o immunogenicity during development.Bactericidal assays have been described or M tuberculosiswith mononuclear cell or whole blood culture.86,91,92,101Immune control o growth in these assays is inerior inpeople with negative tuberculin skin test and in youngchildren; improved by BCG vaccination or vitamin D;impaired by chemokine receptor blockade, T cell depletion,or HIV inection; restored by antiretroviral treatment; andmight be strain specic.86,9199,131 Their predictive role ornew tuberculosis vaccines has yet to be studied.


6/18

Series


Eect o biomarkers on development timelines

The potential eect o biomarkers on the time and costso development o new tuberculosis drugs and vaccinescan be substantial. In the USA, Federal regulations(subpart H o 21CFR314) allow accelerated approval onew drugs or serious or lie threatening illnesses onthe basis o a surrogate endpoint that is reasonablylikely, based on epidemiologic, therapeutic, patho physiologic, or other evidence, to predict clinicalbenet.132 In 2009, an Advisory Committee convenedby the US Food and Drug Administration (FDA)recommended nearly unanimously in avour oaccelerated approval o new drugs or MDR tuberculosison the basis o sputum culture conversion. Suchapproval will shorten the time needed or licensing o

new, more eective treatments to patients with MDRtuberculosis by as much as 3 years. This strategy mightalso be used in development o new regimens, ratherthan single compounds. Here, measurement o serialsputum colony orming unit counts and whole bloodbactericidal activity in trials o 14 weeks duration canprovide a seamless progression rom preclinical studiesthrough trials resulting in licensing with cultureconversion. Such a development plan might reduce byas much as a decade the time required to have newregimens or MDR tuberculosis without cross resistanceto any existing tuberculosis drug.

Strategies or biomarker qualicationAlthough biomarkers have historically been widely usedin drug development and medical practice, only recentlyhave pathways been created to include them in theregulatory review process. In the USA, the impetus orthis change came rom the National Institutes o HealthRoad Map and the FDA Critical Path Initiative, both owhich sought to introduce greater efciency in drugresearch and development.133,134In this context, the termvalidation reers to assay perormance characteristics(eg, how accurately urinary albumin is measured),whereas qualication reers to linkage to biologicalprocesses (eg, to what extent do increases in urinealbumin predict aminoglycoside nephrotoxicity).135

Biomarker review at the FDA includes a voluntary datasubmission that is examined by a biomarker qualicationreview team. Three categories o certainty are described:biological plausibility; prognosis o clinical outcomes indisease; and capturing dierences in efcacy in clinicaltrials.136 The rst category could be described asappropriate only or exploratory purposes, whereas thethird might be needed or registration o a new therapyor vaccine. O all the markers described in this review,only 2 month sputum culture conversion alls into thethird category. Reaching this level o certainty isparticularly challenging in the case o new vaccines ortuberculosis, since there is no eective modern vaccineto which BCG might be compared or the purposes obiomarker qualication, and since innate genetic actors

not amenable to modulation by vaccination might

account or and be detected by biomarkers that predictrisk o disease in natural history studies.

Tuberculosis diagnosticsProgress towards elimination o tuberculosis hasremained elusive despite intensied standard measureso control. Ater a period o global acceleration in 200105,the case detection rate worldwide decelerated in 2006 and2007, reaching 63% in 2007.137 Thus, the target o a casedetection rate o at least 70% by 2005 has not yet beenachieved, and is unlikely to be met until 2014.137

Insufcient access to advanced diagnostic tests hascontributed to this suboptimum perormance. Even in2010, national tuberculosis programmes in disease

endemic countries continue to rely largely on antiquatedand inaccurate methods such as direct smear microscopy,solid culture, chest radiography, and tuberculin skintesting. There is no rapid, point o care test that allowsearly detection o active tuberculosis at health clinics.Diagnosis o smear negative tuberculosis in adults inectedwith HIV and in children continues to pose substantialclinical challenges. Even existing diagnostics are not usedto their ull potential because o poor access to health careand ailures in health care delivery systems, includingpoor synergy between national HIV/AIDS and tuber culosis programmes. Diagnostic delays, misdiagnosis,and inadequate implementation o existing tests result inincreased morbidity and mortality in patients, andallow continued transmission o tuberculosis.138 Theserestrictions o present case detection approaches arestarkly visible in countries with a high prevalence o HIVinection or MDR tuberculosis, or both.139141

Barriers to development o new tuberculosis diagnosticsMarket ailure has been an important actor hindering thedevelopment o new diagnostics or tuberculosis. Industrytends to avoid developing and marketing products thatwill be mainly used or poor patients in resource limitedcountries because such products will not generateprots.142,143 When products are available, neither pricingnor perormance is adapted or developing countries, and

their potential benets are eectively unavailable orpatients and health care providers who need them most.

Furthermore, health systems in developing countriesare generally weak, making them unable to take advantageo tuberculosis diagnostics to achieve best possibleperormance, and to introduce new advances in diagnostictechnologies. This situation is the result o poormanagement, insufcient nancial resources, inadequatehuman resources, and poor laboratory capacity.144Forexample, rapid tests or malaria are a model o the type oassay widely needed or tuberculosis, but only a smallproportion o patients receiving malaria treatment aretested.145 Rapid tests or HIV inection are highly accurate,but undiagnosed HIV inection is very common, and alarge proportion o HIV inected individuals do not


7/18

Series


present or HIV testing until late in inection.146 Only

about 1020% o people inected with HIV in Arica areaware o their status.144Furthermore, less than 3% opeople with HIV inection are screened or tuberculosis,and globally, only about 20% o notied tuberculosispatients are aware o their HIV status.147 These estimatessuggest that even existing diagnostic strategies are poorlyimplemented in many settings.

Knowledge gaps and scientic obstacles impedingprogressOur understanding o the biology oM tuberculosis andinteractions with the human host is incomplete, andthese knowledge gaps impede the development obiomarkers that can distinguish between latent and active

tuberculosis, and distinguish active tuberculosis romother diseases, especially in HIV inected adults andchildren.148150 Present tests or latent M tuberculosisinection do not adequately distinguish resolved rompersistent inection, and are unable to efciently identiyindividuals who are at highest risk o reactivation. 151155Studies into predictive value o IGRAs show only modestpredictive ability, and several studies show similar (andrather low) rates o progression in people with positivetuberculin skin test and IGRA results.156160

Other important knowledge gaps pertain to diagnosiso smear negative tuberculosis in children and HIV inected individuals, and rapid and accurate identica tion o resistance to second line antituberculosis drugs.Although molecular markers have been identied andsuccessully used as rapid and accurate tests orisoniazid and riampicin resistance, testing or theresistance that characterises extensively drug resistanttuberculosis is on a less robust scientic ooting than istesting or MDR tuberculosis.161

The diagnostics pipeline and new WHO policiesOver the past decade, tangible progress has been madein the development o new tuberculosis diagnostics.The increase in investments has resulted in an expandedpipeline o new diagnostic tests.162,163 The private sector,led by unding rom the Bill & Melinda Gates

Foundation, is increasingly engaged in public privatepartnerships such as the Foundation or InnovativeNew Diagnostics (FIND) to develop and delivera pipeline o tests that are appropriate or disease endemic countries. Furthermore, under the umbrellao the Global Laboratory Initiativeone o the WorkingGroups o the Stop TB Partnershipplans are underwayor a large scale up o laboratory services ortuberculosis. For example, UNITAID is providingUS$81 million unding or a programme calledEXPAND TB that will supply rapid diagnostics or MDRtuberculosis to 27 high burden countries. 164Anotherexample is the allocation o substantial resources tolaboratory strengthening by the US PresidentsEmergency Plan or AIDS Relie (PEPFAR).165

For the rst time in many years, progress is being made

in developing a range o diagnostic options orlaboratories in disease endemic countries. The Stop TBPartnerships Retooling Task Force and New DiagnosticsWorking Group produced a summary on the diagnosticspipeline.166 Figure 3 shows an updated version o thepipeline,167 which displays the tests that have beenendorsed by WHO between 2007 and 2009. A completedescription o existing and novel tuberculosis diagnosticsis available elsewhere.139,168

Since 2007, several tuberculosis diagnostics have beenendorsed by WHO or use in disease endemic countries(panel 2). In 2007, WHO endorsed the use o liquidculture systems and rapid tests or species conrmationthrough antigen detection.169 This WHO policy, along

with FINDs negotiations with industry, madeimplementation o liquid culture systems aordable andeasible or the rst time, especially in countries withhigh HIV prevalence.

