Thesis (M.S.)--Boston University

Infection & Immunity / Currently, there are 38 million people living with human immunodeficiency virus (HIV-1) worldwide and there were 680,000 HIV-related deaths in 2020 alone. The greatest cause of mortality in people living with HIV (PLHIV) is infection with opportunistic pathogens such as tuberculosis (TB), which accounts for one third of HIV-related deaths. PLHIV are 20 times more susceptible to TB and co-infection leads to significantly worsened outcomes in terms of both diseases. Humanized mouse (hu-mouse) models, which possess human immune cells for HIV to infect, have been useful for HIV research. Our aim is to create hu- mouse models of HIV, TB and co-infection to investigate disease progression, immune responses, therapeutics, prevention and vaccination. NOD-Rag1null-IL2rgnull (NRG) mice are highly immunocompromised mice that are traditionally used to generate hu-mouse models. We are also developing NRG mice that are transgenic for human HLA-DR4 and HLA-A2 (DRAG-A2) and similar mice have been reported to have improved immune responses. NRG and DRAG-A2 mice were humanized with hematopoietic stem cells obtained from human umbilical cord blood. DRAG-A2 mice had significantly higher engraftment success rates (defined as the percentage of mice with >10% hCD45+) as well as higher overall CD45+ leukocyte, CD4+ T cell, CD19+ B cell and CD14+ monocyte reconstitution in the blood compared to huNRGs. huNRG mice were permissive to infection with JR-CSF or NL4.3-Bal-Env HIV-1 intravaginally or systemically. huDRAG-A2 mice were also infected intravaginally with NL4.3-Bal-Env HIV-1. huDRAG-A2 mice, but not huNRGs, produced HIV-specific IgG, indicating improved immune responses. huNRG mice were infected intranasally with mCherry-Erdman, YFP-H37Rv or H37Rv Mtb. huDRAG-A2 mice were also infected with H37Rv. Human immune cell involvement and human-like granuloma formation was observed using flow cytometry and immunohistopathology. These findings show that the DRAG-A2 model may be optimal for investigating HIV, TB and co-infection, which continue to be serious global health concerns. / Thesis / Master of Science (MSc) / Human immunodeficiency virus (HIV) and tuberculosis (TB) are infectious diseases that affect millions of people worldwide every year. The greatest cause of death in people living with HIV is co-infection with TB and HIV-positive individuals are much more likely to get TB. Humanized mouse (hu-mouse) models possess human immune cells for HIV to infect and are useful for studying HIV. Our goal is to create hu-mouse models of HIV, TB and HIV/TB co-infection that will allow us to study how these diseases interact. We are currently developing a traditional hu-mouse model (known as NRG), as well as an improved next-generation model (known as DRAG-A2) with a more functional immune system. Both models have been successfully infected with HIV or TB. Only DRAG-A2 mice were able to make antibodies against HIV. The improved DRAG-A2 model will enable future studies on HIV, TB and co-infection, which continue to be understudied global problems.

There are an estimated 2 billion individuals infected with Mtb, and 37.7 million people living with HIV (PLWH) worldwide. HIV/Mtb co-infection increases the risk of developing active tuberculosis by over 20-fold, and 210,000 of 1.5 million deaths from TB were among co-infected PLWH in 2020. Therefore, development of effective TB vaccination, particularly within the vulnerable PLWH population, is an urgent global issue. With limited in vivo models to study co-infection, humanized NRG (huNRG) mice and humanized DRAG-A2 mice (a next-generation of huNRG mice expressing HLA class I and II transgenes with improved human immune reconstitution, huDRAG-A2) are promising tools for HIV and TB reserach as they develop robust human immune cell populations and recapitulate many aspects of HIV or TB clinical disease. HIV/Mtb co-infection was investigated using huNRG and hu-DRAG-A2 mice in separate experiments where intravaginal (with DMPA pre-treatment) or intraperitoneal HIV-1 infection was administered, respectively, and intranasal infection of Mtb was administered 3.5 weeks later. Both huNRG and huDRAG-A2 mice recapitulated hallmark features of HIV/Mtb co-infection such as severe granuloma pathology, hCD4+ T cell depletion in lung and spleen tissue, and human like lung pathology such as Mtb-infected foamy macrophages in the granuloma. Co-infected huDRAG-A2 mice also displayed significantly higher bacterial burden in the lungs, increased extrapulmonary dissemination into spleen and liver, and significantly lower hCD4+ T cells in the peripheral blood post-Mtb infection when compared to the Mtb-only infected group. To investigate TB vaccine immunogenicity, huNRG and huDRAG-A2 mice were immunized with a novel trivalent vaccine, AdCh68MV. Upon intranasal immunization, both models showed trends of developing higher Mtb antigen-specific hCD4+ T cell responses in the lung and