Diagnosis and Management of Latent Tuberculosis Infectionperspectivesinmedicine.cshlp.org/content/5/11/a017830... · 2015-10-26 · Diagnosis and Management of Latent Tuberculosis

Diagnosis and Management of LatentTuberculosis Infection

Laura Munoz1,2, Helen R. Stagg2, and Ibrahim Abubakar2

1Infectious Diseases Department, Bellvitge University Hospital-IDIBELL, Barcelona 08970, Spain2Research Department of Infection and Population Health, University College London, LondonWC1E 6JB, United Kingdom

Correspondence: [email protected]

The post-2015 World Health Organization global tuberculosis strategy recognizes that elim-ination requires a focus on reducing the pool of latently infected individuals, an estimated30% of the global population, from which future tuberculosis cases would be generated.Tackling latent tuberculosis infection requires the identification and treatment of asymp-tomatic individuals to reduce the risk of progression to active disease. Diagnosis of latenttuberculosis infection is based on the detection of an immune response to Mycobacteriumtuberculosis antigens using either the tuberculin skin test or interferon-g release assays.Current treatment requires the use of antibiotics for at least 3 months. In this article, wereview the current knowledge of the natural history, immunology, and pathogenesis oflatent tuberculosis, describe key population groups for screening and risk assessment,discuss clinical management in terms of diagnosis and preventative treatment, and identifyareas for future research.

Latent tuberculosis infection (LTBI) is definedby the detection of a specific immune re-

sponse to Mycobacterium tuberculosis complex(MTC) antigens in a healthy subject (i.e., withno symptoms or signs of active tuberculosis[TB]). As M. tuberculosis can only be isolatedfrom humans when it is in an active phase, caus-ing illness, the detection of the presence of LTBIis wholly reliant on indirect measurements ofimmune reactivity to antigenic challenge.

By evading both innate and adaptive immu-nity, bacteria of the MTC are able to persist in adormant phase for several decades, or even forthe lifetime of the host. In �10% of all infected

individuals, a LTBI will progress to active repli-cation and cause TB disease. This can be pre-vented with antibiotic treatment; globally stan-dard regimens are of 629-mo duration with asingle drug, or at least 3 mo of two antibiotics.Increasing the number of drugs within the reg-imen increases the potential for consequentadverse events. Given the small proportion ofLTBI-positive individuals who develop activeTB, the imperfect effectiveness of LTBI treat-ment and associated its side effects, the criticalquestion is, therefore, who should be targetedfor treatment? This question is pertinent be-cause none of the tests currently available are

Editors: Stefan H.E. Kaufmann, Eric J. Rubin, and Alimuddin Zumla

Additional Perspectives on Tuberculosis available at www.perspectivesinmedicine.org

Copyright # 2015 Cold Spring Harbor Laboratory Press; all rights reserved; doi: 10.1101/cshperspect.a017830

Cite this article as Cold Spring Harb Perspect Med 2015;5:a017830

1

ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org

on October 18, 2020 - Published by Cold Spring Harbor Laboratory Presshttp://perspectivesinmedicine.cshlp.org/Downloaded from

http://perspectivesinmedicine.cshlp.org/

able to accurately predict future progression. Itis widely recognized that the risk of progres-sion is highest in young children, the immu-nosuppressed, and shortly after infection. Prag-matically, therefore, these individuals shouldbe targeted for LTBI screening and treatment. Inmost cases, this would lead to an “intention totest is intention to treat” approach, with an ex-ception for close contacts of multidrug-resistantTB cases, in which testing may be used to in-form “watchful waiting.”

In this article, we review the current knowl-edge of the natural history and immunology ofLTBI, describe key population groups for screen-ing and risk assessment, discuss clinical man-agement in terms of diagnosis and preventativetreatment, and finally look to the key fields forfuture research.

HISTORY (CONCEPT OF DORMANTBACILLI)

Although TB is one of the most ancient diseases,LTBI was only recognized at the end of the 19thcentury following (1) the accurate descriptionof the morbid anatomy of TB, its clinical evo-lution, and surrounding circumstances; (2) thefirst experimental animal models; and (3) thedevelopment of essential tools such as the mi-croscope. Soon after the discovery of the caus-ative agent of TB in 1882, Robert Koch createdthe Koch’s fluid (“a glycerine extract of purecultivations of tubercle bacilli” [Koch 1891]),for use as a live vaccine; it turned out to be anineffective preventive agent. Koch’s fluid, whichwas given the name of tuberculin, later provedto be a valuable diagnostic tool for active TB incattle. In the last years of the 19th century, ex-periments showed that cows with confirmed TBdeveloped a febrile reaction 14–18 h after tu-berculin administration. Apparently healthycows who presented the same systemic reactionwere slaughtered; some of the necropsies re-vealed active tubercles (Mi’fadyean 1891; Mar-tin and Robins 1898). Reactive cattle withouttubercles had their positive reaction attributedto dormant TB, as quiescent tuberculous fociwere identified in some of the necropsies (Mar-tin and Robins 1898).

Mantoux’s description of a new techniqueto administer tuberculin (the “tuberculin skintest” [TST]) in 1908, together with its improvedcomposition in 1934 (purified protein derivate[PPD]), made its use as a human diagnosticmethod possible, as systemic reactions werenot observed in tested subjects. The first trialsthat tested tuberculin on TB patients alsoshowed positive results in healthy controls. Itwas at this time that the notion of latent bacteriaor bacillary allergy emerged.

As for preventive purposes, the 20th centurysaw the introduction of several successful vac-cines, and there were great expectations for aneffective preparation against TB. In 1902, Beh-ring proposed to use the bacillus isolated fromhumans for the vaccination of cows. Calmetteand Guerin passaged Mycobacterium bovis,which has antigenic properties similar to thoseof the virulent M. tuberculosis bacilli and theability to provoke antibody formation in hu-mans, to obtain the attenuated Bacillus Cal-mette–Guerin (BCG) strain. BCG was success-fully tested in infants exposed to family casesof TB owing to the high mortality usually seenin such infants (72%) (Calmette 1931).

