Background: Migrant farmworkers are among the highest-risk populations for latent TB infection (LTBI) in the United States with numerous barriers to healthcare access and increased vulnerability to infectious diseases. LTBI is usually diagnosed on the border using the tuberculin skin test (TST). QuantiFERON-TB Gold In-Tube (QFT-GIT) also measures immune response against specific Mycobacterium tuberculosis antigens. The objective of this study is to assess the comparability of TST and QFT-GIT to detect LTBI among migrant farmworkers on the border, as well as to examine the effects of various demographic and clinical factors on test positivity. Methods: Participants were recruited using mobile clinics on the San Luis US-Mexico border and tested with QFT-GIT and TST. Demographic profiles and clinical histories were collected. Kappa coefficients assessed agreement between TST and QFT-GIT using various assay cutoffs. Logistic regression examined factors associated with positive TST or QFT-GIT results. Results: Of 109 participants, 59 of 108 (55 %) were either TST (24/71, 34 %) or QFT-GIT (52/106, 50 %) positive. Concordance between TST and QFT-GIT was fair (71 % agreement,kappa= 0.38, 95 % CI: 0.15, 0.61). Factors associated with LTBI positivity included smoking (OR = 1.26, 95 % CI-1.01-1.58) and diabetes/high blood sugar (OR = 0.70, 95 % CI = 0.51-0.98). Discussion: Test concordance between the two tests was fair, with numerous discordant results observed. Greater proportion of positives detected using QFT-GIT may help avoid LTBI under-diagnosis. Assessment of LTBI status on the border provides evidence whether QFT-GIT should replace the TST in routine practice, as well as identifies risk factors for LTBI among migrant populations.

Gegenstand dieser Arbeit ist die Untersuchung zellulärer Testsysteme bei Patienten mit Verdacht auf latente Infektion mit M. tuberculosis. Eine latente Tuberkulose kann unter Immunsuppression zu einer aktiven Tuberkulose werden. Deshalb wird bei immunsuppressiven Therapien insbesondere mit TNF-alpha-Blockern eine Chemoprävention empfohlen. Daher ist es sehr wichtig, latente Infektionen zu erkennen. Ist ein Patient mit M. tuberculosis infiziert, reagieren seine T-Zellen auf Stimulation mit Antigenen wie ESAT-6 und CFP-10. Diese Immunantwort ist die Grundlage der modernen IFN-γ release Assays (IGRA). ESAT-6 und CFP-10 fehlen bei allen BCG-Stämmen und bei den meisten nicht-tuberkulösen Mykobakterien mit Ausnahme von M. kansasii, M. szulgai und M. marinum (Andersen et al. 2000; Behr et al, 1999; Lalvani 2003). Im Gegensatz dazu haben Personen, die mit M. tuberculosis-Komplex-Organismen infiziert sind, in der Regel T-Zellen im Blut, die diese und andere mykobakterielle Antigene erkennen. Ziel der vorliegenden Arbeit war es, die Nutzung von IGRA in der medizinischen Labordiagnostik dahingehend zu analysieren, ob es Unterschiede zwischen ELISPOT-basierten Tests und Röhrchen-Tests als Testformat gibt, inwieweit diese Tests im klinischen Alltag verlässlich sind und ob sich mit einem anderen Auswertealgorithmus eine sicherere Aussage zum Vorliegen einer latenten Tuberkulose treffen lässt. Im Einzelnen wurden dabei drei Ansätze verfolgt: 1. In einer retrospektiven Studie wurden die Ergebnisse von 2686 Patienten ausgewertet, die