Persisting Mycobacterium tuberculosis (Mtb) evade rapid killing by chemotherapeutic agents and are the reservoir for treatment failure and relapse of tuberculosis (TB) treatment. To date, other than persistently positive cultures, no biomarkers have been identified that effectively predict which TB patients are at risk for treatment relapse after apparent cure. We will recruit a prospective cohort of patients undergoing treatment for MDR TB in Seoul, Korea, using a protocol in place for serial evaluations by [18F]-fiuoro-2-deoxy-Dglucose positron emission tomography/ computer tomography (FDG-PET/CT) during and after completion of therapy. The natural history of the post-treatment phase of TB disease will be investigated, with specific attention to identifying evidence of persistent organisms by FDG-PET/CT scan and defining biomarkers that predict PET/CT scan activity and clinical relapse. Candidate biomarkers include host transcriptome and proteome profiles, effector and memory T cell cells and polyfunctional T cells. These markers will be evaluated and correlated with treatment outcomes. The potential use of biomarker assays to monitor treatment response for MDR TB patients, whose therapy is more toxic and for whom second line drugs are less efficacious, would facilitate determination of more evidence-based treatment durations and provide an earlier endpoint for investigation of novel therapeutics.
Aim 1 will determine the association between abnormalities on FDG-PET/CT scan and relapse of TB patients after completion of treatment;
Aim 2 will develop and validate a model to predict Mtb persistence and treatment relapse of MDR TB patients based on changes in biomarkers;
Aim 3 will prospectively validate the most promising biomarkers as predictors of relapse in drug-susceptible TB. Identification of FDG-PET/CT scan abnormalities associated with persistent disease will provide important insights into the biology of mycobacterial persistence in humans and lead to optimal choice of potential biomarkers to evaluate as predictors of persistent disease.
TB treatment is long and a significant number of treated patients relapse. In this proposal we plan to identify markers that can tell whether a person is going to relapse way before they actually relapse.