The immunological response to M. bovisand M. tuberculosis is highly similar, allowingthe effective use of BCG as a vaccine and tuber-culin as a diagnostic reagent. The proportion ofBCG-vaccinated people who have a positive tu-berculin skin test (TST) has been reported tovary from 0% to 90% (Wang et al. 2002). Thesewide differences may depend on dose, the man-ufacturer of the vaccine, age, and the intervalbetween vaccination and testing. Variation inthe cutoff for positivity used during TST testingalso explains some of the observed variation.Nevertheless, a positive reaction of .15 mmin someone who was vaccinated .15 yr beforeis unlikely to be related to BCG vaccination andwould likely reflect the presence of TB infection.

Questions were raised about the infectious-ness of the tubercle bacilli and how to preventthe spread of the disease. This led to early con-tact studies surrounding single active cases, withthe aim of finding latently infected individuals(positive TSTresponders but without TB symp-toms) and treating them to prevent the devel-

L. Munoz et al.

2 Cite this article as Cold Spring Harb Perspect Med 2015;5:a017830

ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



opment of active TB. Soon after the discovery ofstreptomycin (1944) and isoniazid (1952) foractive TB, the first experiments using isoniazidas a preventive treatment in an animal modelwere undertaken (1955). Initial clinical trials inhumans began in 1957, with results reported inthe mid-1960s. During the next few decades,five major randomized trials showed the bene-fits of isoniazid in preventing active TB.

NATURAL HISTORY AND PATHOGENESIS

Underpinning all attempts at the diagnosis andtreatment of LTBI is our understanding of thenatural history of TB, particularly the processthrough which LTBI is established, and the hostand bacterial factors influential in determininglatency and progression. The most importantreservoir of M. tuberculosis is humans, whereascattle are the equivalent host for M. bovis. Zoo-notic transmission of M. bovis was frequent inancient times; however, pasteurization of milkand the testing of herds decreased the publichealth impact of M. bovis in humans in wealth-ier countries. Many developing countries con-tinue to observe zoonotic transmission.

Humans with active pulmonary TB current-ly play the dominant role in the transmissionand maintenance of TB. Studies by Flugge in1899 (Hillier 1903) showed that both dried tu-berculous sputum dust and the minute dropscoughed out by pulmonary TB patients werecapable of surviving for several days while re-maining infectious. These tiny particles containtubercle bacilli inside droplet nuclei, are easilyinhaled, and constitute the route of entry andfirst contact with the healthy subject (Box 1).

Infection

On inhalation, M. tuberculosis reaches the mainairway and alveoli and faces the first cell-medi-ated barrier of the innate immune system. Alve-olar macrophages phagocytose the pathogenand isolate it through a process of membraneinvaginations that finally culminates in phago-some formation. From now on, two contendingforces compete for control—the bacterium’sstruggle for survival and macrophage’s attemptsto contain the infection.

After phagosome formation, infected mac-rophages and dendritic cells display M. tuber-culosis antigens via major histocompatibilitycomplex (MHC) class II molecules. These pro-fessional antigen-presenting cells expressingMTC molecules travel to the lymph system andthen into local mediastinal lymph nodes, wherethey will activate CD4þ T-helper cells (Th).These cells regulate the secretion of cytokinesvia endocrine and paracrine signaling andfunction in the formation and maintenance ofgranulomas by means of a proinflammatory re-sponse signal cascade (interferon-g [IFN-g],interleukin-2 [IL-2], and tumor necrosis fac-tor-a [TNF-a]), which leads to the attractionof mononuclear cells and T lymphocytes. Thesecell types, together with B lymphocytes anddendritic cells, build up a granuloma, “the hall-mark tissue reaction of TB” (Gengenbacher andKaufmann 2012). Other functions of CD4þ

Th cells include activation of CD8þ cytotoxicT cells and B lymphocytes, as well as differenti-ation into diverse Th cell subtypes.

Depending on the cytokine environment,different types of Th cells will develop. IFN-gdrives Th1 cell production and generates clone

BOX 1. DETERMINANTS OF ACQUISITION OF TB INFECTION

The closeness and duration of contact with a patient with infectious TB (Rieder et al. 2009).

The effective amount of shared air space determined by the level of air circulation and size of theroom, which influences effective exposure.

Patients with a higher bacterial load, such as those with cavitatory pulmonary TB and those AFB (acid-fast bacillus) smear positive, transmit TB more efficiently.

The susceptibility of the host.

Diagnosis and Management of Latent TB Infection

Cite this article as Cold Spring Harb Perspect Med 2015;5:a017830 3

ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



expansion as well as differentiation into mem-ory cells; on the contrary, IL-10 and IL-4 inhibittheir development. Th1 cells regulate cell-medi-ated immunity and activate natural killer cellsthrough IFN-g secretion. High levels of IL-4,IL-5, and IL-10 are needed for the differentia-tion of Th2 cells, their clonal expansion, andgeneration of memory cells. IFN-g down-regu-lates the development of this cell type. Th2 cellsare responsible for controlling humoral immu-nity and activation of B cells into plasma cells.

Adaptive immunity can take as long as 42 dafter MTC exposure and infection to be estab-lished (Wallgren 1948; Poulsen 1950; Berry et al.2010). At this point, a large number of activatedmacrophages are recruited to the site of infec-tion and construct the granuloma, to whichlymphocytes, epithelioid, and giant cells arealso attracted. By depriving the center of thegranuloma of oxygen, the lowering the pH ne-crosis occurs internally, preventing bacterialreplication but enabling dormancy. However,it is likely that some replication still occurs, be-cause without this, isoniazid would not be ef-fective against LTBI, as it is only bactericidalagainst replicating M. tuberculosis.

Macrophages and dendritic cells also trans-port bacteria to the local lymph nodes duringthe construction of granulomas, allowing thedispersal of infection, particularly to well-oxy-genated areas such as lung apices, suprarenalglands, the brain cortex, and bone and growingjoints.

Disease

Only �10% of infected individuals develop ac-tive disease; of these, most will progress in thefirst 2 years after exposure. There are two possi-ble methods for the development of active dis-ease—an active primary infection, and reactiva-tion of LTBI. Reactivation may occur years afterinfection in older individuals, usually caused bya failure of the host immune response (Box 2).