im Labor des Instituts für Klinische Immunologie des Universitätsklinikums Leipzig von 2013 bis 2016 als Routineuntersuchungen erhoben wurden. Bei klinisch unplausiblen Ergebnissen wurden bei einem Teil der Patienten eine Wiederholungsuntersuchungen durchgeführt. Die analytische Sensitivität und Spezifität sowie den positiven und negativen prädiktiven Wert konnten wir nur unter der Annahme abschätzen, dass der Ausfall in der Wiederholungsuntersuchung einen Hinweis auf falsch- oder richtig-positive oder -negative Werte zulässt. Wir kommen damit zu einer Sensitivität von nur 28 %, einer Spezifität von immerhin 91 %, einem positiven prädiktiven Wert von 32 % und einem negativen prädiktiven Wert von 90 %. Damit sind 68 % der positiven Werte falsch positiv und 10 % der negativen Werte falsch negativ. Unsere Untersuchungen zur Wiederholbarkeit der ELISPOT-Tests im eigenen Labor bestätigen, dass negative Ergebnisse meist wiederholbar sind, positive Werte jedoch skeptisch betrachtet werden müssen. 2. Seit 2012 sendet Instand e.V. zweimal jährlich Ringversuchsproben für den IGRA aus (Ringversuch 650, https://www.instand-ev.de/). Wir analysierten, wie viele Labors bei Teilnahme an der externen Qualitätssicherung (sogenannten Ringversuchen) für IGRA ein korrektes Ergebnis erzielt hatten und ob Unterschiede zwischen ELISPOT-Assay und Röhrchen-Test bestehen. Im Ringversuch waren die Ergebnisse von ELISPOT (z.B. TB-Spot, Oxford Immunotec) und Röhrchentest (Quantiferon Gold bzw. Gold-Plus, Qiagen oder Diasorin) bis auf den 2. Ringversuch 2019 vergleichbar und unterschieden sich nicht. 3. Obwohl die meisten Labore an den Ringversuchen erfolgreich teilnehmen, ist die recht häufige Anzahl vor allem falsch positiver Ergebnisse im diagnostischen Alltag problematisch. Wir haben daher in einem dritten Untersuchungsschritt die Validierungsdaten eines Labors detailliert untersucht, um ein definiertes Verfahren zur Ermittlung positiver Testergebnisse vorzuschlagen. Zwischen 2011 und 2013 erfolgte im Labor Ettlingen eine umfangreiche Qualitätssicherung des ELISPOT unter Nutzung von 70 Proben in Doppelbestimmung. Dabei wurden jeweils die Messungen für die Positiv- und die Negativkontrolle sowie die Werte nach Stimulation mit ESAT-6 und CFP-10 analysiert. Um relevante Unterschiede zwischen Negativkontrolle und der eigentlichen Messung zu bewerten, wurden die Unterschiede mit der Wiederholpräzision in Beziehung gesetzt: Die Unterschiede sollten größer sein als die daraus abgeleitete Ungenauigkeit (Impräzision) der Messungen. Daraus wurde ein Cut-off-Wert kalkuliert, der unmittelbar auf den Daten des Labors beruht. Hierzu ist die Homogenität der Standardabweichungen über alle Proben erforderlich. Sie wird durch die Wurzeltransformation aller Daten erreicht. Für die verwendeten Daten ist sie anwendbar (Altman 1991, Bland 2000). Dies bedeutete bei den untersuchten Daten, dass bei Doppelbestimmungen die Unterschiede zwischen dem Testergebnis immer um den Faktor 0,76 größer sein muss als die Negativkontrolle. Dazu sind allerdings zwei Voraussetzungen zu erfüllen: 1. die Werte müssen vor Analyse wurzeltransformiert werden. 