spleen. Overall, this project sets the initial stages of a pre-clinical HIV/Mtb co-infection model in huNRG and huDRAG-A2 mice appropriate for immune investigations, therapeutic and vaccination development. / Thesis / Master of Science in Medical Sciences (MSMS) / There are over 2 billion individuals infected with TB and 37.7 million people living with HIV (PLWH) worldwide. When someone is co-infected with both diseases, the risk of death is greatly increased. Research in co-infection and developing effective TB vaccination for PLWH are urgent global issues. Animal studies are currently limited because studying HIV requires human immune cells. Our lab has established humanized mice (hu-mice) that develop many different human immune cells and are useful for HIV/Mtb co-infection research. When hu-mice were co-infected, they showed more dying lung tissue, immune cell loss, and bacteria in the lungs. Hu-mice were also used to study human immune responses to a novel TB vaccine delivered to the lungs. Trends of higher immune responses towards TB were observed in the lung and spleen of immunized hu-mice. Overall, this project shows the utility of hu-mice as pre-clinical models of HIV/Mtb co-infection and Mtb vaccine studies.

No description available.

No description available.

Circulating monocytes have the capacity to mature into either macrophages or dendritic cells in tissue, both of which play important roles in the induction and effector phase of immune response. In TB, the macrophages are the central player in the host-bacteria interaction as the main mycobacterial reservoir. In HIV disease, monocyte lineage cells are one of the two main cell types (along with CD4+ T-cells) in sustaining intracellular HIV infection. Monocytes are heterogeneous population with three functionally distinct subsets namely classical, intermediate and non-classical monocytes. The three subsets exist in a continuum, and have a certain plasticity or flexibility to develop into multiple roles depending on the local and tissue environment. In the current study we sought to evaluate the frequencies of these three subsets in participants with TB, HIV and TB/HIV co-infection. While previous studies had shown that the intermediate and non-classical monocyte subsets were elevated relative to classical monocytes, very little had been done in these disease groups regarding more comprehensive characterization of these subsets. In particular, we wished to quantitate the expression of multiple sets of cell surface and intracellular molecules of high relevance using multi-parameter flow cytometry from a functional point of view. In publication I, we evaluated Toll-like receptors (TLRs) expression in each of the study cohorts. TLRs are vital pattern recognition receptors by monocyte lineage cells and signal the induction of crucial functions. We focused on three such TLRs (TLR2, TLR4 and TLR9) which have been shown to be involved in many monocyte lineage cell interactions with mycobacterial and HIV infections. We observed enhanced expression of TLR2 and TLR4, but not TLR9 in TB and HIV. TLR4 was particularly high in patients with TB, but also in HIV. We observed comparable increase of TLR4 irrespective of monocyte subset. However, TLR2 expression exhibited a different pattern. Levels among the most prominent classical monocyte subsets were identical in all four cohorts, healthy controls (HC), HIV, TB, and TB/HIV co-infection. In contrast, TLR2 expression was significantly elevated in both participants with HIV and TB, but not with participants with TB/HIV co-infection in the intermediate monocyte subset. We also observed correlations between TLRs and plasma cytokines that were disease and TLR specific. In publication II, we observed elevated chemokine receptors (CRs) expression which above healthy controls and exhibit a pattern of disease preference. Thus, CCR2 and CX3CR1 were the highest in participants with TB, followed by HIV and TB/HIV co-infection, whereas CCR4 and CCR5 were highest in participants with HIV, and less elevated in TB. CCR2 and CX3CR1 are critical for migration of monocytes to sites of TB infection, as determined by murine models. CCR4 and especially CCR5 have been implicated in migration of cells to distant organs but more as co-receptors for HIV infection. Thus, the observed pattern of CRs expression in these monocytes in different disease states would predict greater availability of these cells or their receptors for interaction with either TB or HIV organisms. From the perspective of the pathogen this would lead to enhanced “substrate”, whereas from the perspective of the host, this could lead to greater immune potential. As a final point, we also observed that the pattern of disease association of CRs was independent of the monocyte subset. In publication III, we explored the expression of Programmed cell death-ligand 1 (PDL1) on the three monocyte subsets. Like many of the other molecules we have addressed in this thesis, PDL1 expression was enhanced in