IMMUNOLOGY AND PATHOGENESIS

The breadth of information available on theimmunology and pathogenesis of LTBI means

that the intention of this section is not to com-prehensively cover all the available material,but rather to provide a summary of the key con-cepts to allow readers to better understand thediagnostic and therapeutic management ofLTBI.

MTC Factors

M. tuberculosis has a series of tools to enablesurvival inside the human host. First, a uniquecell wall structure with high lipid content pro-tects the bacterium from the highly acidic pHof the phagosome. Second, it is able to arrestphagosome–lysosome fusion. Third, it inter-feres in antigen presentation as well as in thenormal function of CD8þ T cells, natural kill-er cells, and the complement attack complex.Fourth, M. tuberculosis can resist host-derivedantimicrobials (reactive nitrogen and oxygenintermediates).

Host Factors

As described above, the human immune re-sponse against the MTC includes both the innatedefense system (macrophage and natural killercells, complement activation, ciliary clearance,and secretions of the main airways) and theadaptive response, comprising both cellularand humoral immunity. T cells are the maincomponent of the cellular response that providesprotection against TB and promotes memoryresponse and granuloma formation. Althoughhistorically B cells and humoral immunitywere not considered to offer any protection,some studies have shown they do contribute toimmunological control (Gupta et al. 2012).

The main benefit of the adaptive response isthe presence of immunological memory, so thata copy of each antigen found is stored in mem-ory T cells and will lead to a rapid and enhancedresponse to subsequent encounters with thatsame antigen. This is the basis for the TST, adelayed-type hypersensitivity reaction causedby cytokines released from previously sensitizedT cells when they are exposed to PPD antigens.Although hypersensitivity is linked to protec-tive immunity, protection is not complete. Dif-

L. Munoz et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



ferent studies have shown that TST-positivereactors are less likely to be reinfected than non-responders, but there is no way to prevent reac-tivation without specific treatment.

From 2 to 4 wk after the first encounter be-tween bacteria of the MTC and macrophages,the human host prepares two types of responses.The first one is delayed-type hypersensitivityto tubercle bacilli, which is tissue damaging asit promotes destruction of nonactivated macro-phages containing proliferating mycobacteri-um. The second is a cell-mediated response,which activates macrophages to kill bacteria.There is a dynamic equilibrium between thesetwo responses, which can lead to either TB pro-gression or containment of MTC.

BURDEN IMPORTANCE

In 1993, the World Health Organization (WHO)declared TB to be a global public health emer-gency, stating that: “Tuberculosis today is hu-manity’s great killer, and it is out of control inmany parts of the world. The disease, prevent-able and treatable, has been grosslyneglected andno country is immune to it.” The first priority forTB elimination is early diagnosis and treatmentof pulmonary TB, to reduce transmission. Sub-sequent steps include screening for clinicallyasymptomatic LTBI in high-risk populations.In addition to effective active TB control, thepost-2015 WHO global TB strategy recognizesthat elimination requires a focus on reducing the

BOX 2. RISK FACTORS

For M. tuberculosis Infection

† Children are more prone to primary infection, as their immune system is not yet completelymature.

† Close contacts of persons known or suspected of having active TB.

† Health-care workers, as well as residents and employees of congregate settings whose clients areat increased risk for active TB (correctional facilities, homeless shelters).

† Being born in areas that have a high incidence of active TB.

For Progression of Infection to Active Tuberculosis (TB)

† HIV infection

† Treatment with anti-TNF agents

† Solid organ transplantation

† Immunosuppressive treatment

† End-stage renal disease

† Hematological malignancies

† Head and neck neoplasm

† Gastrectomy

† Silicosis

† Underweight and malnutrition

† Corticosteroid therapy

† Diabetes mellitus

† Untreated fibrotic lesions suggesting previous TB in chest X-ray

Some patients have no obvious predisposing factors and may just be naturally susceptible to pro-gression.



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



pool of latently infected individuals from whichfuture TB cases would be generated.

It is estimated that one-third of the world’spopulation is latently infected with MTC. How-ever, given the 10% progression rate, clinical andpublic health services must focus their energiesprimarily on groups at a higher risk of disease.These include recently infected persons, immu-nocompromised individuals such as HIV (hu-man immunodeficiency virus) patients, anti-TNF and transplant candidates, and householdcontacts of pulmonary TB cases. Additionally,health-care workers and migrants from high-incidence countries to low-incidence settingsmay be tested and given preventive treatment.

DIAGNOSTIC TESTS

Introduction

Individuals should not be tested for LTBI unlessthe responsible clinician has a managementplan; “intention to test is intention to treat.”The low risk of progression, treatment comple-tion, and effectiveness and the adverse effects oftreatment should be taken into account whentesting is offered.

As previously stated, there is no way to di-rectly diagnose LTBI, as the MTC cannot berecovered from the host unless active TB is pre-sent. An indirect immunological assessment ofexposure is made instead by ascertaining thereactivity of host lymphocytes to mycobacterialantigens, either by testing the in vivo responsewith the TST or in vitro with INF-a release as-says (IGRAs).

TST

The TST consists of an intradermal injection oftuberculin. Although a variety of approacheshave been used, the Mantoux technique, whichis administered on the inner surface of the fore-arm, is one of the most widely used globally.Tuberculin material can be PPD or RT-23. Thestandard doses are 5 PPD units (0.1 mL) or 2RT-23 units, which are equivalent. Both typesof tuberculins contain a complex mixture ofantigens, including those of M. bovis–BCG

strains and several antigens of nontuberculousmycobacteria (NTM), as well as M. tuberculosisantigens. Tuberculin will stimulate a delayed-type hypersensitivity response via T lympho-cytes. A positive TST can be detected by an in-duration on the site of injection after 48–72 h.The result is the transverse diameter of the in-duration, which is usually recorded in millime-ters. The interpretation of the result should takeinto account immunosuppression as well as theprevalence of TB in the setting in question—5 mm is considered positive if the patient isimmunosuppressed, but it will be considerednegative in healthy patients living in high-prev-alence areas, for whom a 15-mm cutoff willapply. In some settings, a uniform cutoff of10 mm is applied. These different and subjec-tive considerations make it difficult to compareTST properties between studies in differentcountries and populations.