Line probe assays, which are based on reversehybridisation technology, have consistently shown excellentaccuracy or rapid detection o MDR tuberculosis.175 Asa result, in 2008, WHO endorsed the use o these assaysor rapid detection o MDR tuberculosis in smear positivepatients.170 Several non commercial and less expensiveoptions have been explored or MDR screening o clinicalspecimens with a variety o culture methods withincentralised reerence laboratories, including microscopicobservation drug susceptibility, thin layer agar, directnitrate reductase, and colorimetric redox indicator assays.WHO considered evidence or their accuracy and role, andrecommended that selectednon commercial culture anddrug susceptibility testing methods be used as an interimsolution in resource constrained settings, in reerencelaboratories, or in other laboratories with sufcient culture

Figure 3: The tuberculosis diagnostics pipeline

TechnologiesinboxeshavebeenendorsedbyWHO.DST=drug-susceptibilitytest.MODS=microscopicobservation

drugsusceptibility.NRA=nitratereductaseassay.CRI=colorimetricredoxindicatorassay.LPA=line-probeassay.

NAAT=nucleic acid amplication test. LED=light-emitting diode. POC=point o care. LTBI=latent tuberculosis

inection. *Manual NAAT: technology or Mycobacterium tuberculosis drug-susceptibility testing. Manual NAAT:

technology or M tuberculosis detection at the peripheral laboratory. Manual NAAT: technology or M tuberculosis

detection at the community health-care level. Source: adapted rom Stop TB Partnership. Global Plan to Stop TB,20062015,167 and reproduced with permission rom author and publisher.

Autom

atedNAA

T

(detectionan

dDST)

Rapidcolorim

etricDST

ManualNAAT

(detectio

n)

POCtest

(detectio

n)

Predictiv

eLTBI

ManualN

AAT

(detectio

n)

Liquidcu

ltureand

DST

Rapidsp

eciation

LEDmicr

oscopya

ndsame

-daydiag

nosis

LPAsma

nualNAA

T(DST)*

Non-com

mercialc

ultures

(MODS,N

RA,CRI)

First referral level 1040%

70%

Accessafter5years(%)

95%

Peripheral laboratory

Community health careDistancefrompatients

2007 2008 2009 2010 2011Year

2012 2013 2014 2015 2016


8/18

Series


capacity, while capacity or genotypic or automated liquid

culture and drug susceptibility testing is being developed.

171

Although non commercial assays have similar accuracy asdo commercial liquid culture systems and cost less, thesetests are not standardised and need extensive training,optimisation, and quality assurance beore clinical use.

Panel 2: Summary o WHO policies and statements on

tuberculosis diagnostics

Liquid media or culture and DST (introduced in 2007)

WHO recommends, as a step-wise approach:

TheuseofliquidmediumforcultureandDSTin

middle-income and low-income countries.

Rapidspeciesidenticationtoaddresstheneedsfor

culture and DST, taking into consideration that

implementation o liquid systems will be phased, will be

integrated into a country-specic comprehensive plan or

laboratory-capacity strengthening, and will address

several issues including biosaety and training.

Denition o a new sputum-smear-positive tuberculosis

case (introduced in 2007)

The revised denition o a new sputum-smear-positive case

o pulmonary tuberculosis is based on the presence o at least

one acid ast bacilli in at least one sputum sample in countries

withawellfunctioningexternalquality-assurancesystem.

Reduction o number o smears or diagnosis o

pulmonary tuberculosis (introduced in 2007)

WHOrecommendsthenumberofspecimenstobeexaminedor screening o tuberculosis cases can be reduced rom three

totwo,inplaceswhereawellfunctioningexternalquality-

assurancesystemexists,wheretheworkloadisveryhigh,and

human resources are scarce.

Molecular line-probe assays or rapid screening o patients

at risk o MDR tuberculosis (introduced in 2008)

The use o line-probe assays is recommended by WHO, with

the ollowing guiding principles:

Adoptionofline-probeassaysforrapiddetectionofMDR

tuberculosis should be decided by ministries o health

withinthecontextofcountryplansforappropriate

management o patients with MDR tuberculosis,

including the development o country-specic screeningalgorithms and timely access to quality-assured second-

line antituberculosis drugs.

Directuseofline-probeassaysonsmear-negativeclinical

specimens is not recommended.

Theuseofcommercialline-probeassays,ratherthanin-

house assays, is recommended to ensure reliability and

reproducibility o results.

Adoptionofline-probeassaysdoesnoteliminatethe

need or conventional culture and DST capability; culture

remains necessary or denitive diagnosis o tuberculosis

in smear-negative patients, whereas conventional DST is

needed to diagnose XDR tuberculosis.

(Continuesinnextcolumn)

(Continued rom previous column)

LED-based microscopy (introduced in 200910)

WHOrecommendsthatconventionaluorescence

microscopy be replaced by LED microscopy using auramine

staininginallsettingswhereuorescencemicroscopyis

currently used, and that LED microscopy be phased in as an

alternative or conventional Ziehl-Neelsen light microscopy

in both high-volume and low-volume laboratories.

TheswitchtoLEDmicroscopyshouldbeundertaken

through a careully phased implementation plan, with use

o LED technologies that meet WHO specications.

Non-commercial culture DST methods (introduced

in 200910)

WHO recommends that selected non-commercial culture andDST methods be used as an interim solution in resource-

constrained settings, in reerence laboratories, or in those

with sufcient culture capacity, while capacity or genotypic

and/or automated liquid culture and DST are being

developed. With due consideration o the above issues, WHO

endorses the selective use o one or more o the ollowing

non-commercial culture and DST methods:

Microscopicallyobserveddrugsusceptibilityasdirector

indirect tests, or rapid screening o patients suspected o

having MDR tuberculosis.

Nitratereductaseassay,asdirectorindirecttests,for

screening o patients suspected o having MDR

tuberculosis, and acknowledging that time to detection o

MDR tuberculosis in indirect application would not beaster than conventional DST methods using solid culture.

Colorimetricredoxindicatormethods,asindirecttestson

Mycobacterium tuberculosis isolates rom patients suspected

o having MDR tuberculosis, and acknowledging that time

to detection o MDR tuberculosis would not be aster (but

wouldbelessexpensive)thanconventionalDSTmethods

using commercial liquid culture or molecular line-probe

assays.

Same-day diagnosis by microscopy (introduced in

200910):

WHOrecommendsthatcountriesthathavesuccessfully

implemented the current WHO policy or a two-specimen

case-nding strategy consider a switch to the same-day-diagnosis approach, especially in settings where patients

are likely to deault rom the diagnostic process.

Countriesthatarestillusingthethree-specimencase-

nding strategy consider a gradual change to the same-

day-diagnosisapproach,onceWHO-recommendedexternal

microscopy quality-assurance systems are in place and good

quality microscopy results have been documented.

Changestoasame-day-diagnosisstrategybeprecededbya

detailed situation assessment o the programmatic,

logistical, and operational implications within countries,

and supported by a careully phased implementation plan.

Source: WHO.169174 DST=drug-susceptibility testing. MDR=multidrug resistant.

XDR=extensivelydrugresistant.LED=light-emittingdiode.