participants with HIV, TB, and TB/HIV co-infection. Among participants with HIV, PDL1 was correlated with HIV-1 viral load. The enhanced expression was apparent in all three subsets, but it was particularly prominent in the intermediate monocyte subset. Moreover, PDL1 expression was the highest in participants with TB/HIV co-infection. The implications behind these observations is that the subset thought to have the greatest potential for T cell antigen presentation had the highest levels of the T cell down-regulatory PDL1 molecule in the cohort of patients particularly participants with TB/HIV co-infection. Participants with TB/HIV co-infection have the greatest potential to be immuno-compromised and as a result the very need for enhanced not depressed APCs function. In addition, we also observed the PDL1 levels were correlated with multiple plasma, mostly pro-inflammatory,cytokines. We analyzed cytokine mRNA levels of total monocytes to address the source of the cytokines but mRNA levels did correlate with neither plasma cytokine nor PDL1 levels. Considering all the phenotype analysis in each of the three studies together we could see two patterns emerging. In one scenario,surface molecules expression patterns were disease specific but independent of monocyte subset expression. In other words, whatever the underlying mechanism(s) involved in their regulation, those mechanisms apparently acted similarly in all three subsets. In another scenario, expression of surface molecules showed disease specific patterns, but molecules were particularly enhanced in the intermediate monocyte subsets. These findings imply that there exist mechanisms to modulate surface phenotypes and functions that are unique to a given subset. In conclusion, we have comprehensively defined the density of multiple molecules expressed by different subsets of monocytes and explored their differences in participants with TB, HIV and TB/HIV co-infection, as well as their correlations with microbial indices and plasma cytokines. Many molecules levels were elevated to some extent in all disease cohorts, but we observed patterns of expression which were particularly elevated in TB (CCR2, CX3CR1, and TLR2), those in HIV (CCR4, CCR5) and those on both (TLR4, PDL1). Molecule-disease associations were either independent of monocyte subset, or most readily revealed in a single monocyte subset. TB/HIV co-infection did not follow a consistent pattern in association with monocytes markers, in some cases more resembling TB, in others HIV, in others neither. Finally, to proof one possible mechanism of association between disease and monocyte phenotype, we explored correlations between monocyte markers and plasma cytokines. We observed significant positive and negative associations, frequently unique to a single monocyte subset or disease cohort, such as TB/HIV co-infected cohort. Collectively, the results imply that there are likely multiple mechanisms involved at many levels regulating the phenotype and function of monocytes, and these differ in different disease states.:Abbreviations ............................................................................................................ 3 Abstract ..................................................................................................................... 4 1. Introduction ........................................................................................................... 6 1.1 Epidemiology of Tuberculosis and Human Immunodeficiency virus ................... 6 1.2 The immunological response to TB and HIV ....................................................... 7 1.2.1 Innate immunity of TB ...................................................................................... 7 1.2.2 Innate immunity of HIV .................................................................................... 11 1.2.3 Immune checkpoint regulation in TB and HIV.................................................. 13 1.3 The role of monocytes in TB, HIV and TB/HIV ................................................... 14 1.3.1 Monocytes ....................................................................................................... 14 1.3.2 Abnormalities of monocytes in TB ................................................................... 15 1.3.3 Abnormalities of monocytes in HIV ................................................................. 16 1.3.4 Abnormalities on monocytes in TB/HIV co-infection ....................................... 17 1.4 The rationale for the thesis ................................................................................ 17 2. Objectives ............................................................................................................ 19 3. Publications .......................................................................................................... 20 4. Summary .............................................................................................................. 65 References ............................................................................................................... 69 Annex I: Author contribution ..................................................................................... 76 Annex II: Declaration of independent writing of the work ......................................... 77 Annex III: Curriculum Vitea ....................................................................................... 78 Annex V: Acknowledgment........................................................................................ 82 Annex VI ................................................................................................................... 84