The main limitation of TST is its low spe-cificity because of previous BCG vaccinationand NTM infections. There are also other draw-backs in this apparently simple test. In subjectswith chronic debilitating conditions or immu-nosuppression, TSTmay have a lower sensitivitythan in healthy people. It must be administratedand read by trained professionals so as to avoidvariability of the results. It also requires twoclinical visits, and a positive reaction may inter-fere with confidentiality issues.

IGRAs

These tests use a blood sample taken from thepatient to measure T-cell release of IFN-g invitro after stimulation by specific MTC antigens.Two IGRAs have been developed: the Quanti-FERON-TBw Gold In-Tube (QFT) (CellestisLimited, Carnegie, Victoria, Australia) and theT-SPOT.TBw (Oxford Immunotec, Oxford,UK). Both tests use the early secreted antigenictarget 6 (ESAT-6) and culture filtrate protein 10(CFP10), which are encoded by genes located onthe region of difference 1 (RD-1) segment of theMTC genome. QFT also includes a third anti-gen, TB7.7 (RD-11).

QFT is an enzyme-linked immunosorbent-assay (ELISA)-based test performed on whole

L. Munoz et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



blood. It measures levels of IFN-g in the super-natant of a cell suspension. The result is easilyprocessed by software and is reported in inter-national units per mL (IU/mL). T-SPOT.TB isan enzyme-linked immunospot assay (ELI-SPOT) performed on separated T lymphocytes.The result is reported as number of IFN-g pro-ducing T cells.

The main advantage of IGRAs is their spe-cificity, as ESAT-6, CFP10, and TB7.7 are foundin the MTC genome, but not in M. bovis–BCG,or in the majority of NTMs. Moreover, IGRAscan differentiate a negative result from anergyby means of a positive control, which uses phy-tohemagglutinin to stimulate production ofIFN-g. In this review, we will talk about IGRAswithout distinction, assuming that both QFTand T-SPOT.TB properties are comparable.

Neither TST nor IGRAs are able to detectTB infection in its first stages. Because they mea-sure the adaptive immune response to MTC, an�8-wk period after infection is required for areliable result.

There are three notable disadvantagesshared by both types of immunologic tests.The first one is their poor positive predictivevalue (PPV) for progression to active TB. Severalstudies have tried to correlate the risk of pro-gression to the quantitative response of both invivo tests (TST measurement) and in vitro tests(i.e., IFN-g concentration in QFT tubes), butno conclusive results have been found. A recentmeta-analysis showed a slightly better PPV withIGRAs compared with TST (2.4% for TSTand 6.8% for IGRAs in high-risk groups) forthe development of active disease (Diel et al.2012), i.e., in a group of 100 IGRA-positive sub-jects tested for LTBI, and left without preemptivetreatment, seven individuals would develop TB.However, to date, it is not possible to identifythese seven before they progress to active disease.Furthermore, another review concluded thatthere is little difference between TST and IGRAPPVs (Rangaka et al. 2012), contributing to theWHO deendorsement of the use of IGRAs indeveloping countries.

The second common disadvantage is theinability of TSTs and IGRAs to distinguish be-tween LTBI and active disease. The WHO has

recommended against the use of IGRAs for ac-tive TB diagnosis, as a negative result does notrule out disease.

The third issue is that the tests are unable todistinguish recent from remote infection. Theresponse profiles for past and recent infectionare not different from each other. Each resultmust be carefully examined and tailored, as pe-culiarities of the host may modify sensitivityand specificity of both TSTand IGRAs (Abuba-kar et al. 2013).

Evidence of Diagnostic Accuracy

Sensitivity

Because there is no “gold standard” for LTBIdiagnosis, it is not possible to properly measurethe sensitivity and specificity of the availablediagnostic tests. The only individuals with prov-en TB infection are active TB patients, fromwhom bacteria of the MTC can be isolated.Data from a recent meta-analysis of diagnostictests in TB patients showed a global sensitivity of0.70 (TST), 0.81 (QFT), and 0.88 (T-SPOT.TB)(Diel et al. 2010). The immune response of la-tently infected and active TB patients is likely tonot be equivalent, as TB itself is a well-knowncause of anergy. Thus, sensitivity may be evenhigher in healthy people. In this population, anegative TST or IGRA rules out TB infectionand has an almost 100% negative predictive val-ue (NPV) for progression to active TB. A nega-tive result, however, may not be reliable in otherclinical scenarios, such as clinical suspicion ofactive TB or immunosuppressed patients. Sev-eral studies have tested how IGRAs perform inthese specific circumstances; IGRAs seem to beless affected by immunosuppression than theTST (Redelman-Sidi and Sepkowitz 2013).

Specificity

Specificity is the main difference between theTSTand IGRAs. Tuberculin contains several an-tigens shared by the MTC, M. bovis–BCG, andNTMs; thus, a positive TST does not necessarilyequate to TB infection, especially in BCG-vac-cinated subjects, among whom a higher propor-tion of false-positive results can be found. On



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



the contrary, IGRAs have excellent specificity asthey use antigens encoded in specific sequencesof the MTC genome. Pooled specificity datashow a range of 0.86–0.99 for IGRAs versus0.97 and 0.59 for TST in BCG-vaccinated andnon-BCG-vaccinated individuals, respectively(Diel et al. 2010).

Improved specificity prevents the excessiveuse of treatment to prevent progression, as false-positive results will be excluded. If a group of1000 healthy recent contacts of pulmonary TBcases were tested by TST, �50% would test pos-itive and be treated. If an IGRAwas used instead,only 30% would be treated, resulting in largesavings by preventing unnecessary treatment.

Other Issues: Two Steps versus One Stepand Cost-Effectiveness

Data on the cost-effectiveness of the differentapproaches to LTBI diagnosis are limited be-cause of the paucity of information on predic-tive values. Individually, it is generally more ex-pensive to test with an IGRA than with a TST. Inthe community, the better the selection of peo-ple at risk of TB for testing, the less unnecessarytreatment and fewer adverse events, as well asreduced expenditure on testing and follow-upvisits.