9/18

Series


Fluorescence microscopy is widely used in high income countries since it oers increased sensitivity,and has logistical advantages such as less techniciantime,176 but is rarely used in resource limited countries.Several light emitting diode (LED) microscopes that canbe used in uorescence microscopy have been developedin the past ew years.177 They are inexpensive, robust,consume little electricity, are highly sensitive, and needless technician time than does Ziehl Neelsen microscopy.WHO recommended that conventional uorescencemicroscopy be replaced by LED microscopy in allsettings and that LED microscopy be phased in as analternative or conventional Ziehl Neelsen microscopyin both high volume and low volume laboratories. 171Eorts are also underway to minimise diagnostic delaysand to improve system efciency by optimising thenumber o specimens that are needed and the way inwhich they are collected (eg, so called same daydiagnosis, using two sputum smears collected on thesame day).178 In act, WHO recently endorsed the use othe same day microscopy approach. 171

The growing evidence base or tuberculosis diagnosticsEvidence presented to WHO expert committees over thepast ew years that has inormed the endorsement onew technologies (panel 2) included easibility studiesassessing the technology aspect, evaluation studies o

the nal manuactured product, and large scaledemonstration projects ocused on cost, eect, andpracticability o use in real world settings. 179 Thisextensive and robust platorm o evidence is timeconsuming and expensive to generate, but necessary tolend support to evidence based policies or tuberculosisdiagnosis.172 An outline or the development otuberculosis diagnostics, published by the Stop TBPartnerships New Diagnostics Working Group,ormalised this evidence platorm by describing thedevelopment pathway or new tuberculosis diagnosticsin detail (gure 4), rom initial concept to development,evaluation, delivery, scale up, and impact assessment. 180

WHO has assumed a leadership role in ensuring thatnew tuberculosis diagnostic policies are evidence based,179

and in line with the grading o recommendationsassessment, development and evaluation (GRADE)approach to guideline development.172 To enable and helpwith this process, existing systematic reviews ontuberculosis diagnostics, policies, guidelines, andresearch agendas or diagnosis have been compiled bythe Stop TB Partnerships New Diagnostics WorkingGroup.181 Panel 3 summarises the ndings o systematicreviews o various tuberculosis diagnostics.

Optimism or the utureThe product pipeline or the uture looks promising. In2009, data were published on the rst automatedmolecular test or tuberculosis, the Xpert MTB/RIF,which was co developed by the Foundation or InnovativeNew Diagnostics, Cepheid (Sunnyvale, CA, USA), and theUniversity o Medicine and Dentistry o New Jersey, NJ,USA.182 This assay, which was CE (Conormit Europenne)marked in 2009, avoids most o the pitalls o conventionalnucleic acid amplication tests (saety, contamination,ease o use, etc), can be done by sta with little training,and can be used or case detection or MDR screening.Data rom evaluation trials showed excellent perormancein both smear positive and smear negative patients, andhigh accuracy or determination o riampicin resistance.Thus, this highly sensitive and simple to use system candetect M tuberculosis directly rom sputum in less than

2 h.182 Data rom ongoing demonstration projects arelikely to be reviewed by WHO in 2010.

For the diagnosis o latent M tuberculosis inection,commercially available IGRAs have emerged as a strongalternative to the tuberculin skin test. These assays havevery high specicity and have specic logistical advantagescompared with the tuberculin skin test183 or diagnosis.IGRAs, however, have no role as rule in tests or activetuberculosis diagnosis or adults in endemic settings.183 Theuse o IGRAs is steadily increasing, with several countrieswith low and intermediate incidence opting to use them,mostly as ollow up tests in people with positive resultsrom tuberculin skin tests, especially in BCG vaccinatedpopulations.184A survey o IGRA guidelines showed muchdiversity in how various countries recommend and use

Figure 4: Schematic showing the pathway to tuberculosis diagnostics, rom concept to delivery

Source: Stop TB Partnerships New Diagnostics Working Group. Pathways to better diagnostics or tuberculosis: a blueprint or the development o TB diagnostics

(2009),180 and reproduced with permission rom author and publisher.

Epidemiologicaland public healthimpact

Needs

assessment

Upstream Downstream

Concept,feasibility,

and proof-of-principle

Development,optimisation(design-lock)

Evaluationstudies of

accuracy andreliability

Demonstrationstudies includingpatient outcomes

Evidence forscale-up, scale-up,

delivery andaccess

Policy


10/18

Series


IGRAs.184The two step approach (initial tuberculin skin test

ollowed by conrmatory IGRA testing) seems to be themost common strategy, partly because o economicconsiderations. The optimum strategy or IGRA use is yetto be established.

Although targeted testing and preventive therapy orlatent M tuberculosis inection is well established in low incidence countries, the exact role o testing and treatmentin disease endemic countries remains controversial.However, testing or latent M tuberculosis inection isreceiving increased attention in vulnerable subgroups,such as HIV inected people and childhood contacts oactive tuberculosis cases.185,186 A WHO policy or IGRAs isunder consideration or 2010.

Limitations o the existing diagnostics pipelineA simple, rapid, inexpensive point o care test or activetuberculosis that can perorm as well or better thanconventional smear microscopy, and can deliver resultswithin minutes without sophisticated equipment orlaboratory requirements, is still missing rom thedevelopment pipeline. Point o care diagnostic tests oerimportant potential advantages or control o diseasessuch as tuberculosis that need lengthy standardised,decentralised therapy.142,187,188 Patient, community, andactivist groups have urged or increased unding andresources to develop point o care tests, and specicationsor an ideal point o care test have been proposed. 189

The existing tuberculosis diagnostics pipeline is alsorestricted with respect to tests that address importantdiagnostic challenges, especially in HIV inected people,and children and adults with smear negative tuberculosis.Unortunately, most existing tests have showndisappointing perormance in smear negative tuberculosis.Conventional nucleic acid amplication tests haveinadequate sensitivity in patients with smear negativetuberculosis. Improved tests such as Xpert MTB/RIFmight have improved sensitivity in these patients, buturther validation is needed.182

Childhood tuberculosis is a well known diagnosticchallenge, and all available tests do poorly in cases opaucibacillary tuberculosis.190 Furthermore, since young

children are unable to produce sputum, alternativespecimens such as urine, saliva, or breath condensatewould be helpul to use. The absence o a gold standardor childhood tuberculosis and smear negativetuberculosis is an important impediment to rapidassessment o new diagnostic methods in these high risksubgroups. One potential solution to the problem o aninadequate gold standard would be to ollow up wellcharacterised cohorts o patients ater initial testing untiltuberculosis is denitely ruled in or out. This type ostudy could also assess whether use o a new test actuallyimproved patient important outcomes, rather thanexamining sensitivity and specicity only.

Although serological antibody tests or tuberculosishave potential as point o care tests, their perormance

Panel 3: Summary o ndings rom systematic reviews on tuberculosis

diagnostic tests

Diagnosis o active tuberculosis

Sputum-smear microscopy or pulmonary tuberculosis

FMisonaverage10%moresensitivethanisconventionalmicroscopy.Specicityof

both FM and conventional microscopy is similar. FM is associated with improved time

efciency.

LEDFMperformsequivalentlytoconventionalFM,withaddedbenetsoflowcost,

durability, and ability to use without a darkroom.

Centrifugationandovernightsedimentation,precededwithanyofseveralchemical

methods(includingbleach),isslightlymoresensitive(69%)thanisdirect

microscopy;specicitymightbeslightlydecreased(13%)bysputumprocessing

methods.

Whenserialsputumspecimensareexamined,themeanincrementalyieldand/or

increaseinsensitivityfromexaminationofthirdsputumspecimenrangesbetween

2%and5%.

Asame-day-diagnosisapproach(microscopyoftwoconsecutivespot-spotsputum

specimens) is equivalent, in terms o diagnostic accuracy, to conventional case-nding

strategies by microscopy.

NAATs or pulmonary and extrapulmonary tuberculosis

NAATshavehighspecicityandpositivepredictivevalue.NAATs,however,have

relatively lower (and highly variable) sensitivity and negative predictive value or all

formsoftuberculosis,especiallyinsmear-negativeandextrapulmonarydisease.

In-house(so-calledhomebrew)NAATsproducehighlyinconsistentresultscompared

with commercial, standardised NAATs.

Serological antibody detection tests or pulmonary and extrapulmonary tuberculosis

Commercialserologicaltestsforbothpulmonaryandextrapulmonarytuberculosis

produce highly inconsistent estimates o sensitivity and specicity; none o the assays

do well enough to replace microscopy.