2. benötigt werden mindestens Doppelbestimmungen. Da keine Doppelbestimmungen vorliegen, konnte das Verfahren nicht an einem eigenen Datensatz verifiziert werden. Wir schlussfolgern daraus zusammenfassend: 1. in der Praxis gibt es keine wesentlichen Unterschiede zwischen ELISPOT und Röhrchen-Test als Testformat, was Sensitivität und Spezifität anbelangt 2. IGRAs sind ein verlässliches Werkzeug im klinischen Alltag, insbesondere wenn es um den Ausschluss einer latenten Tuberkulose geht 3. die Auswertung von ELISPOT-Daten lässt sich über Mehrfachbestimmung und quadratwurzeltransformierte Auswertung optimieren (wobei das noch in einer Folgearbeit zu beweisen ist).:Inhalt Abkürzungsverzeichnis 3 1 EINFÜHRUNG 5 1.1 Epidemiologie der Tuberkulose 5 1.2 Pathogenese der Tuberkulose 8 1.3 Prävention der Tuberkulose 10 1.4 Diagnostik der Tuberkulose 11 1.4.1 Radiologische Untersuchungen 12 1.4.2 Mikrobiologische Untersuchungen 12 1.4.3 Molekularbiologischer Nachweis 13 1.5 Nachweis der Immunreaktion gegenüber M. tuberculosis 14 1.5.1 Tuberkulin-Hauttest 14 1.5.2 Serologische Tests auf M. tuberculosis-Infektion 16 1.5.3 Zelluläre Labortests auf M. tuberculosis-Reaktivität 18 2 AUFGABENSTELLUNG 22 3 MATERIAL UND METHODEN 23 3.1 Patienten 23 3.2 Blutentnahme und Zellpräparation 23 3.3 ELISPOT 25 3.4 Ringversuch zur externen Qualitätssicherung 26 3.5 Retrospektive Analyse der Daten des eigenen Labors und der Ringversuchsdaten 26 3.6 Analyse der Validierungsdaten aus Ettlingen 27 4 ERGEBNISSE 29 4.1 Analyse der ELISPOT-Daten des eigenen Labors 29 4.2 Ergebnisse des Instand-Ringversuchs 33 4.3 Analyse der ELISPOT-Wiederholungsmessungen 35 5 DISKUSSION 41 6 Zusammenfassung 48 Literaturverzeichnis 52 Lebenslauf 63 Danksagung 64 Erklärung über die eigenständige Abfassung der Arbeit 65

As pessoas vivendo com HIV tÃm probabilidade aumentada de desenvolver, apresentar formar graves, ter cepas multirresistentes e morrer por tuberculose. A profilaxia para infecÃÃo latente por Mycobacteium tuberculosis (ILTB) diminui a chance de ativaÃÃo de tuberculose (TB) numa mÃdia de 62% nessa populaÃÃo. Entretanto, o diagnÃstico da TB na sua forma latente à controverso. O teste tuberculÃnico (TT) à o Ãnico exame aprovado no Brasil para avaliaÃÃo dessa infecÃÃo, embora existam problemas tanto na sua realizaÃÃo quanto na sua interpretaÃÃo. Exames de liberaÃÃo de interferon-gama foram criados recentemente com o objetivo de aumentar a especificidade e a praticidade da investigaÃÃo da ILTB. Esse estudo se propÃs a avaliar como vem sendo feita a investigaÃÃo da ILTB e o desempenho do TT e do QuantiFERON-TB Gold In-Tube (QTF-GIT) em portadores de HIV. Foram selecionados ao todo 351 pacientes portadores de HIV e sem evidÃncia de TB ativa, admitidos em dois centros de referÃncia de Fortaleza-CE, no perÃodo de 2007-2010. Na admissÃo, 41,8% dos pacientes realizaram TT, 36,3% foram avaliados quanto a contato com TB e 28,4% tiveram radiografia de tÃrax. A profilaxia foi realizada para 73,3% dos pacientes com TT positivo. Houve diagnÃstico de ILTB em 25,3% dos pacientes de acordo com o TT e em 6,7% pelo QTF-GIT (p<0,001). A correlaÃÃo entre os resultados dos dois testes foi considerada fraca (k= -0,037). Resultado positivo do TT esteve associado com drogadiÃÃo (OR 7 CI: 1,53-32,11; p=0,01), contato com TB bacilÃfera (OR 13 CI: 2,7-62,83; p=0,001), profilaxia para ILTB prÃvia (OR 17,5 CI: 3,4-90,4; p<0,001), procedÃncia do interior do estado (OR 