Guidance on diagnostic strategies for LTBIin individuals at risk varies considerably be-tween countries. Although there are several dis-crepancies between guidelines in terms of theTST cutoff for positivity, the repetition of TST(looking for a boosting effect), and the use ofone or more types of tests, there is growing ev-idence on the superiority of IGRAs over the TSTin terms of sensitivity and specificity. Giventheir higher specificity and their at least compa-rable NPV for reactivation of TB, IGRAs may bepreferable to TST for targeting people for pre-ventive treatment. There are very few countries(United States, Spain, and Portugal) where thetwo-step TST is heeded. This strategy aims toobtain a LTBI diagnosis in older and immuno-suppressed patients, in which an initial TSTmayboost the immune system against an older latentinfection that would be evident after the seconddose of tuberculin. However, the effectiveness of

this second test is usually low, and several stud-ies have reported a nonnegligible proportion offalse-positive results, mainly in BCG-vaccinatedindividuals.

A more widely used two-step approach,usually based on cost-effectiveness analysis, isthe application of IGRAs to confirm an initialTST. The rationale for this is that the cheaperand potentially equally sensitive test (TST) isapplied first, and the diagnosis is confirmedwith an IGRA test. Which strategy (IGRAs inplace of TST or as a confirmatory test) is moresuitable for testing individuals at risk is un-clear, but it will depend on the immune statusof the host and the expected prevalence of TBinfection in the population tested (Table 1). Inimmunocompetent individuals, in whom thesensitivity of the two tests can be regarded asroughly comparable, positive IGRA results areexpected to also be TST-positive. In this sce-nario, using IGRAs as a confirmatory test maybe appropriate if the expected proportion ofinfection is low. Conversely, if the expected pro-portion of infection is high, testing with anIGRA instead of TST may be more cost-saving.

In immunocompromised patients, the useof IGRA plus TST to increase sensitivity mayavoid missing some infected subjects at a highrisk of progression to active disease. However,the results from several studies in patients aboutto start biologic therapeutics suggest that screen-ing for LTBI and decisions regarding preventivetreatment can be reliably based on IGRAs, irre-spective of TST results (Laffitte et al. 2009;Chang et al. 2011; Qumseya et al. 2011; Garco-vich et al. 2012; Hsia et al. 2012, 2013; Jung etal. 2012; Mariette et al. 2012; Greveson et al.2013). Furthermore, the additional cost of un-dertaking two tests for the limited gain in sensi-tivity likely means that dual testing may not becost-effective.

TREATMENT

Drugs—Trial Evidence of Effectiveness

The treatment of LTBI has a long history, start-ing from the early studies of the 1950s throughto modern trials of novel combination regimens

L. Munoz et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



in high-risk populations. Drugs for treating la-tent infection should possess sterilizing activityagainst dormant mycobacteria, as such is thestatus of bacteria in these patients. As previouslydescribed, the very fact that isoniazid is activeagainst LTBI suggests that at least some replica-tion is occurring, as the target of this drug’sactivity is the inhibition of the synthesis of my-colic acid, required for the mycobacterial cellwall, during replication. Numerous randomizedcontrolled trials and a handful of systematic re-views have shown the safety and efficacy of iso-niazid in the general population, HIV-infectedindividuals, and posttransplant patients for pre-venting TB reactivation.

Rifampicin, the second proven active drugfor LTBI, is a well-known antibiotic with intra-cellular activity that inhibits the bacterial DNA-dependent RNA polymerase and has been used

against other bacterial infections in their latencyperiod.

Globally, it is most common to use 6–9 moof daily isoniazid monotherapy (dose 300 mg)although recent data suggest that 4 mo of dailyrifampicin (dose 600 mg), or 3 mo of combinedtherapy with isoniazid 300 mg and rifampicin600 mg may be equally effective (Stagg et al.2014). A novel regimen consisting of 12 dosesof a weekly regimen of rifapentin (900 mg)and high-dose isoniazid (900 mg) has been re-cently described as effective and is currently usedin the United States under directly observedtherapy.

The first and major reason for treatmentfailure, and thus for the development of activeTB, is poor medication adherence. Unfortu-nately, the direct observation of each patient isunaffordable for most countries. LTBI treat-

Table 1. Differences between TST and IGRAs according to immune status

Healthy subjects

Immunosuppressed subjects (including

HIV, IMID patients attributable to starting

anti-TNF therapy, patients awaiting for

transplantation)

First step Always rule out active TB(Clinical history and examination, chest X-ray, and sputum microscopy and

culture as well as other tests if extrapulmonary TB is suspected)

When to test? Recent infection (contact tracingstudy)

Occupational risk (HCW)

At HIV diagnosisBefore the beginning of IS treatment

How do TST perform? Good sensitivity Less sensitivity because of weakeningcondition

Poor specificity: high proportion offalse-positive reactions in BCG-vaccinated subjects

Poor specificity: high proportion of false-positive reactions in BCG-vaccinatedsubjects

Window period (recent infection) No window period needed

How do IGRAs perform? Good sensitivity Less sensitivity than in healthy subjects, butbetter sensitivity than TST

Good specificity Good specificityWindow period (recent infection) No window period needed

Recommended strategiesfor LTBI diagnosis andtargeting patients forpreventive treatment

High TB infection prevalenceexpected: IGRA

Most guidelines recommend TST ANDIGRA to maximize sensitivity; severalstudies used ONLY IGRAs with no issuesLow TB infection prevalence

expected: two-step strategyPositive TST followed by an IGRA

TST, tuberculin skin test; IGRA, interferon-g release assay; HIV, human immunodeficiency virus; IMID, immune-mediated

inflammatory disease; TNF, tumor necrosis factor; HCW, health-care workers; IS, immunosuppressive; BCG, Bacillus

Calmette–Guerin.



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



ments are lengthy, and a close monitoring ofsubjects is required to ensure high treatmentcompliance. Several methods for checking treat-ment compliance have been validated. They in-clude the Eidus-Hamilton test and commer-cially available reactive strips that can easilydetect urine metabolites of isoniazid metabo-lites. Checking the orange-colored urine of sub-jects taking rifampicin may also be useful foradherence control.

Adverse Events

The three common treatment regimens are gen-erally well tolerated. Patients must always betold about symptoms that may indicate the de-velopment of adverse effects so that they seekcare promptly. The most common life-threat-ening adverse event resulting from isoniazid isliver toxicity. Not only is a clinical assessmentneeded, but liver function tests are also essentialin the third or fourth week of treatment, andthen in the second and third month, as well as inthe instance of patients reporting hepatitis signsand symptoms. Biochemistry variations on al-anine transaminase (ALT) and aspartate trans-aminase (AST) precede symptoms as indicatorsof liver damage and that is the reason for testingapparently healthy patients. Another frequentadverse event of isoniazid is peripheral neurop-athy, which can easily be prevented with dailypyridoxine.