Severalpotentialcandidateantigensforinclusioninanantibodydetection-based

diagnostic test or pulmonary tuberculosis in HIV-inected and uninected individuals

were identied.

Combinationsofselectantigensprovidehighersensitivitiesthandosingleantigens.

ADA or tuberculosis pleuritis, pericarditis, peritonitis

MeasurementofADAconcentrationsinpleural,pericardial,andasciticuidhashigh

sensitivityandspecicityforextrapulmonarytuberculosis.

Intereron or tuberculosis pleuritis

Pleuraluidinterferon-determinationisasensitiveandspecictestforthe

diagnosis o tuberculosis pleuritis.

Phage amplifcation assays or pulmonary tuberculosis Phage-basedassayshavehighspecicitybutlowerandvariablesensitivity.Current

commercial phage-based assays are limited by high rates o indeterminate results.

Automated liquid cultures or pulmonary tuberculosis

Automatedliquidculturesaremoresensitivethanaresolidcultures;timetodetection

is more rapid than or solid cultures.

Diagnosis o latent tuberculosis

TST or latent tuberculosis inection

IndividualswhohadreceivedBCGvaccinationaremorelikelytohaveapositive

TST; the eect o BCG on TST results is less ater 15 years; positive TST with

indurations o >15 mm are more likely to be the result o tuberculosis inection than

o BCG vaccination.

(Continuesonnextpage)


11/18

Series


thus ar has been disappointing.191,192 Urine mycobacterial

antigen (eg, lipoarabinomannan) detection tests areattractive options or point o care testing. Results ostudies o detection o urinary lipoarabinomannan havebeen variable but generally suboptimum, althoughsomewhat better in patients with advanced HIVinection.193197 Alternative detection targets are beingsought, such as M tuberculosis tr DNA. 198

Several options are being explored or simpler, lessexpensive point o care and multiplexed assay ormats inthe uture, including manual molecular testing that can bedone in peripheral settings, lab on chip approaches thatcan be used to detect several inections simultaneously,antigen detection on highly sensitive platorms, andantibody detection with panels o recently identied

antigens o diagnostic value. Owing to growing interest andunding or new methods and biomarkers, several agencies,industries, and groups are working on developing point o care platorms or tuberculosis, including novel serologicalassays, detection o volatile organic compounds in breath,handheld molecular devices, microchip technologies, andtests that exploit approaches such as microuidics,nanotechnology, proteomics, and metabolomics.139,162,168,187

Overcoming barriers or implementation in tuberculosiscontrol programmesWhat will be the outcome o all this technologydevelopment or tuberculosis diagnostics, and how canthis progress be translated into concrete gains in controlo tuberculosis? The eect o new tests will dependlargely on the extent o their introduction into the globalpublic sector, which will itsel depend partly on policydecisions made by international technical agencies suchas WHO, and by donors, and ultimately by nationaltuberculosis programmes in countries o low and middleincome. So ar, most evaluations o diagnostic methodshave reported only sensitivity and specicity; many othese studies were poorly designed and incompletelyreported.172,199 Some have assessed time to test result, anda ew have reported unit costs. However, to lend supportto the introduction o new diagnostic technologies,broader evidence is needed, including implementation

issues.180 For example, the perormance o new tests inprogrammatic conditions should be studied; tests doneby experts in careully controlled research settings arenot likely to be indicative o uture eld perormance. Inaddition to unit costs, costing studies should includecosts or labour, equipment depreciation, initial andongoing training, supervision, and quality control.200

Future studies o completed diagnostic products have togo beyond test accuracy and aim to generate evidence orthe incremental value o new tests, their eect on patientoutcomes, and their use or diagnostic decision makingand cost eectiveness. 172,199 Operational research is alsoessential to improve service delivery and to understandwhy diagnosis is delayed or missed, and to guide optimumimplementation o new methods. To help with these types

(Continued rom previous page)

TheeectonTSTofBCGreceivedininfancyissmall,especially10yearsafter

vaccination. BCG received ater inancy produces more requent, more persistent, and

larger TST reactions. NTM inection is not a clinically important cause o alse-positive

TST, apart rom in populations with a high prevalence o NTM sensitisation and a very

low prevalence o tuberculosis inection.

T-cell-based IGRAs or latent tuberculosis inection

IGRAshaveexcellentspecicity(higherthantheTST),andareunaectedbyprevious

BCG vaccination.

IGRAscannotdistinguishbetweenlatenttuberculosisinfectionandactive

tuberculosis, and have no role or active tuberculosis diagnosis in adults.

IGRAscorrelatewellwithmarkersoftuberculosisexposureinlow-incidence

countries.

IGRAsensitivityvariesacrosspopulationsandtendstobelowerinhigh-endemic

countries and in HIV-inected individuals.

Diagnosis o drug-resistant tuberculosis

Phage amplifcation assays or rapid detection o riampicin resistance

Commercialphageamplicationassaysproducevariableresultswhenuseddirectlyon

sputum specimens.

Studieshaveraisedconcernsaboutcontamination,falsepositiveresults,andtechnical

assay ailures.

Line-probe assays: INNO-LiPA Ri and GenoType MTBDR assaysor rapid detection o riampicin

resistance

TheINNO-LiPARifassayisahighlysensitiveandspecictestforthedetectionof

riampicin resistance in culture isolates. The test has lower sensitivity when used

directly on clinical specimens.

GenoTypeMTBDRassayshaveexcellentsensitivityandspecicityforrifampicinresistance, even when directly used on clinical specimens.

CRI methods and NRA or rapid detection o riampicin and isoniazid resistance

Colorimetricmethodsaresensitiveandspecicforthedetectionofrifampicinand

isoniazidresistanceincultureisolates.CRIsuseinexpensivenon-commercialsupplies

and equipment and have a rapid turnaround time (7 days).

NRAhashighaccuracywhenusedtodetectrifampicinandisoniazidresistancein

culture isolates. Data or its use when directly applied to clinical specimens are scarce,

butresultsarepromising.NRAissimple,usesinexpensivenon-commercialsupplies

and equipment, and has a rapid turnaround time (714 days) compared with

conventional methods.

MODS or rapid detection o riampicin and isoniazid resistance

MODShashighaccuracywhentestingforrifampicinresistance,butshowsslightly

lower sensitivity when detecting isoniazid resistance. MODSseemstodoequallywellwithuseofdirectpatientspecimensandculture

isolates.

MODSusesnon-commercialsuppliesandequipment,andhasarapidturnaroundtime

(10 days) compared with conventional methods.

TLA or rapid detection o riampicin and isoniazid resistance

DataassessingTLAforthedetectionofdrugsusceptibilityarescarce;

however,allstudiessofarhavereported100%concordancewiththeirreference

standards.

TLAusesinexpensivenon-commercialsuppliesandequipment,andhasarapid

turnaround time (11 days) compared with conventional methods.

FM=uorescencemicroscopy.LED=light-emittingdiode.NAATs=nucleicacidamplicationtests.ADA=adenosinedeaminase.

TST=tuberculinskintest.NTM=non-tuberculousmycobacterial.IGRAs=interferon-releaseassays.CRI=colorimetric

redoxindicator.NRA=nitratereductaseassays.MODS=microscopicallyobserveddrugsusceptibility.TLA=thin-layeragar.

Adapted rom reerence 172 and reproduced with permission rom the author, under creative commons open access license.