Syndrome inflammatoire lié à la reconstitution immunitaire chez le patient coinfecté par le VIH et la tuberculose est une complication du traitement par des antirétroviraux, appelé TB-IRIS. Ce syndrome est souvent rencontré dans les pays en voie de développement. Son diagnostic se pose essentiellement sur la présentation clinique et nécessiter de se différentier des autres pathologies. Son évolution est souvent favorable ou sous corticoïde mais certaine forme est sévère et/ou mortelle. L'étude de leur mécanisme permettra d'identifier de marqueur prédictif, applicable à leur diagnostic précoce et à l'amélioration de leur prise en charge. Alors, nous avons proposés d'étudier le rôle de cellule NK et le rôle de lymphocyte T dans l'essai clinique de CAMELIA au Cambodge. Le résultat a montré l'élévation de la capacité de dégranulation de cellule NK est associé à la survenu de TB-IRIS et il peut être un marqueur prédictif. De plus, l'hyperactivation de cellule T effectrice et la diminution de cellule T régulatrice sont aussi observées. Le rôle de cellule T régulatrice n'est pas encore préciser. Le mécanisme régulateur de ce phénomène doit être ultérieurement exploré. / Inflammatory syndrome associated with immune reconstitution in patients coinfected with HIV and TB is one complication of antiretroviral treatment, called TB-IRIS. This syndrome more often encountered in developing countries. Diagnosis of this syndrome is mainly based of clinical presentation and needs to differentiate from other diseases. Their evolution is usually favorable or under corticosteroids, but some cases are severe and / or fatal. The study of their mechanism could lead to identify predictive markers, applicable to their early diagnosis and improved their management. Thus, we proposed to study the role of NK cell and T cell in the CAMELIA clinical trials which have conducted in Cambodia. The result showed that higher increase of NK cell degranulation capacity was associated with the occurrence of TB-IRIS and it could be a predictive marker. Furthermore, the hyperactivation of effector T cell and decrease of regulatory T cell were also observed. The implication of the regulatory T cell in this syndrome was not clear yet. The regulatory mechanism of this phenomenon should be further explored

Master of Public Health - MPH / Background: In South Africa, tuberculosis (TB) detection remains a major problem, as notified cases are estimated to account for only 68% of all incident cases. Health services have relied on passive case finding and this leads to missed or delayed diagnosis. In Cape Town, City Health has embarked on an active surveillance programme to systematically screen all adults seeking health care at PHC facilities for active TB, in order to identify undiagnosed incident TB cases and avert missed opportunities for treating TB. Aim: The aim of this study was to evaluate the effects of an intensified TB screening strategy on changes in facility level TB case finding in City Health PHC facilities in Cape Town.

A pesquisa teve como objetivo analisar o Programa de Controle da Tuberculose (PCT) no município de Ribeirão Preto -São Paulo, no período de 1998 a 2006. Utilizou-se para este propósito indicadores epidemiológicos e de desempenho construídos a partir de dados das fichas de notificação de tuberculose (TB) armazenadas no sistema de informação o EPI-Tb, da Secretaria municipal de saúde deste município e os dados populacionais de estimativas do DATASUS, do período de estudo. Selecionaram-se para o estudo casos novos de TB notificados e residentes no município, por ano de diagnóstico, excluindo-se casos atendidos em outros municípios e presidiários. No período compreendido entre 1998 a 2006 foram notificados no EPI-Tb da SMS/Ribeirão Preto 1623 casos novos de tuberculose, sendo que houve queda no número absoluto de casos e no coeficiente de incidência de 47,8% (50,01- 26,08) dos casos no período ou 5,3% ao ano. O risco de ser um caso novo de TB foi 2,4 vezes maior para homens que para as mulheres. Apesar do número de casos notificados serem maiores na faixa etária de 15 a 49 anos, o risco de adoecer por TB foi maior na faixa etária acima de 50 anos, a partir de 2001. O percentual de co-infecção HIV/TB ficou em 27,1% (prevalência mínima), mas a prevalência máxima