Patients taking rifamycins can present flu-like symptoms in their first week of treatment.This adverse event is more common in inter-mittent regimens (i.e., weekly doses).

Hypersensitivity reactions are more fre-quently related to rifamycins, although theymight also be caused by isoniazid. These reac-tions are mostly seen in the second to fourthweek of treatment and consist of a rash and fever,ranging from harmless itching of the skin tosevere clinical pictures including Stevens–John-son syndrome.

Special Issues

Although preventing drug-resistant TB cases isof great concern for clinicians and health au-

thorities, there are no recommended regimensfor LTBI treatment of those exposed to multi-drug-resistant TB, as there are no sterilizingdrugs for dormant bacilli apart from isoniazidand the rifamycins. Guidelines advise only“watchful waiting” of these individuals to assurean early diagnosis in case of the developmentof active TB. Young children, who are house-hold contacts of multidrug-resistant TB cases,may benefit from prophylactic treatment. Thisshould always be supervised by a physician withknowledge and experience of multidrug-resis-tant TB prophylactic treatment, based on drugsusceptibility results of the index case.

For all LTBI patients diagnosed within acontact study, it is mandatory for the clinicianto check the drug susceptibility of the index caseand to ensure that the transmitted bacteria arelikely to be sensitive to the chosen LTBI treat-ment regimen.

MANAGEMENT

The most important issue for LTBI manage-ment is to properly target candidates for screen-ing and treating, that is, to identify an individ-ual’s risk of developing active TB. The maincontraindication for LTBI treatment is activeTB, and thus the first commitment of the clini-cian is to rule out the disease through a clinicalinterview for the specific symptoms of TB and acareful examination of a recent chest X-ray.Medical and nursing staff skilled in treatingTB patients and dealing with adverse events ofantituberculosis drugs should ideally manageLTBI screening.

Individuals at Risk

Newly Infected Individuals

As mentioned above, the first 2 yr after infectioncarry the highest risk for developing active dis-ease. For this reason, individuals at risk forMTC infection found to have converted froma negative to a positive result for a LTBI diag-nostic test, as well as individuals testing positiveafter a contact tracing study, will be the mostlikely to benefit from preventive treatment.

L. Munoz et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



Contact Tracing (Erkens et al. 2010). Afterdiagnosing an individual with pulmonary TB,it is mandatory to get information regardingtransmission settings to assign priorities forthe investigation. The purpose of conducting acontact examination of a TB case is to evaluateindividuals for active TB disease (to find eitherthe source case or secondary cases) so that treat-ment can be given and further transmissionstopped, as well as diagnosing and treatingthose with LTBI.

Priorities for contact investigation dependon the characteristics of the index case (cavita-tion on chest X-ray and positive AFB smearsputum are the most infectious), susceptibilityand vulnerability of contacts (children under5 yr and immunocompromised patients are as-signed the highest priority), and circumstancesof the exposure (household contacts, congre-gate settings).

All referred contacts should be specificallyasked for respiratory symptoms and currentgeneral health. Then, a TST or an IGRA (eitheralone or in a two-step strategy) should be un-dertaken. Patients with negative results ob-tained before 8 wk after the end of exposureshould have the test repeated so as to avoidfalse-negative results in the window period.

Subjects reporting any suggestive symptomof TB and all individuals with positive results onthe LTBI diagnostic tests should have a chest X-ray to rule out active TB. After TB is excluded, allLTBI patients should be offered treatment witheither isoniazid (6 mo), rifampicin (4 mo), or acombined 3-mo regimen with daily isoniazidand rifampicin or weekly isoniazid and rifapen-tine, with close follow-up. This includes adher-ence, asking for symptoms attributable to adverseevents, and liver function tests, with at least onetest before the beginning of treatment and an-other in 3–4 wk. Further tests may be needed.

Household contacts who are children (orthose with a close relationship with the TB pa-tient) should have TB ruled out with a chestX-ray. They should also be started on preven-tive treatment even though a negative resultmay be found in the first LTBI test (TST or anIGRA). Children, especially those under 5 yr ofage, have an immature immune system that fa-

vors both false-negative results and the develop-ment of active TB. If active TB has been exclud-ed and a second LTBI diagnostic test remainsnegative after the window period, treatmentcan be stopped.

When a child is found to have TB disease, acontact study should be performed aroundthem, because although TB in children is rarelyinfectious, it is usually a sign of a close contactwho is the source case.

Health-Care Workers. An IGRA or TST maybe used for persons with occupational exposureto the MTC. If TB infection is systematicallytested for and ruled out when the individualcommences their new job, and then every 2 yras long as they remain negative responders, aconversion from negative to positive resultmay be detected and thus treatment for a recentinfection could be offered to prevent TB devel-opment.

Immunocompromised Patients

Immunocompromised persons with LTBI are atan increased risk for progression to active TB;treatment of LTBI in this population can reducethis risk.

People Living with HIV. Patients coinfectedwith HIV and MTC are particularly prone to areactivation of LTBI and the development ofdisseminated disease. For this reason, system-atic screening for TB infection in individualsnewly diagnosed with HIV has been recom-mended. After ruling out active disease, patientsshould be tested with either a TST or an IGRA,and those with positive results should receivepreventive treatment. Several studies haveshown the better performance of IGRAs overTSTs in patients with lower CD4 cells counts(,200 cells/mm3). These patients should betreated with at least 6 mo of isoniazid 300 mgdaily.

Subjects Considered for Anti-TNF Drugs. In-dividuals diagnosed with immune-mediatedinflammatory diseases have an impaired im-mune system attributable to their disease or itstreatment. Those receiving anti-TNF (especiallyinfliximab and adalimumab) are at particularrisk for reactivation, as these drugs break up



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



granulomas, which are partially maintained byself-released TNF.