12/18

Series


o research, the TB Research Movementrecently

initiated by the Stop TB Partnership and WHOisengaging tuberculosis researchers, tuberculosis pro gramme managers, and aected communities in acollaborative and concerted strategic eort to increase thescope, scale, and speed o tuberculosis research across thecontinuum, linking together basic research, developmento new methods, and operational research.201

The GRADE approach,202 now being used by WHO,was originally designed or interventions such as drugsand vaccines, or which the product is the healthintervention itsel and the use o the product can belargely judged by its saety and eectiveness alone.Diagnostics, however, are only the start o a healthintervention, and their eect will depend on where and

how they are used, and what clinical decisions they canlend support to. The GRADE approach has been adaptedand applied to diagnostic tests,203,204but will need to beurther adapted or supplemented by careul considerationso the diversity and challenges o health systems whenexamining diagnostics aimed at public sector populationsin developing countries. Although GRADE has itslimitations and can be improved and adapted ortuberculosis diagnostics, it is a major advance comparedwith the conventional policy making process. 205

Inadequate unding is another major barrier that needsto be overcome. New tuberculosis diagnostics will be ono practical value i they are not readily available at pointso care in endemic areas, and i they are not takenseriously by governments o developing countries.Insufcient commitment to tuberculosis control by manydeveloping country governments is largely responsibleor poor programme perormance. The Global Plan toStop TB, 20062015, estimated that at least US$9 billion($900 million per year) should be spent on tuberculosisresearch and development between 2006 and 2015 todevelop new drugs, diagnostics, and vaccines.167 Thebudget needed or tuberculosis diagnostics was$516 million; yet according to the 2009 Treatment ActionGroup and Stop TB Partnership reports, development ortuberculosis diagnostics received only $50 million in2008.206,207 This amount represented only 10% o the total

unding or tuberculosis research and development.207Furthermore, philanthropic grants are outstrippinggovernment unding or tuberculosis research.207

To overcome this worrisome trend in declining public sector investment, governments in all countries,especially industrialised countries, need to increasetheir unding or tuberculosis research and develop ment.167,206,207 Emerging and rapidly growing economiessuch as China, India, Brazil, and South Arica can andshould increase their investments in tuberculosis,especially since these countries account or a largeproportion o the global tuberculosis burden. Countriessuch as China and India can also make a big contributionby producing locally manuactured, low cost generictuberculosis drugs, diagnostics, and vaccines. In the

long term, these countries have the potential to

spearhead the next wave o innovation in tuberculosisresearch and development.

ConclusionsThe need or a more accurate, inexpensive point o caretuberculosis diagnostic test that is applicable in tuber culosis and HIV endemic areas is greater nowadaysthan ever beore, and will be crucial or achieving globaltuberculosis control. Several modelling studies208215suggest that new diagnostics or tuberculosis diseaseand MDR tuberculosis could have an important eectwithin populations, especially in disease endemiccountries, although improving population health andhealth services, and economic growth, might be as

important.216,217 Clinical and eld studies are needed toassess whether programmatic introduction o newdiagnostics contributes to improved individual patientoutcomes and a measurable benecial public healtheect. Ater nearly a century o neglect and under investment, the tuberculosis diagnostics pipeline hasrapidly grown, with several technologies showing greatpromise. Indeed, several have already been endorsed byWHO and are being introduced into clinical use. Thisprogress needs to be translated into improving the liveso patients with tuberculosis, and reducing the utureincidence o tuberculosis. This aim can and must beachieved, but will need strong political commitment,sustained unding, and engagement o public andprivate stakeholders and civil society. Donors andgovernments have to synergise their activities to ensuremaximum programme perormance or optimum careor patients with tuberculosis and with both tuberculosisand HIV inection.

To advance the area o tuberculosis biomarkers to thatneeded or registration, substantial investments will berequired to undertake the necessary studies. Studies oMDR tuberculosis have been advocated by some as a richsource o poor outcomes or biomarkers research.However, whether markers predicting ailure due toresistance will necessarily also predict relapse (whichseems somewhat paradoxically to occur inrequently in

MDR tuberculosis) is uncertain.218 As studies areundertaken to shorten MDR treatment, we might need torely on biomarkers or relapse developed in drug sensitivedisease to guide them. Tuberculosis incidence rarelyapproaches 1% in the general population even in high prevalence countries, hampering prospective studies.Ethical concerns preclude natural history studies in high risk patients, such as children or people with HIVinection, meaning that isoniazid preventive therapyshould be oered. As a result, studies to validate markersthat predict the transition rom health to illness (or viceversa) in these populations will necessarily be large andprotracted. I the plethora o potential biomarkersdescribed here is to be converted into clinically useultests, we not only need continuing research, but also


13/18

Series


improved unding to synergise and improve

multidisciplinary cross cutting collaborations betweenscientists working with cohorts o patients and contactsparticipating in clinical trials o new drug regimens,diagnostics, and vaccines.

Contributors

AZ conceived the article outline and selected and assigned authorsroles. The literature search and the rst and subsequent drats o thereport were developed by RSW (biomarkers) and MP (diagnostics).TMD, GW, and AZ contributed to writing o the biomarkers section, andMDP, DM, and AZ contributed to writing o the diagnostics section. AZmerged the two sections as the nal article with contributions rom allauthors. All authors read and approved the nal versions o the twosections beore submission.

Steering committee

This article is part oThe LancetSeries on tuberculosis, which wasdeveloped and coordinated by Alimuddin Zumla (University College

London Medical School, London, UK); Mario C Raviglione (Stop TBDepartment, WHO, Geneva, Switzerland); and Ben Marais (Universityo Stellenbosch, Stellenbosch, South Arica).

Conficts o interests

RSW is employed by Pzer, USA. AZ is principal investigator o theEuropeAID Active Detection o Active Tuberculosis (ADAT) andEuropean Union Framework 7 Trans renal DNA (EU FW7 TrDNA)projects, which are assessing new tuberculosis diagnostics, and serveson the Stop TB Research Movement Task Force. MDP is the ChieScientic Ofcer o Foundation or Innovative New Diagnostics(FIND), Geneva, Switzerland, a non prot agency that works withseveral industry partners in developing and evaluating newdiagnostics or neglected inectious diseases. MP serves as an externalconsultant or FIND; serves as a co chair o the Stop TB PartnershipsNew Diagnostics Working Group; and serves as chair o the TaskForce o the Stop TB Research Movement. GW receives support rom

GlaxoSmithKline, Bill & Melinda Gates Foundation, AerasFoundation, TB Alliance, and EDCTP. TMD receives support romEDCTP and EU FP7. None o the unding agencies had any role in thedevelopment or submission o this report. DM declares that he has noconicts o interest.

Acknowledgments

AZ receives support rom the EU FP7, EDCTP, Global Alliance or TBDrug development, EuropeAID ADAT, UK Medical Research Council,and the UK NIHR CBRC. TMD receives support rom the EU FP7;EDCTP; Research Council o Norway; Netherlands Arican partnershipor capacity development and clinical interventions against poverty related diseases; and the Danish Agency or Science, Technology andInnovation. MP is a recipient o a New Investigator Award rom theCanadian Institutes o Health Research (CIHR), and receives supportrom EDCTP and European Commission (TBSusgent, EU FP7). DM isrecipient o a career award rom the Fonds de la Recherche en Sant duQubec (FRSQ). None o the unding agencies had any role in the

development or submission o this report.

Reerences1 Biomarkers working group. Biomarkers and surrogate endpoints:

preerred denitions and conceptual ramework.Clin Pharmacol Ther2001; 69: 8995.

2 Mitchison DA. Assessment o new sterilizing drugs or treatingpulmonary tuberculosis by culture at 2 months. Am Rev Respir Dis1993; 147: 106263.

3 The Tuberculosis Trials Consortium. Riapentine and isoniazid oncea week versus riampicin and isoniazid twice a week or treatment odrug susceptible pulmonary tuberculosis in HIV negative patients: arandomised clinical trial. Lancet2002; 360: 52834.

4 Wallis RS, Wang C, Doherty TM, et al. Biomarkers ortuberculosis disease activity, cure, and relapse. Lancet Inect Dis2010; 10: 6869.

5 Rustomjee R, Diacon AH, Allen J, et al. Early bactericidal activityand pharmacokinetics o the Diarylquinoline TMC 207 in

pulmonary tuberculosis. Antimicrob Agents Chemother2008;52: 283135.

6 Davies GR, Brindle R, Khoo SH, Aarons LJ. Use o nonlinearmixed eects analysis or improved precision o early

pharmacodynamic measures in tuberculosis treatment.Antimicrob Agents Chemother2006; 50: 315456.

