This thesis investigates active TB case finding and antiretroviral therapy for tuberculosis control in a setting with high HIV prevalence in Cape Town, South Africa. Many countries in sub-Saharan Africa have seen a worsening tuberculosis epidemic since the 1990s. Rising tuberculosis incidence rates have largely been attributed to high HIV prevalence in this region. Conventional tuberculosis control efforts focus on passive case finding and high cure rates in smear-positive patients, achieved through short course chemotherapy. These control strategies are insufficient in controlling the tuberculosis epidemic where HIV prevalence is high. Additional control strategies have been proposed, including active tuberculosis case finding, isoniazid preventive therapy for HIV infected individuals, infection control and antiretroviral therapy. The feasibility, uptake, yield, treatment outcomes and costs of population-based active tuberculosis case finding are investigated in the first part of the thesis. The second part determines losses along the HIV care pathway, community antiretroviral coverage and the association between coverage and tuberculosis risk. The main finding is that population-based active tuberculosis case finding linked to a mobile HIV testing service had a high uptake and yield. Treatment outcomes in patients diagnosed through active case finding were as good as outcomes in patients diagnosed through passive case finding in primary care clinics in Cape Town. Costs were USD 1,177 per TB case diagnosed and USD 2,458 per 3 successfully treated TB case, in an incremental costing analysis adopting a health service provider perspective. Analysis of the HIV care pathway in a peri-urban impoverished settlement in the greater area of Cape Town highlighted substantial losses along the pathway between HIV diagnosis and antiretroviral therapy. These results illustrate the operational challenges in achieving high treatment coverage. Antiretroviral coverage in this community increased from 18% in 2004 to 84% in 2009. Increasing antiretroviral coverage was associated with decreasing tuberculosis risk among patients receiving antiretroviral therapy, even controlled for timeupdated CD4 count, suggesting an effect on transmission, not just on individual risk reduction. The impact of active tuberculosis case finding and antiretroviral therapy on tuberculosis incidence on a population level was beyond the scope of this thesis. Large scale cluster randomized controlled trials are needed to investigate the effect of these strategies on tuberculosis control. In the meantime researchers conducting active tuberculosis case finding studies should be encouraged to collect data on treatment outcomes and costs. In addition further interventions are needed to increase retention and linkage to care in individuals prior to initiating antiretroviral therapy.

Background: Tuberculosis progresses despite potent Thl responses. A putative explanation is the presence of a subversive Th2 response. However interpretation is confounded by a novel cytokine, IL-452, a splice variant and inhibitor of IL-4. Methods: The expression of Thl cytokines (IFN-y), IL-452, Th2 cytokines (IL-4) and sCD30 was investigated in whole blood, lung lavage and mononuclear cell cultures from donors with TB, TB-HIV co-infection, and matched controls. Results: After validation of a fluorogenic real-time RT-PCR assay, the half-life of IL- 4 mRNA, but not IL-452, was found to be prolonged in TB vs controls (P<0.002). mRNAs for IL-4 and IL-452 were elevated in unstimulated cells from blood and lung lavage of patients vs controls (p<0.005). Patients with TB expressed significantly greater mRNA levels of both cytokines in T-cells (p<0.05 compared to controls where expression was predominantly in non-T cells). Radiological disease correlated with the IL-4/IFN-y ratio and sCD30 (p<0.005). Tuberculosis antigen upregulated expression of IL-4 relative to IL-462 in mononuclear cell cultures from tuberculosis patients (P<0.05). By contrast, though HIV-TB co-infected donors had increased IL-4 in blood and lung lavage, in lung the predominant form was IL-452. After chemotherapy, in tuberculosis and in HIV-TB co-infection, IL-4 mRNA levels remained unchanged whilst IL-462 increased (p<0.05). Conclusions: A Th2-like response, prominent in T cells, and driven by TB antigen, is present in TB and is modulated by treatment suggesting a role for IL-4 and its antagonist, IL-452 in the pathogenesis of TB and their ratio as a possible marker of disease activity. Furthermore, enhancement of IL-4 mRNA stability, a hitherto undescribed regulatory mechanism in human TB, may facilitate the immunopathological effect of IL-4. The specific antigens inducing the IL-4 response require identification to facilitate future vaccine development strategies. Further studies are required to determine whether IL-4 facilitates systemic HIV progression in co-infected patients.