One strategy proposed in candidates to re-ceive anti-TNF is systematically screening forLTBI before treatment begins. How to test thispopulation is still an area of debate in differentcountries. To ensure the greatest sensitivity, sev-eral nations have encouraged the use of bothTST and an IGRA and the treatment of any pa-tient with a positive result.

Transplant Candidates. Among solid organtransplant recipients, the risk of developing ac-tive TB has been calculated to be 36- to 74-foldhigher than in the general population (Singhand Paterson 1998), with lung transplantationbeing the most risky. Active screening of LTBI isincreasingly becoming part of the routine in thecheck-up of transplant candidates before thesurgical procedure.

Patients with end-stage kidney disease andthose with advanced liver cirrhosis waiting forkidney and liver transplantation deserve specialcomment. These two conditions are risk factorsfor skin test anergy and thus an increased rate offalse-negative results in TST. For this reason, anIGRA is preferred to determine which patientsshould receive LTBI treatment.

Migrants from Endemic Areas Arrivingto High-Income Countries

As the majority of TB in migrants arises throughthe reactivation of infections acquired overseasand developed in the first few years postentry,screening for LTBI in this population on arrivalis an essential element of any TB eliminationstrategy. This group may be a hard-to-reachpopulation, with low rates of returning to haveTSTs read; thus, an IGRA test might be pre-ferred. The use of IGRAs can increase test com-pletion rates, so control efforts should focus onthose most likely to benefit from further evalu-ation and treatment.

LOOKING TO THE FUTURE

The control and eventual global elimination ofTB would depend on either an effective post-exposure vaccine or better tools and drugs to

detect and treat LTBI. Current research focusedon improving the diagnosis and positive pre-dictive value of LTBI tests ranges from improve-ments of IGRAs by investigating additionalantigens as well as alternative cytokines, to re-search based on proteomics, transcriptomics,and metabolomics. For example, several ongo-ing studies in humans are measuring differentcytokine profiles (e.g., IFN-g and IL2) to differ-entiate active TB and LTBI in vitro, whereasothers have focused on transcriptional profilesapparently associated with latently infected in-dividuals who might progress to disease. It islikely that advances in our understanding of la-tency in TB would be revolutionized in the nextdecade based on ongoing research in large-pop-ulation cohorts using these approaches.

New research must also identify treatmentregimens that are considerably shorter, moreeffective, and less toxic. The new rifapentineregimen of only 12 doses has been made possi-ble because of the long half-life of this drug.Future research must consider whether newerdrugs with a long half-life, such as bedaquiline,may be used to shorten treatment. Such researchmust also investigate the adverse effect profile ofany new regimen.

REFERENCES

Abubakar I, Stagg HR, Whitworth H, Lalvani A. 2013. Howshould I interpret an interferon g release assay result fortuberculosis infection? Thorax 68: 298–301.

Berry MP, Graham CM, McNab FW, Xu Z, Bloch SA, Oni T,Wilkinson KA, Banchereau R, Skinner J, Wilkinson RJ,et al. 2010. An interferon-inducible neutrophil-drivenblood transcriptional signature in human tuberculosis.Nature 466: 973–977.

Calmette A. 1931. Preventive vaccination against tuberculo-sis with BCG. Proc R Soc Med 24: 1481–1490.

Chang B, Park HY, Jeon K, Ahn JK, Cha HS, Koh EM, KangES, Koh WJ. 2011. Interferon-g release assay in the diag-nosis of latent tuberculosis infection in arthritis patientstreated with tumor necrosis factor antagonists in Korea.Clin Rheumatol 30: 1535–1541.

Diel R, Loddenkemper R, Nienhaus A. 2010. Evidence-based comparison of commercial interferon-g release as-says for detecting active TB. A meta analysis. Chest 137:952–968.

Diel R, Loddenkemper R, Nienhaus A. 2012. Predictive val-ue of interferon-g release assays and tuberculin skin test-ing for progression from latent TB infection to diseasestate: A meta-analysis. Chest 142: 63–75.

L. Munoz et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



Erkens CGM, Kamphorst M, Abubakar I, Bothamley GH,Chemtob D, Haase W, Migliori GB, Rieder HL, ZellwegerJP, Lange C. 2010. Tuberculosis contact investigation inlow prevalence countries: A European consensus. EurRespir J 36: 925–949.

Garcovich S, Ruggeri A, D’Agostino M, Ardito F, De SimoneC, Delogu G, Fadda G. 2012. Clinical applicability ofQuantiferon-TB-Gold testing in psoriasis patients duringlong-term anti-TNF-a treatment: A prospective, obser-vational study. J Eur Acad Dermatol Venereol 26: 1572–1576.

Gengenbacher M, Kaufmann SH. 2012. Mycobacterium tu-berculosis: Success through dormancy. FEMS MicrobiolRev 36: 514–532.

Greveson K, Goodhand J, Capocci S, Woodward S, MurrayC, Cropley I, Hamilton M, Lipman M. 2013. Yield andcost effectiveness of mycobacterial infection detectionusing a simple IGRA-based protocol in UK subjectswith inflammatory bowel disease suitable for anti-TNFa therapy. J Crohn’s Colitis 7: 412–418.

Gupta A, Kaul A, Tsolaki AG, Kishore U, Bhakta S. 2012.Mycobacterium tuberculosis: Immune evasion, latencyand reactivation. Immunobiology 217: 363–374.

Hillier A. 1903. The nature of the infectivity of phthisis: Astudy of the views of Koch, Flugge, and others. Br Med J 1:593–596.

Hsia EC, Schluger N, Cush JJ, Chaisson RE, Matteson EL,Xu S, Beutler A, Doyle MK, Hsu B, Rahman MU. 2012.Interferon-g release assay versus tuberculin skin test priorto treatment with golimumab, a human anti-tumor ne-crosis factor antibody, in patients with rheumatoid ar-thritis, psoriatic arthritis, or ankylosing spondylitis. Ar-thritis Rheum 64: 2068–2077.

Hsia EC, Cush JJ, Matteson EL, Beutler A, Doyle MK, Hsu B,Xu S, Rahman MU. 2013. Comprehensive tuberculosisscreening program in patients with inflammatory arthrit-ides treated with golimumab, a human anti-tumor ne-crosis factor antibody, in Phase III clinical trials. ArthritisCare Res (Hoboken) 65: 309–313.