7 Epstein MD, Schluger NW, Davidow AL, Bonk S, Rom WN,Hanna B. Time to detection o Mtb tuberculosis in sputumculture correlates with outcome in patients receiving treatmentor pulmonary tuberculosis. Chest1998; 113: 37986.

8 Wallis RS, Perkins M, Phillips M, et al. Induction o the antigen85 complex oM. tuberculosis in sputum: a determinant o outcomein pulmonary tuberculosis. J Inect Dis 1998; 178: 111521.

9 Rustomjee R, Lienhardt C, Kanyok T, et al. A Phase II study o thesterilising activities o ooxacin, gatioxacin and moxioxacin inpulmonary tuberculosis. Int J Tuberc Lung Dis 2008; 12: 12838.

10 Wallis RS, Perkins M, Phillips M, et al. Predicting the outcome otherapy or pulmonary tuberculosis. Am J Respir Crit Care Med2000; 161: 107680.

11 Wallis RS, Phillips M, Johnson JL, et al. Inhibition o INH induced expression oM. tuberculosis antigen 85 in sputum:a potential surrogate marker in TB chemotherapy trials.Antimicrob Agents Chemother2001; 45: 130204.

12 Desjardin LE, Perkins MD, Wolski K, et al. Measurement osputum Mtb tuberculosis messenger RNA as a surrogate orresponse to chemotherapy. Am J Respir Crit Care Med1999;160: 20310.

13 Li L, Mahan CS, Palaci M, et al. Sputum MycobacteriumtuberculosismRNA as a marker o bacteriologic clearance inresponse to anti tuberculosis therapy. J Clin Microbiol2010;48: 4651.

14 Garton NJ, Waddell SJ, Sherratt AL, et al. Cytological andtranscript analyses reveal at and lazy persister like bacilli intuberculous sputum. PLoS Med2008; 5: e75.

15 Cannas A, Goletti D, Girardi E, et al. Mycobacterium tuberculosisDNA detection in soluble raction o urine rom pulmonarytuberculosis patients. Int J Tuberc Lung Dis 2008; 12: 14651.

16 Cannas A, Calvo L, Chiacchio T, et al. Mycobacterium tuberculosisDNA detection in the urine rom pulmonary tuberculosis patients

(abstract). Seventh International Conerence on the Pathogenesiso Mycobacterial Inections; June, 2008; Saltsjobaden, Sweden(abstr P17: 39).

17 Shah M, Variava E, Holmes CB, et al. Diagnostic accuracy o aurine lipoarabinomannan test or tuberculosis in hospitalizedpatients in a high HIV prevalence setting.J Acquir Immune Defc Syndr2009; 52: 14551.

18 Mutetwa R, Boehme C, Dimairo M, et al. Diagnostic accuracy ocommercial urinary lipoarabinomannan detection in Aricantuberculosis suspects and patients. Int J Tuberc Lung Dis 2009;13: 125359.

19 Reither K, Saatho E, Jung J, et al. Low sensitivity o a urineLAM ELISA in the diagnosis o pulmonary tuberculosis.BMC Inect Dis 2009; 9: 141.

20 Daley P, Michael JS, Hmar P, et al. Blinded evaluation ocommercial urinary lipoarabinomannan or active tuberculosis:a pilot study. Int J Tuberc Lung Dis 2009; 13: 98995.

21 Boehme C, Molokova E, Minja F, et al. Detection o mycobacteria l

lipoarabinomannan with an antigen capture ELISA inunprocessed urine o Tanzanian patients with suspectedtuberculosis. Trans R Soc Trop Med Hyg2005; 99: 893900.

22 Tessema TA, Bjune G, Assea G, Svenson S, Hamasur B,Bjorvatn B. Clinical and radiological eatures in relation to urinaryexcretion o lipoarabinomannan in Ethiopian tuberculosispatients. Scand J Inect Dis 2002; 34: 16771.

23 Choudhry V, Saxena RK. Detection o Mycobacterium tuberculosisantigens in urinary proteins o tuberculosis patients.Eur J Clin Microbiol Inect Dis 2002; 21: 15.

24 Singh KK, Dong Y, Hinds L, et al. Combined use o serum andurinary antibody or diagnosis o tuberculosis. J Inect Dis 2003;188: 37177.

25 Kashino SS, Pollock N, Napolitano DR, Rodrigues V Jr,Campos Neto A. Identication and characterization oMycobacterium tuberculosis antigens in urine o patients withactive pulmonary tuberculosis: an innovative and alternativeapproach o antigen discovery o useul microbial molecules.

Clin Exp Immunol2008; 153: 5662.


14/18

Series


26 Napolitano DR, Pollock N, Kashino SS, Rodrigues V Jr,Campos Neto A. Identication o Mycobacterium tuberculosis

ornithine carboamyltranserase in urine as a possible molecularmarker o active pulmonary tuberculosis. Clin Vaccine Immunol2008; 15: 63843.

27 Syhre M, Chambers ST. The scent o Mycobacterium tuberculosis.Tuberculosis (Edinb ) 2008; 88: 31723.

28 Phillips M, Cataneo RN, Condos R, et al. Volatile biomarkers opulmonary tuberculosis in the breath. Tuberculosis (Edinb) 2007;87: 4452.

29 Wallis RS, Palaci M, Vinhas S, et al. A whole blood bactericidalassay or tuberculosis. J Inect Dis 2001; 183: 130003.

30 Janulionis E, Soer C, Song HY, Wallis RS. Lack o activity o oralcloazimine against intracellular M. tuberculosis in whole bloodculture. Antimicrob Agents Chemother2004; 48: 313335.

31 Wallis RS, Vinhas SA, Johnson JL, et al. Whole blood bactericidalactivity during treatment o pulmonary tuberculosis. J Inect Dis2003; 187: 27078.

32 Veenstra H, Crous I, Brahmbhatt S, et al. Changes in the kinetics o

intracellular IFN gamma production in TB patients duringtreatment. Clin Immunol2007; 124: 33644.

33 Carrara S, Vincenti D, Petrosillo N, Amicosante M, Girardi E,Goletti D. Use o a T cell based assay or monitoring efcacy oantituberculosis therapy. Clin Inect Dis 2004; 38: 75456.

34 Sauzullo I, Mengoni F, Lichtner M, et al. In vivo and in vitro eectso antituberculosis treatment on mycobacterial intereron gammaT cell response. PLoS One 2009; 4: e5187.

35 Aiken AM, Hill PC, Fox A, et al. Reversion o the ELISPOT testater treatment in Gambian tuberculosis cases. BMC Inect Dis2006; 6: 66.

36 Kobashi Y, Mouri K, Yagi S, Obase Y, Miyashita N, Oka M. Transitionalchanges in T cell responses to Mycobacterium tuberculosis specicantigens during treatment. J Inect2009; 58: 197204.

37 Ribeiro S, Dooley K, Hackman J, et al. T SPOT.TB responses duringtreatment o pulmonary tuberculosis. BMC Inect Dis 2009; 9: 23.

38 Wassie L, Demissie A, Asea A, et al. Ex vivo cytokine mRNA levelscorrelate with changing clinical status o ethiopian TB patients and

their contacts over time. PLoS One 2008; 3: e1522.39 Dheda K, Chang JS, Breen RA, et al. In vivo and in vitro studies o

a novel cytokine, interleukin 4delta2, in pulmonary tuberculosis.Am J Respir Crit Care Med2005; 172: 50108.

40 Siawaya JF, Bapela NB, Ronacher K, Beyers N, van Helden P,Walzl G. Dierential expression o interleukin 4 (IL 4) and IL 4delta 2 mRNA, but not transorming growth actor beta (TGF beta),TGF beta RII, Foxp3, gamma intereron, T bet, or GATA 3 mRNA,in patients with ast and slow responses to antituberculosistreatment. Clin Vaccine Immunol2008; 15: 116570.

41 Wallis RS, Heland MS, Whalen C, et al. Immune activation,allergic drug toxicity, and mortality in HIV positive tuberculosis.Tuber Lung Dis 1996; 77: 51623.