foi de 32,7%. A forma clínica mais freqüente para os casos novos foi a pulmonar com 85%, enquanto para os casos co-infectados esta forma esteve presente em 58,3% deles e a extra pulmonar em 27,8%. O local de descoberta dos casos de TB foi 51% em ambulatórios (públicos e privados), 39% em hospitais (universitários, público e privados) e 10% outras formas. Uma das fragilidades observadas foi a baixa detecção de casos de TB no município que nos últimos anos ficou próxima de 45% e que as unidades básicas de saúde e PCTs realizam apenas um quinto das baciloscopias de escarro para diagnóstico que deveriam realizar (segundo estimativas do MS). No entanto uma das fortalezas foi a implantação do tratamento supervisionado no município, que iniciou efetivamente em 1998, e foi aumentando gradativamente, chegando em 2006 a supervisionar 76% dos casos. Houve melhora nas taxas de cura, ficando próximo à 72% e 50,5%, para os casos novos sem co-infecção e com co-infecção HIV/TB, respectivamente. A taxa de mortalidade por TB no município apresentou ligeira tendência de queda no período. Apesar da baixa letalidade no período, 50,8% dos óbitos por TB só foram diagnosticados e notificados após o óbito; indicando dificuldade de acesso ao diagnóstico e tratamento da TB nestes casos. Conclui-se que para melhorar a detecção de casos de TB no município serão necessárias mudanças na forma de acolher os indivíduos suspeitos de TB na atenção básica de saúde, facilitando seu acesso a essas unidades, além de investigar mais sintomáticos respiratórios na comunidade. Algumas ações de controle da doença poderiam ser descentralizas, como o tratamento supervisionado e controle de comunicantes. Para os pacientes co-infectados HIV/TB apenas o tratamento supervisionado não está sendo suficiente para alcançarem sucesso no tratamento. / The objective of the present investigation was to analyze the Program of Tuberculosis Control (PTC) in the municipality of Ribeirão Preto- São Paulo, during the period from 1998 to 2006. Epidemiological and performance indicators were used for this purpose, constructed from data of the charts of tuberculosis (TB) notification stored in the information system of EPI-Tb, of the municipal health Secretariat of this municipality and from estimate population data of DATASUS regarding the study period. New TB cases notified regarding patients residing in the municipality were selected according to year of diagnosis, with cases attended in other municipalities and prisoners being excluded. A total of 1623 new cases of TB were notified to EPI-Tb of the SMS/Ribeirão Preto during the period from 1998 to 2006, with a fall in the absolute number of cases and a 47.8% (50,01-26,08) reduction of the coefficient of incidence being observed during this period, corresponding to 5.3% per year. The risk of being a new TB case was 2.4 times higher for men than for women. Although the number of notified cases was higher for the 15 to 49 year age range, starting in 2001 the risk of becoming ill with TB was higher in the age range above 50 years. The percentage of HIV/TB co-infection was 27.1% (minimum prevalence), but the maximum prevalence was 32,7%. The most frequent clinical form for the new cases was the pulmonary one (85%), while this form was present in 58.3% of co-infected cases and the extrapulmonary form in 27.8%. The site of detection of TB cases was public and private outpatient clinics in 51% of cases, university, public and private hospitals in 38%, and other sites in 10%. One of the fragilities observed was the low detection of TB cases in the municipality, which remained close to 45% over the last few years and the fact that the basic health units and PTCs perform only one fifth of the sputum bacilloscopies they should perform for diagnosis (according to Health Ministry estimates). However, one of the strong points was the implantation of supervised treatment in the municipality, which was effectively started in 1998 and increased gradually, with 76% of cases being supervised in 2006. There was an improvement in cure rates, that reached 72% and 50.5% for non-co-infected new cases and HIV/TB co-infected cases, respectively. The TB mortality rate in the municipality showed a slight tendency to a fall during the period. Despite the low lethality observed, 50.8% of the TB deaths were only diagnosed and notified after death, indicating a difficulty in access to diagnosis and treatment of TB in these cases. We conclude that, in order to improve the detection of TB cases in the municipality, changes are needed in the reception of individuals suspected to have TB at basic health units, facilitating their access to these units, in addition to the investigation of more persons with respiratory symptoms in the community. Some actions for the control of the disease such as supervised treatment and the control of communicants could be decentralized. Supervised treatment alone is not sufficient for HIV/TB-co-infected patients to achieve successful treatment.