Jung YJ, Lyu J, Yoo B, Lee CK, Kim YG, Yang SK, Byeon JS,Kim KJ, Ye BD, Lee KH, et al. 2012. Combined use of aTST and the T-SPOTw.TB assay for latent tuberculosisinfection diagnosis before anti-TNF-a treatment. Int JTuberc Lung Dis 16: 1300–1306.

Koch R. 1891. A further communication on a remedy fortuberculosis. Br Med J 1: 125–127.

Laffitte E, Janssens JP, Roux-Lombard P, Thielen AM, BardeC, Marazza G, Panizzon RG, Saurat JH. 2009. Tubercu-

losis screening in patients with psoriasis before anti-tumour necrosis factor therapy: Comparison of aninterferon-g release assay vs. tuberculin skin test. Br JDermatol 161: 797–800.

Mariette X, Baron G, Tubach F, Liote F, Combe B, Miceli-Richard C, Flipo RM, Goupille P, Allez M, Salmon D,et al. 2012. Influence of replacing tuberculin skin testwith ex vivo interferon g release assays on decision toadminister prophylactic antituberculosis antibiotics be-fore anti-TNF therapy. Ann Rheum Dis 71: 1783–1790.

Martin CF, Robins GD. 1898. On the diagnostic value oftuberculin. Br Med J 1: 357–359.

Mi’fadyean J. 1891. Experiments with tuberculin on cattle.Br Med J 1: 634–635.

Poulsen A. 1950. Some clinical features of tuberculosis. ActaTuberc Scand 24: 311–346.

Qumseya BJ, Ananthakrishnan AN, Skaros S, Bonner M,Issa M, Zadvornova Y, Naik A, Perera L, Binion DG.2011. QuantiFERON-TB Gold testing for tuberculosisscreening in an inflammatory bowel disease cohort inthe United States. Inflamm Bowel Dis 17: 77–83.

Rangaka MX, Wilkinson KA, Glynn JR, Ling D, Menzies D,Mwansa-Kambafwile J, Fielding K, Wilkinson RJ, Pai M.2012. Predictive value of interferon-g release assays forincident active tuberculosis: A systematic review andmeta-analysis. Lancet Infect Dis 12: 45–55.

Redelman-Sidi G, Sepkowitz KA. 2013. IFN-g release assaysin the diagnosis of latent tuberculosis infection amongimmunocompromised adults. Am J Respir Crit Care Med188: 422–431.

Rieder H, Chen-Yuan C, Gie RP, Enarson D. 2009. Generalbackground to clinical tuberculosis. Infection and dis-ease. Crofton’s clinical tuberculosis, 3rd ed., pp. 1–10.Macmillan, Oxford.

Singh N, Paterson D. 1998. Mycobacterium tuberculosis in-fection in solid-organ transplant recipients: Impact andimplications for management. Clin Infect Dis 27: 1266–1277.

Stagg HR, Zenner D, Harris RJ, Munoz L, Lipman MC,Abubakar I. 2014. Treatment of latent tuberculosis: Anetwork meta-analysis. Ann Intern Med (in press).

Wallgren A. 1948. The time-table of tuberculosis. Tubercle29: 245–251.

Wang L, Turner MO, Elwood RK, Schulzer M, FitzGeraldJM. 2002. A meta-analysis of the effect of Bacille CalmetteGuerin vaccination on tuberculin skin test measure-ments. Thorax 57: 804–809.



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



June 8, 20152015; doi: 10.1101/cshperspect.a017830 originally published onlineCold Spring Harb Perspect Med

Laura Muñoz, Helen R. Stagg and Ibrahim Abubakar Diagnosis and Management of Latent Tuberculosis Infection

Subject Collection Tuberculosis

of TuberculosisTransmission and Institutional Infection Control

Edward A. Nardell

Clinical Aspects of Adult Tuberculosis

ZumlaRobert Loddenkemper, Marc Lipman and Alimuddin

InfectionMycobacterium tuberculosisto Innate and Adaptive Cellular Immune Responses

Katrin D. Mayer-Barber and Daniel L. Barber

Advances in Diagnostic Assays for TuberculosisStephen D. Lawn

and Noncommunicable DiseasesTuberculosis Comorbidity with Communicable

Matthew Bates, Ben J. Marais and Alimuddin ZumlaTuberculosis InfectionDiagnosis and Management of Latent

AbubakarLaura Muñoz, Helen R. Stagg and Ibrahim

Host-Directed Therapies for TuberculosisDavid M. Tobin

Mycobacterial Growth

al.Iria Uhía, Kerstin J. Williams, Vahid Shahrezaei, et

Tuberculous GranulomaImmunity and Immunopathology in the

Antonio J. Pagán and Lalita RamakrishnanDrug-Resistant TuberculosisMultidrug-Resistant Tuberculosis and Extensively

Michael L. RichKwonjune J. Seung, Salmaan Keshavjee and

Evolution of the Mechanism-Based Paradigm?Tuberculosis Drug Development: History and

Sumit Chakraborty and Kyu Y. RheeArabinogalactan

Peptidoglycan and−−The Mycobacterial Cell Wall

Lloyd, et al.Luke J. Alderwick, James Harrison, Georgina S.

Drug DevelopmentGenetic Approaches to Facilitate Antibacterial

Dirk Schnappinger

Tuberculosis and HIV Coinfection

Gunilla KälleniusJudith Bruchfeld, Margarida Correia-Neves and

TargetsDevelopment Pipeline and Emerging Drug The Tuberculosis Drug Discovery and

Khisimuzi Mdluli, Takushi Kaneko and Anna Upton

Imaging in Tuberculosis

Gulati, et al.Jamshed B. Bomanji, Narainder Gupta, Parveen

http://perspectivesinmedicine.cshlp.org/cgi/collection/ For additional articles in this collection, see

Copyright © 2015 Cold Spring Harbor Laboratory Press; all rights reserved


http://perspectivesinmedicine.cshlp.org/cgi/collection/

http://perspectivesinmedicine.cshlp.org/cgi/collection/


Documents

Diagnosis and Management of Latent Tuberculosis Infectionperspectivesinmedicine.cshlp.org/content/5/11/a017830... · 2015-10-26 · Diagnosis and Management of Latent Tuberculosis