2,74 CI:1,04-7,22; p= 0,04). NÃo houve associaÃÃo entre QTF-GIT positivo e fatores de risco para TB. A mÃdia de contagem de linfÃcitos T CD4+ nos indivÃduos com TT positivo foi superior à mÃdia dos com TT negativo (535,8 vs. 373,4 cÃl/mm3; p=0,006), enquanto o inverso ocorreu em relaÃÃo ao QTF-GIT (277 vs. 438,3 cÃl/mm3; p= 0,055). A mÃdia do logaritmo da carga viral foi superior naqueles com QTF-GIT positivo (4,81 vs. 2,11 log10 cÃp/ml; p= 0,005). Mais da metade dos pacientes nÃo realizou TT, apesar da alta prevalÃncia de ILTB. O TT contou com maior nÃmero de testes positivos. O QTF-GIT mostrou-se superior para pacientes com elevada viremia e imunossupressÃo. Sugere-se o uso de ambos os testes de forma complementar para aumentar a chance de diagnÃstico de ILTB e diminuir os riscos de progressÃo da doenÃa. / People living with HIV have an enhanced chance to develop and to die of tuberculosis (TB). Many studies demonstrate that chemoprophylaxis for latent tuberculosis infection (LTBI) reduces the progression to active TB. Indeed, the diagnosis of LTBI is controversial. In Brazil, the only test approved for use is the tuberculin skin test (TST), however, this test is complicated by several problems due to application and interpretation of the exam. Recently developed interferon-gamma release assays (IGRA) using Mycobacterium tuberculosis-specific antigens have the advantage of decreased cross-reactivity and, therefore, increased specificity. The purpose of this study is to evaluate the adherence of LTBI diagnosis and to compare the results of the QuantiFERON-TB Gold In-Tube test (QTF-GIT) and TST in a population of HIV-positive individuals from a country with high prevalence of TB. A cross-sectional study was carried out with 351 HIV patients without active tuberculosis, attending outpatient in two reference centers, from November 2007- 2010. At admission, 41.8% had realized TST, 36.3% had been interrogated about TB exposure and 28.4% had performed a chest X-ray. Chemoprophylaxis was offered to 73.3% of TST positive patients. The TST and QTF-GIT results were positive in 25.3% and 6.7% (p<0.001) of the individuals, respectively. The agreement between the two tests was poor (k= -0.037). Drug use (OR 7, 95% CI 1.5-32.1; p=0.01), TB exposure (OR 13, 95% CI 2.7-62.83; p=0.001), previous LTBI prophylaxis (OR 17.5, 95% CI 3.4-90.4; p<0.001), and living outside the state capÃtal (OR 2.7, 95% CI 1-7.2; p= 0.04) were associated with a positive TST result. There is no association between QTF-GIT positive result and risk factors for TB. TST positive individuals had a higher mean CD4+ cell count than those with TST negative result (535.8 cell/mm3 vs. 373.4 cell/mm3; p=0.006), in contrast to QTF-GIT positive result (277 cell/mm3 vs. 438.3 cell/mm3; p= 0.055). Higher viral load was associated with QTF-GIT positive result (4.8 log10 cop/ml vs. 2.1 log10 cop/ml; p= 0.005). Despite of Brazil being a country with a high burden of TB, more than half the patients have not realized TST, which appears to be more sensitive than QTF-GIT for diagnosis of LTBI. Otherwise, QTF-GIT shows better results in patients with advanced immunosuppression and high viral load. We suggest the use of both tests to increase LTBI diagnosis and decrease the risk of disease progression.