A pesquisa teve como objetivo analisar o Programa de Controle da Tuberculose (PCT) no município de Ribeirão Preto -São Paulo, no período de 1998 a 2006. Utilizou-se para este propósito indicadores epidemiológicos e de desempenho construídos a partir de dados das fichas de notificação de tuberculose (TB) armazenadas no sistema de informação o EPI-Tb, da Secretaria municipal de saúde deste município e os dados populacionais de estimativas do DATASUS, do período de estudo. Selecionaram-se para o estudo casos novos de TB notificados e residentes no município, por ano de diagnóstico, excluindo-se casos atendidos em outros municípios e presidiários. No período compreendido entre 1998 a 2006 foram notificados no EPI-Tb da SMS/Ribeirão Preto 1623 casos novos de tuberculose, sendo que houve queda no número absoluto de casos e no coeficiente de incidência de 47,8% (50,01- 26,08) dos casos no período ou 5,3% ao ano. O risco de ser um caso novo de TB foi 2,4 vezes maior para homens que para as mulheres. Apesar do número de casos notificados serem maiores na faixa etária de 15 a 49 anos, o risco de adoecer por TB foi maior na faixa etária acima de 50 anos, a partir de 2001. O percentual de co-infecção HIV/TB ficou em 27,1% (prevalência mínima), mas a prevalência máxima foi de 32,7%. A forma clínica mais freqüente para os casos novos foi a pulmonar com 85%, enquanto para os casos co-infectados esta forma esteve presente em 58,3% deles e a extra pulmonar em 27,8%. O local de descoberta dos casos de TB foi 51% em ambulatórios (públicos e privados), 39% em hospitais (universitários, público e privados) e 10% outras formas. Uma das fragilidades observadas foi a baixa detecção de casos de TB no município que nos últimos anos ficou próxima de 45% e que as unidades básicas de saúde e PCTs realizam apenas um quinto das baciloscopias de escarro para diagnóstico que deveriam realizar (segundo estimativas do MS). No entanto uma das fortalezas foi a implantação do tratamento supervisionado no município, que iniciou efetivamente em 1998, e foi aumentando gradativamente, chegando em 2006 a supervisionar 76% dos casos. Houve melhora nas taxas de cura, ficando próximo à 72% e 50,5%, para os casos novos sem co-infecção e com co-infecção HIV/TB, respectivamente. A taxa de mortalidade por TB no município apresentou ligeira tendência de queda no período. Apesar da baixa letalidade no período, 50,8% dos óbitos por TB só foram diagnosticados e notificados após o óbito; indicando dificuldade de acesso ao diagnóstico e tratamento da TB nestes casos. Conclui-se que para melhorar a detecção de casos de TB no município serão necessárias mudanças na forma de acolher os indivíduos suspeitos de TB na atenção básica de saúde, facilitando seu acesso a essas unidades, além de investigar mais sintomáticos respiratórios na comunidade. Algumas ações de controle da doença poderiam ser descentralizas, como o tratamento supervisionado e controle de comunicantes. Para os pacientes co-infectados HIV/TB apenas o tratamento supervisionado não está sendo suficiente para alcançarem sucesso no tratamento. / The objective of the present investigation was to analyze the Program of Tuberculosis Control (PTC) in the municipality of Ribeirão Preto- São Paulo, during the period from 1998 to 2006. Epidemiological and performance indicators were used for this purpose, constructed from data of the charts of tuberculosis (TB) notification stored in the information system of EPI-Tb, of the municipal health Secretariat of this municipality and from estimate population data of DATASUS regarding the study period. New TB cases notified regarding patients residing in the municipality were selected according to year of diagnosis, with cases attended in other municipalities and prisoners being excluded. A total of 1623 new cases of TB were notified to EPI-Tb of the SMS/Ribeirão Preto during the period from 1998 to 2006, with a fall in the absolute number of cases and a 47.8% (50,01-26,08) reduction of the coefficient of incidence being observed during this period, corresponding to 5.3% per year. The risk of being a new TB case was 2.4 times higher for men than for women. Although the number of notified cases was higher for the 15 to 49 year age range, starting in 2001 the risk of becoming ill with TB was higher in the age range above 50 years. The percentage of HIV/TB co-infection was 27.1% (minimum prevalence), but the maximum prevalence was 32,7%. The most frequent clinical form for the new cases was the pulmonary one (85%), while this form was present in 58.3% of co-infected cases and the extrapulmonary form in 27.8%. The site of detection of TB cases was public and private outpatient clinics in 51% of cases, university, public and private hospitals in 38%, and other sites in 10%. One of the fragilities observed was the low detection of TB cases in the municipality, which remained close to 45% over the last few years and the fact that the basic health units and PTCs perform only one fifth of the sputum bacilloscopies they should perform for diagnosis (according to Health Ministry estimates). However, one of the strong points was the implantation of supervised treatment in the municipality, which was effectively started in 1998 and increased gradually, with 76% of cases being supervised in 2006. There was an improvement in cure rates, that reached 72% and 50.5% for non-co-infected new cases and HIV/TB co-infected cases, respectively. The TB mortality rate in the municipality showed a slight tendency to a fall during the period. Despite the low lethality observed, 50.8% of the TB deaths were only diagnosed and notified after death, indicating a difficulty in access to diagnosis and treatment of TB in these cases. We conclude that, in order to improve the detection of TB cases in the municipality, changes are needed in the reception of individuals suspected to have TB at basic health units, facilitating their access to these units, in addition to the investigation of more persons with respiratory symptoms in the community. Some actions for the control of the disease such as supervised treatment and the control of communicants could be decentralized. Supervised treatment alone is not sufficient for HIV/TB-co-infected patients to achieve successful treatment.