42 Immanuel C, Rajeswari R, Rahman F, Kumaran PP,Chandrasekaran V, Swamy R. Serial evaluation o serum neopterinin HIV seronegative patients treated or tuberculosis.Int J Tuberc Lung Dis 2001; 5: 18590.

43 Turgut T, Akbulut H, Deveci F, Kacar C, Muz MH. Seruminterleukin 2 and neopterin levels as useul markers or treatmento active pulmonary tuberculosis. Tohoku J Exp Med2006;209: 32128.

44 Hosp M, Elliott AM, Raynes JG, et al. Neopterin, beta2 microglobulin, and acute phase proteins in HIV 1 seropositiveand seronegative Zambian patients with tuberculosis. Lung1997;175: 26575.

45 Fuchs D, Hausen A, Koer M, Kosanowski H, Reibnegger G,Wachter H. Neopterin as an index o immune response in patientswith tuberculosis. Lung1984; 162: 33746.

46 Baylan O, Balkan A, Inal A, et al. The predictive value o serumprocalcitonin levels in adult patients with active pulmonarytuberculosis. Jpn J Inect Dis 2006; 59: 16467.

47 Nyamande K, Lalloo UG. Serum procalcitonin distinguishes CAPdue to bacteria, Mycobacterium tuberculosis and PJP.Int J Tuberc Lung Dis 2006; 10: 51015.

48 Polzin A, Pletz M, Erbes R, et al. Procalcitonin as a diagnostic tool

in lower respiratory tract inections and tuberculosis. Eur Respir J2003; 21: 93943.

49 Prat C, Dominguez J, Andreo F, et al. Procalcitonin and neopterincorrelation with aetiology and severity o pneumonia. J Inect2006;

52: 16977.50 Schleicher GK, Herbert V, Brink A, et al. Procalcitonin and

C reactive protein levels in HIV positive subjects with tuberculosisand pneumonia. Eur Respir J2005; 25: 68892.

51 Kandemir O, Uluba B, Polat G, Sezer C, Camdeviren H, Kaya A.Elevation o procalcitonin level in patients with pulmonarytuberculosis and in medical sta with close patient contact.Arch Med Res 2003; 34: 31114.

52 Wallis RS, van Vuuren C, Potgieter S. Adalimumab treatment olie threatening tuberculosis. Clin Inect Dis 2009; 48: 142932.

53 Tang BM, Eslick GD, Craig JC, McLean AS. Accuracy oprocalcitonin or sepsis diagnosis in critically ill patients: systematicreview and meta analysis. Lancet Inect Dis 2007; 7: 21017.

54 Lawn SD, Wiktor S, Coulibaly D, Ackah AN, Lal RB. SerumC reactive protein and detection o tuberculosis in personsco inected with the human immunodeciency virus.Trans R Soc Trop Med Hyg2001; 95: 4142.

55 Bajaj G, Rattan A, Ahmad P. Prognostic value o C reactive proteinin tuberculosis. Indian Pediatr1989; 26: 101013.

56 Scott GM, Murphy PG, Gemidjioglu ME. Predicting deteriorationo treated tuberculosis by corticosteroid reserve and C reactiveprotein. J Inect1990; 21: 6169.

57 Plit ML, Theron AJ, Fickl H, Van Rensburg CE, Pendel S,Anderson R. Inuence o antimicrobial chemotherapy and smokingstatus on the plasma concentrations o vitamin C, vitamin E,beta carotene, acute phase reactants, iron and lipid peroxides inpatients with pulmonary tuberculosis. Int J Tuberc Lung Dis 1998;2: 59096.

58 Lee JH, Chang JH. Changes o plasma interleukin 1 receptorantagonist, interleukin 8 and other serologic markers duringchemotherapy in patients with active pulmonary tuberculosis.Korean J Intern Med2003; 18: 13845.

59 Baynes R, Bezwoda W, Bothwell T, Khan Q, Mansoor N. Thenon immune inammatory response: serial changes in plasmairon, iron binding capacity, lactoerrin, erritin and C reactive

protein. Scand J Clin Lab Invest1986; 46: 695704.60 Walzl G, Ronacher K, Djoba Siawaya JF, Dockrell HM. Biomarkers

or TB treatment response: Challenges and uture strategies. J Inect2008; 57: 10309.

61 Lai CK, Wong KC, Chan CH, et al. Circulating adhesion moleculesin tuberculosis. Clin Exp Immunol1993; 94: 52226.

62 Demir T, Yalcinoz C, Keskinel I, Demiroz F, Yildirim N.sICAM 1 as a serum marker in the diagnosis and ollow up otreatment o pulmonary tuberculosis. Int J Tuberc Lung Dis 2002;6: 15559.

63 Mukae H, Ashitani J, Tokojima M, Ihi T, Kohno S, Matsukura S.Elevated levels o circulating adhesion molecules in patients withactive pulmonary tuberculosis. Respirology2003; 8: 32631.

64 Djoba Siawaya JF, Bapela NB, Ronacher K, et al. Immuneparameters as markers o tuberculosis extent o disease and earlyprediction o anti tuberculosis chemotherapy response. J Inect2008; 56: 34047.

65 Eugen Olsen J, Gustason P, Sidenius N, et al. The serum level o

soluble urokinase receptor is elevated in tuberculosis patients andpredicts mortality during treatment: a community study romGuinea Bissau. Int J Tuberc Lung Dis 2002; 6: 68692.

66 Rosas Taraco AG, Salinas Carmona MC, Revol A, Rendon A,Caballero Olin G, Arce Mendoza AY. Expression o CDllc in bloodmonocytes as biomarker or avorable response to antituberculosistreatment. Arch Med Res 2009; 40: 12831.

67 Agrano D, Fernandez Reyes D, Papadopoulos MC, et al.Identication o diagnostic markers or tuberculosis by proteomicngerprinting o serum. Lancet2006; 368: 101221.

68 Brahmbhatt S, Black GF, Carroll NM, et al. Immune markersmeasured beore treatment predict outcome o intensive phasetuberculosis therapy. Clin Exp Immunol2006; 146: 24352.

69 Mistry R, Cli JM, Clayton C, et al. Gene expression patterns inwhole blood identiy subjects at risk or recurrent tuberculosis.J Inect Dis 2007; 195: 35765.

70 Jacobsen M, Repsilber D, Gutschmidt A, et al. Candidatebiomarkers or discrimination between inection and diseasecaused by Mycobacterium tuberculosis. J Mol Med2007; 85: 61321.


15/18

Series


71 Lalvani A, Pathan AA, Durkan H, et al. Enhanced contact tracingand spatial tracking o Mycobacterium tuberculosisinection by

enumeration o antigen specic T cells. Lancet2001; 357: 201721.72 Jackson Sillah D, Hill PC, Fox A, et al. Screening or tuberculosis

among 2381 household contacts o sputum smear positive cases inThe Gambia. Trans R Soc Trop Med Hyg2007; 101: 594601.

73 Bakir M, Millington KA, Soysal A, et al. Prognostic value oa T cell based, intereron gamma biomarker in children withtuberculosis contact. Ann Intern Med2008; 149: 77787.

74 Aichelburg MC, Rieger A, Breitenecker F, et al. Detection andprediction o active tuberculosis disease by a whole bloodintereron gamma release assay in HIV 1 inected individuals.Clin Inect Dis 2009; 48: 95462.

75 Doherty TM, Demissie A, Olobo J, et al. Immune responses to theMycobacterium tuberculosis specic antigen ESAT 6 signalsubclinical inection among contacts o tuberculosis patients.J Clin Microbiol2002; 40: 70406.

76 Diel R, Loddenkemper R, Meywald Walter K, Niemann S,Nienhaus A. Predictive value o a whole blood IFN gamma assay orthe development o active tuberculosis disease ater recent inectionwith Mycobacterium tuberculosis. Am J Respir Crit Care Med2008;177: 116470.