Thesis (PhD (Molecular Biology and Human Genetics))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: The growing global epidemic of human tuberculosis (TB) results in 8 million new cases of this disease and 2 million deaths annually. Control thereof will require greater insight into the biology of the causative organism, Mycobacterium tuberculosis, and into the pathogenesis of the disease. This will benefit the design of new vaccines and diagnostic assays which may reduce the degree of both disease transmission and progression. Animal models have played a vital role in the understanding of the aetiology, pathogenesis, and treatment of TB. Much of such insight has been obtained from experimental infection models, and the development of new vaccines, for example, is dependant on these. Nonetheless, studies utilising naturally occurring TB in animals, such as those which have investigated the use of interferon-gamma release assays (IGRA) for its diagnosis, have contributed substantially to the body of knowledge in this field. However, there are few such examples, and this study sought to identify and investigate naturally occuring animal TB in South Africa as an opportunity to gain further insight into this disease. During the course of this study, the dassie bacillus, a distinctly less virulent variant of M. tuberculosis, was isolated from a rock hyrax from the Western Cape Province of South Africa. This has provided new insight into the widespread occurrence of this organism in rock hyrax populations, and has given impetus to further exploring the nature of the difference in virulence between these pathogens. Also investigated was M. tuberculosis infection in dogs in contact with human TB patients. In so doing, the first reported case of canine TB in South Africa was described, v a novel canine IGRA was developed, and a high level of M. tuberculosis infection in these animals was identified. This supports human data reflecting high levels of transmission of this pathogen during the course of human disease. Additionally, the fact that infected companion animals may progress to disease and potentially act as a source of human infection was highlighted. However, an attempt to adapt a flow cytometric assay to study cell-mediated immune responses during canine TB revealed the limitations of such studies in species in which the immune system remains poorly characterised. The use of IGRAs to diagnose TB was further explored by adapting a human assay, the QuantiFERON-TB Gold (In-Tube Method), for use in non-human primates. These studies have shown that such an adaption allows for the sensitive detection of TB in baboons (Papio ursinus) and rhesus macaques (Macaca mulatta) and may be suitable for adaption for use in other species. However, they have also evidenced the limitation of this assay to specifically detect infection by M. tuberculosis. Finally, to contextualise the occurrence of the mycobacterial infections described above, and other similar examples, these have been reviewed as an opinion piece. Together, these investigations confirm that animal models will continue to make important contributions to the study of TB. More specifically, they highlight the opportunities that naturally occuring animal TB provides for the discovery of novel insights into this disease. / AFRIKAANSE OPSOMMING: Wêreldwye tuberkulose (TB) epidemie veroorsaak agt miljoen nuwe gevalle en twee miljoen sterftes jaarliks. Ingryping by die beheer hiervan vereis begrip van die biologie van die mikroörganisme Mycobacterium tuberculosis, die oorsaak van TB, asook van die patogenese van die siekte self. Hierdie kennis kan lei tot ontwerp van nuwe entstowwe en diagnostiese toetse wat gevolglik beide die oordrag- en vordering van die siekte mag bekamp. Dieremodelle speel lankal 'n rol in ons begrip van die etiologie-, patogenese- en behandeling van TB. Insig is grotendeels verkry vanaf eksperimentele infeksiemodelle, en ontwikkeling van entstowwe, onder andere, is afhanklik van soortgelyke modelle. Desnieteenstaande, studies wat natuurlike TB voorkoms in diere ondersoek, byvoorbeeld dié wat op die ontwikkeling van interferon-gamma vrystellingstoetse (IGVT) fokus, het merkwaardige bydrae gemaak tot kennis en begrip in hierdie studieveld. Daar is slegs enkele soortgelyke voorbeelde. Om hierdie rede is die huidige studie uitgevoer waarbinne natuulike diere-TB geïdentifiseer en ondersoek is in Suid-Afrika om verdere kennis en insig te win aangaande TB. Die "dassie bacillus", bekend om beduidend minder virulent te wees as M. tuberculosis, is tydens hierdie studie geïsoleer vanuit 'n klipdassie (Procavia capensis) in die Wes-Kaapse provinsie, Suid-Afrika. Insig in die wydverspreide voorkoms van hierdie organisme in klipdassie bevolkings is gevolglik verkry en verskaf momentum om die aard van verskil in virulensie tussen dié patogene te bestudeer. vii Voorts is M. tuberculosis infeksie bestudeer in honde wat in kontak is met menslike TB pasiënte en word die eerste geval van honde TB dus in Suid-Afrika beskryf. In hierdie groep diere, is 'n hoë vlak van M. tuberculosis infeksie geïdentifiseer deur gebruik te maak van 'n nuut ontwikkelde IGVT vir die diagnose van honde TB. Gevolglik ondersteun dié studie bevindinge van menslike studies wat toon dat besondere hoë vlakke van M. tuberculosis oordrag voorkom gedurende die verloop van die siekte. Verder toon die studie dat geïnfekteerde troeteldiere 'n bron van menslike infeksie kan wees. 