Tuberculosis is a major threat to human health. About one third of the world's population is latently infected with Mycobacterium tuberculosis . In these cases the bacillus is in a state of low metabolic activity, making eradication difficult with conventional chemotherapy, which targets actively metabolizing organisms. The mechanisms by which M. tuberculosis reactivates to cause disease are currently unknown but a better understanding could greatly improve the treatment of tuberculosis. Resuscitation-promoting factor is a protein first identified in the supernatant of stationary phase cultures of Micrococcus luteus. It is active in picomolar concentrations, increasing the number of culturable M. luteus cells from dormant populations and shortening the lag phase of growth of small inocula. Bioinformatic searches reveal over 40 examples of rpf-ke genes in the high G-C cohort of Gram-positive bacteria, including M. tuberculosis , which contains five rpf gene orthologues. The work presented here investigated aspects of the M. tuberculosis Rpfs. Improvements in solubility of recombinant mycobacterial (M. tuberculosis and M. smegmatis) Rpfs were achieved by manipulating induction times and temperatures during protein expression and by using new hosts and vectors and producing novel fusion proteins. New assays were devised to measure the biological activity of recombinant Rpfs, using ATP bioluminescence of M. luteus cultures. A phage display library for M. tuberculosis was constructed, in an attempt to identify a protein receptor for Rpf. Rpf expression in human infection was investigated for the first time, using immunocytochemistry. Anti-Rpf antibodies were applied to human tissue sections infected with M. tuberculosis. Rpf was found to be located within epithelioid giant cells and in the immediate vicinity of acid-fast bacilli in necrotic centres. The presence of Rpf in human tuberculosis infection demonstrated in this work suggests that Rpfs may have a role in controlling dormancy of the bacilli in human disease.

Drug resistant tuberculosis (TB) is increasing worldwide and it is now estimated by the World Health Organisation (WHO) that 7% of TB cases globally are resistant to isoniazid (INH). INH is a key drug for the treatment of TB, alongside rifampicin (RIF). Understanding factors which influence the emergence and propagation of INH resistance and the influence of INH resistance on TB treatment success are vital in improving global TB control efforts. Vietnam is ranked 12th of 22 high burden countries for TB and has a high prevalence of INH resistance (25%) with a relatively modest prevalence of TB resistant to both INH and RIF (multi-drug resistant, MDR TB) (2.7%) The first study in this thesis developed rapid screening tests for both RIF and INH resistance in Mycobacterium tuberculosis isolates which showed high accuracy. The specificity and the sensitivity of the MAS-PCR test for INH resistance compared to the conventional phenotypic DST were 100% [95% CI 92.9-100%] and 90% [95% CI 82.4-95.1 %], respectively. The second study demonstrated that the minimum inhibitory concentration (MIC) for INH is influenced by both the mutation responsible for resistance to INH and the lineage backbone of the M.tuberculosis isolate. A two-way ANOVA of MIC including both strain lineage (p=0.003) and resistance mutation (p<0.001) showed highly significant independent effects of both factors on MIC level. MIC to INH of isolates with a katG315 mutation (2ug/ml) was significantly higher compared to MIC to INH of isolates with an inhA-15 mutation (0.25p.g/ml) and wild-type isolates (0- 0.lug/ml) (p<0.00l).

This thesis examines the contrasting roles of genetic drift and selection on the emergence of drug resistance in Mycobacterium tuberculosis . In clinical practice some alleles of rifampicin resistance are isolated more frequently than others. To identify if this variation is due to genetic drift or selection, the mutation rate to rifampicin resistance in M. tuberculosis (H37Rv) was determined. PCR-SSCP analysis revealed only three patterns from the rifampicin resistant isolates, each pattern arising at the same mutation rate (Mann-Whitney U test P>0.5). Fitness, defined as the ratio of generations of resistant and susceptible cells formed in mixed culture, of the differing rifampicin resistant alleles was determined relative to the parent fully susceptible strain. There was a significant correlation between fitness and the clinical isolation rate of each allele (regression analysis P=0.026). The fitness of two isolates, with identical IS6110 RFLP pattern isolated from 2 siblings, was determined. One isolate had developed multi drug resistance, the second isolate had remained fully drug susceptible. The fitness of the drug resistant isolate was significantly lower than the drug susceptible isolate (matched pair t test p=0.002). The decreased relative fitness of the resistant isolate implied a physiological cost for the development of drug resistance. Isolates of M. tuberculosis from three patients involved in a hospital outbreak of multi drug resistant tuberculosis were obtained. The fitness of these isolates was determined relative to H37Rv. Isolates obtained from the same patient did not vary in fitness (one way ANOVA p=0.34). However, the isolates from the three different patients had differing fitness values (one way ANOVA p=<0.001). This implied that there is adaptation of the isolates to the individual patient. In conclusion, selection has a major role in adaptation of drug resistance in M. tuberculosis. This adaptation includes adaptation to the infected host as well as drug resistance.

No description available.

No description available.

No description available.

No description available.

No description available.