77 Higuchi K, Harada N, Fukazawa K, Mori T. Relationship betweenwhole blood intereron gamma responses and the risk o activetuberculosis. Tuberculosis (Edinb) 2008; 88: 24448.

78 Petrucci R, Abu AN, Gurgel RQ, et al. Intereron gamma,intereron gamma induced protein 10, and tuberculin responses ochildren at high risk o tuberculosis inection. Pediatr Inect Dis J2008; 27: 107377.

79 Whittaker E, Gordon A, Kampmann B. Is IP 10 a better biomarkeror active and latent tuberculosis in children than IFNgamma?PLoS ONE2008; 3: e3901.

80 Ruhwald M, Bjerregaard Andersen M, Rabna P, Eugen Olsen J,Ravn P. IP 10, MCP 1, MCP 2, MCP 3, and IL 1RA hold promise asbiomarkers or inection with M. tuberculosis in a whole bloodbased T cell assay. BMC Res Notes 2009; 2: 19.

81 Demissie A, Wassie L, Abebe M, et al. The 6 kilodalton early

secreted antigenic target responsive, asymptomatic contacts otuberculosis patients express elevated levels o interleukin 4 andreduced levels o gamma intereron. Inect Immun 2006; 74: 281722.

82 Demissie A, Abebe M, Asea A, et al. Healthy individuals thatcontrol a latent inection with Mycobacterium tuberculosisexpresshigh levels o Th1 cytokines and the IL 4 antagonist IL 4delta2.J Immunol2004; 172: 693843.

83 Weldingh K, Andersen P. ESAT 6/CFP10 skin test predicts diseasein M. tuberculosis inected guinea pigs. PLoS One 2008; 3: e1978.

84 Edwards LB, Acquaviva FA, Livesay VT. Identication otuberculous inected: dual tests and density o reaction.Am Rev Respir Dis 1973; 108: 133439.

85 Ozdemir D, Cesur S, Annakkaya AN, et al. Serum neopterinconcentrations in healthy healthcare workers compared withhealthy controls and patients with pulmonary tuberculosis.Med Sci Monit2006; 12: CR52124.

86 Hot DF, Blazevic A, Abate G, et al. A new recombinant bacilleCalmette Guerin vaccine saely induces signicantly enhanced

tuberculosis specic immunity in human volunteers. J Inect Dis2008; 198: 1491501.

87 Soares AP, Scriba TJ, Joseph S, et al. Bacillus Calmette Guerinvaccination o human newborns induces T cells with complexcytokine and phenotypic proles. J Immunol2008; 180: 356977.

88 Hawkridge T, Scriba TJ, Gelderbloem S, et al. Saety andImmunogenicity o a New Tuberculosis Vaccine, MVA85A, inHealthy Adults in South Arica. J Inect Dis 2008; 198: 54452.

89 Davids V, Hanekom WA, Mansoor N, et al. The eect o bacillecalmette guerin vaccine strain and route o administration oninduced immune responses in vaccinated inants. J Inect Dis 2006;193: 53136.

90 Hanekom WA, Hughes J, Mavinkurve M, et al. Novel application oa whole blood intracellular cytokine detection assay to quantitatespecic T cell requency in eld studies. J Immunol Methods 2004;291: 18595.

91 Kampmann B, Gaora PO, Snewin VA, Gares MP, Young DB,Levin M. Evaluation o human antimycobacterial immunity usingrecombinant reporter mycobacteria. J Inect Dis 2000; 182: 895901.

92 Cheon SH, Kampmann B, Hise AG, et al. Bactericidal activity inwhole blood as a potential surrogate marker o immunity ater

vaccination against tuberculosis. Clin Diagn Lab Immunol2002;9: 90107.

93 Kampmann B, Tena GN, Mazazi S, Young D, Eley B, Levin M.A novel human in vitro system to evaluate antimycobacterialvaccines. Inect Immun 2004; 72: 640107.

94 Tena GN, Young DB, Eley B, et al. Failure to control growth omycobacteria in blood rom children inected with humanimmunodeciency virus, and its relationship to T cell unction.J Inect Dis 2003; 187: 154451.

95 Kampmann B, Tena Coki GN, Nicol M, Levin M, Eley B.Reconstitution o antimycobacterial immune responses inHIV inected children receiving HAART. AIDS 2006; 20: 101118.

96 Martineau AR, Wilkinson RJ, Wilkinson KA, et al. A single dose ovitamin D enhances immunity to mycobacteria.Am J Respir Crit Care Med2007; 176: 20813.

97 Saliu O, Soer C, Stein DS, Schwander SK, Wallis RS. Tumornecrosis actor blockers: dierential eects on mycobacterialimmunity. J Inect Dis 2006; 194: 48692.

98 Hot DF, Worku S, Kampmann B, et al. Investigation o therelationships between immune mediated inhibition omycobacterial growth and other potential surrogate markers oprotective mycobacterium tuberculosis immunity. J Inect Dis 2002;186: 144857.

99 Wallis RS, Vinhas S, Janulionis E. Strain specicity oantimycobacterial immunity in whole blood culture ater cure otuberculosis. Tuberculosis (Edinb) 2009; 89: 22124.

100 Canaday DH, Wilkinson RJ, Li Q, Harding CV, Silver RF,Boom WH. CD4(+) and CD8(+) T cells kill intracellularMycobacterium tuberculosis by a perorin and Fas/Fasligand independent mechanism. J Immunol2001; 167: 273442.

101 Cheng SH, Walker L, Poole J, et al. Demonstration o increasedanti mycobacterial activity in peripheral blood monocytes ater BCGvaccination in British school children. Clin Exp Immunol1988;74: 2025.

102 Conde MB, Eron A, Loredo C, et al. Moxioxacin versus

ethambutol in the initial treatment o tuberculosis: a double blind,randomised, controlled phase II trial. Lancet2009; 373: 118389.

103 Dorman SE, Johnson JL, Goldberg S, et al. Substitution omoxioxacin or isoniazid during intensive phase treatment opulmonary tuberculosis. Am J Respir Crit Care Med2009;180: 27380.

104 Burman WJ, Goldberg S, Johnson JL, et al. Moxioxacin versusethambutol in the rst 2 months o treatment or pulmonarytuberculosis. Am J Respir Crit Care Med2006; 174: 33138.

105 Nuermberger EL, Yoshimatsu T, Tyagi S, et al. Moxioxacin containing regimens o reduced duration produce a s table cure inmurine tuberculosis. Am J Respir Crit Care Med2004; 170: 113134.

106 Rebollo MJ, San Juan GR, Folgueira D, et al. Blood and urinesamples as useul sources or the direct detection o tuberculosis bypolymerase chain reaction. Diagn Microbiol Inect Dis 2006;56: 14146.

107 Torrea G, Van de Perre P, Ouedraogo M, et al. PCR based detectiono the Mycobacterium tuberculosis complex in urine o HIV inected

and uninected pulmonary and extrapulmonary tuberculosispatients in Burkina Faso. J Med Microbiol2005; 54: 3944.

108 Kawabulula M, Ahmed K, Nagatake T, et al. Evaluation o PCR based methods or the diagnosis o tuberculosis by identication omycobacterial DNA in urine samples. Int J Tuberc Lung Dis 2002;6: 73237.

109 Aceti A, Zanetti S, Mura MS, et al. Identication o HIV patientswith active pulmonary tuberculosis using urine based polymerasechain reaction assay. Thorax1999; 54: 14546.

110 Kumar V, Jakubiec W, Li X, et al. Saety, tolerability, pk, and wholeblood bactericidal activity (WBA) against Mycobacterium tuberculosiso single ascending doses o PNU 100480. ICAAC2009;49: F11217a (abstr).

111 Williams KN, Stover CK, Zhu T, et al. Promising anti tuberculosisactivity o the oxazolidinone PNU 100480 relative to linezolid in themurine model. Antimicrob Agents Chemother2008; 53: 131419.

112 Kim K, Visintin I, Alvero AB, Mor G. Development and validationo a protein based signatur

Biomarkers and diagnostics for tuberculosis

Documents

Transcript of Biomarkers and diagnostics for tuberculosis