'n Poging om 'n vloeisitometriese toets te ontwikkel om die aard van selgefundeerde immuunreaksies te bestudeer in honde met TB toon die beperkings van dergelike studies in spesies waarin die immuunsisteem gebrekkig gekarakteriseer is. Die gebruik van IGVT'e in die diagnose van TB is verder ondersoek deur 'n menslike toets (QuantiFERON-TB Gold, In-Tube Method) aan te pas vir die gebruik van nie-menslike primaat gevalle. Hierdie studies toon gevolglik dat so 'n aanpassing toepaslik is vir hoogs sensitiewe deteksie van TB in chacma bobbejane (Papio ursinus) en rhesus ape (Macaca mulatta), en mag ook aangepas word vir gebruik in ander spesies. Tog word die beperkings van hierdie toets om infeksie wat spesifiek deur M. tuberculosis veroorsaak uitgelig. Ter afsluiting word hierdie studie in konteks geplaas deur 'n oorsig te gee van bogenoemde- en soortgelyke gevalle van dierlike infeksie deur mikobakterieë in Suid-Afrika. Hierdie studies bevestig dat dieremodelle steeds belangrike toevoegings maak tydens die bestudering van TB en lig veral die moontlikhede uit dat bestudering van natuulike TB in diere kan lei tot die ontdekking van nuwe insigte ten opsigte van die siekte self.

Tuberculosis (TB) remains the main cause of death in people living with HIV (PLHIV). Indeed, PLHIV have a 20-30% greater risk of developing TB compared to HIV-uninfected subjects and have lower TB treatment success rates. In 2014, among the 9.6 million incident cases of TB reported worldwide, 12% occurred in PLHIV and 0.4 million deaths from HIV-associated TB were recorded.Mycobacterium tuberculosis is the main etiological agent for TB. For a majority of individuals, the immune response upon infection by M. tuberculosis is sufficient to prevent the development of disease, but insufficient to clear the bacteria. This leads to the persistence of viable M. tuberculosis in diverse cells with no resulting clinical manifestations, an entity known as latent tuberculosis infection (LTBI). The resulting reservoir of M. tuberculosis is vast, and an estimated one third of the world population is concerned. For subjects with LTBI, the life-time risk of reactivation and progression to TB lies between 5 and 10%. However, if co-infected with HIV, the risk is much greater and reaches 10% per year. According to a Cochrane review in 2010, the screening and treatment of LTBI in PLHIV reduces this risk by 30-60%. This prevention strategy is therefore widely recommended. However, the implementation of LTBI screening and treatment into standard HIV-care has been limited. In this work, three different approaches have been used to understand and address this issue, focusing on a low TB-incidence and high-income setting.The first approach was to assess the implementation of LTBI screening in HIV-care across Belgium and identify its barriers as perceived by the caregivers on the field. Raising awareness to this issue was an indirect objective of the study. A multi-choice questionnaire was sent to 55 physicians working in a Belgian AIDS reference center or satellite clinic. A response rate of 62% was obtained. Only 20% of participants performed LTBI screening on all their patients and notable variations in the screening methods used were observed. A large majority of participants were in favor of targeting LTBI screening to HIV-infected patients at highest risk of TB rather than a systematic screening of all PLHIV. These results have been communicated to the Belgian LTBI working group, currently updating the national LTBI screening guidelines. Indeed, targeting screening to those at highest risk of TB is an attractive strategy in low-TB incidence countries and is already recommended in the United Kingdom. However, to date, no score assessing the risk of TB in PLHIV has been validated. Among the barriers to LTBI screening identified by the participants of this first study, the most frequently reported were lack of sensitivity of screening tools, risk associated to polypharmacy and toxicity of treatment. Improving the sensitivity of LTBI screening was the cornerstone of the second approach. The available screening tools for LTBI are the tuberculin skin test (TST) and two Interferon-gamma release assays (IGRAs): the QuantiFERON-TB Gold-IT (QFT-GIT) and the T-SPOT.TB®. All three lack sensitivity in PLHIV. Various strategies to discover superior LTBI screening tools are therefore being explored, including the development of IGRAs in response to alternative M. tuberculosis antigens to those used in the QFT-GIT or T-SPOT.TB®. A potential candidate is the native Heparin-Binding Haemagglutin (nHBHA), a methylated M. tuberculosis protein regarded as a latency-associated antigen. An in-house IGRA based on nHBHA (nHBHA-IGRA) has been shown to be a promising LTBI screening tool both in immunocompetent adults and in hemodialysed patients. The contribution of this nHBHA-IGRA to the detection of M. tuberculosis in PLHIV was therefore investigated. Treatment-naïve HIV-infected subjects were recruited from 4 Brussels-based hospitals. Subjects underwent screening for latent TB using the nHBHA-IGRA in parallel to the classical method consisting of medical history, chest X-ray, TST and QFT-GIT. Prospective clinical and biological follow-up ensued, with repeated testing with nHBHA-IGRA. Among 48 candidates enrolled for screening, 9 were diagnosed with LTBI by combining the TST and QFT-GIT results (3 TST+/QFT-GIT+, 1 TST+/QFT-GIT- and 5 TST-/QFT-GIT+). All 3 TST+/QFT-GIT+ patients, the TST+/QFT-GIT- patient as well an additional 3 subjects screened positive with the nHBHA-IGRA. These 3 additional patients had known M. tuberculosis exposure risks compatible with LTBI. During follow-up (median 14 months) no case of TB was reported and nHBHA-IGRA results remained globally constant. Multiplex analysis confirmed IFN- as the best read-out for the assay. From this study, we concluded that the nHBHA-IGRA appears complementary to the QFT-GIT for the screening of LTBI in PLHIV and the combination of the two tests may increase the sensitivity of screening. A large-scale study is however necessary to determine whether combining nHBHA-IGRA and QFT-GIT offers sufficient sensitivity to dismiss TST, as suggested by our results. In the same study, a group of HIV-infected adults with clinical suspicion of active TB were also recruited and tested with nHBHA-IGRA. Contrary to results in HIV-uninfected subjects, the nHBHA-IGRA could not discriminate between LTBI and active TB in PLHIV. This is an important caveat as HIV-infected subjects may present subclinical TB.A different angle was used for the third approach to the problem of LTBI in PLHIV. Systemic immune activation (SIA) is one of the principal driving forces in the natural course of HIV-infection. Despite long-term viral suppression by combination antiretroviral treatment (cART), a low-level SIA persists and is associated with an early-onset of age-associated disorders such as cardiovascular disease, dementia and osteoporosis. Causes of SIA in PLHIV are multiple and certain chronic infections appear to be implicated. A recent study in South Africa found that LTBI in PLHIV was associated with an increase in circulating activated CD8+ T-cells. If LTBI should contribute to the persistence of SIA, its screening and treatment could have an additional benefit on the clinical outcome of PLHIV. To investigate this theory, the expression of T-cell activation markers (CD38 and HLADR) as well as the level of plasmatic markers of immune activation (IL-6, sCD14, D-Dimers) were compared between subjects presenting active TB, subjects with LTBI and M. tuberculosis-free persons, with and without HIV-infection. In accordance with previous studies, active TB was associated with higher levels of SIA biomarkers in both HIV-infected and -uninfected groups. Among the HIV-uninfected subjects, no significant difference in biomarker level was found between those presenting LTBI and those with no evidence of M. tuberculosis. The effect of LTBI on activation biomarkers in the HIV-infected groups remained inconclusive because of the small number of individuals in the HIV+/LTBI group. Further investigation is therefore warranted. Interestingly, it was found that plasmatic markers may have a greater sensitivity for the detection of M. tuberculosis-associated SIA than the T-cell activation markers, an important result for future studies.Overall, LTBI in PLHIV is a challenging topic, in particular because of the lack of a gold-standard for the diagnosis of LTBI. Despite suboptimal tools, the evident clinical impact of LTBI screening and treatment in PLHIV on TB incidence justifies its implementation in standard HIV-care. In low TB-incidence countries, who, when and how to screen for LTBI in PLHIV remains unclear. This work offers an overview on the subject with particular focus on possible measures for improvement in the field. / Doctorat en Sciences médicales (Médecine) / info:eu-repo/semantics/nonPublished