Engidawork,Ephrem(PhD, Professor)Aklillu,Eleni (PhD, Professor)Shibesh,Workineh (PhD, Assoc. Professor)Sidamo,Temesgen2025-08-132025-08-132023-06https://etd.aau.edu.et/handle/123456789/6744Background: Multi-drug resistant tuberculosis (MDR-TB) has hindered the therapeutic success of TB. Patient, drug, and pathogen-related factors are critical to understanding, characterizing, and predicting drug resistance, which help to tailor drug dosing taking into account inter-individual differences and patient needs. The highly active World Health Organization (WHO) Group A fluoroquinolones, FQs (moxifloxacin, MXF, and levofloxacin, LFX), exhibit inter-individual pharmacokinetic variability (IIV) that may result in low drug exposure. The 24-hour area under the concentration-time curve over the minimum inhibitory concentration (AUC0-24/MIC) is a strong predictor of FQ exposure and positive treatment outcome. Ethiopia is one of the world nations with a high prevalence of MDR-TB. However, there is a lack of data on factors affecting MDR-TB treatment success and one-size-fits-all scenario is the single most common therapeutic strategy of TB in Ethiopia. Objective: This dissertation aimed to study treatment outcome and its predictors, characterize and predict MXF and LFX resistance in MDR-TB patient population using phenotypic and genotypic methods as well as a PK/PD model-based study. Methodology: A prospective observational study of 80 MDR-TB patients was carried out. Treatment outcomes in patients receiving MXF-and LFX-based regimens were compared at the end of treatment. We determined the plasma concentrations of MXF and LFX and their PK parameters. A PK/PD (AUC0- 24/MIC) analysis to predict the recommended probability of target attainment (PTA) was carried out for IV both drugs. The potential patient covariates which impact the model estimated PK profiles were also screened. Drug resistance was characterized by phenotypic and genotypic methods, and the association between patient and pathogen related variables, FQ-resistance and treatment outcome were evaluated. Result: The total treatment success in this study was lower than the report in previous local and international studies. The LFX-based MDR-TB regimen outperformed the MXF-based regimen. Unsuccessful treatment outcome was predicted by a delayed culture conversion rate, history of alcohol intake, lesion of lung cavities, serum levels of creatinine (Cr.). A one-compartment model with adjustments was fit to the LFX and MXF concentrations. Cr. and body mass index (BMI) were covariates identified to have impact on clearance and apparent volume distribution (Vd) of LFX and MXF, respectively. Exposures to LFX and MXF were lower in study participants than in other settings. However, LFX-treated groups experienced higher drug levels and exposure with dose. Nine clinical isolates (11.3%) were resistant overall, all of which were at least resistant to FQs, while 3 were resistant to both FQs and injectable drugs. Mutation in gyrA 94 was identified in 5 isolates. The MIC values were associated with patient treatment outcomes. Cavitary lesion and serum creatinine predict FQ-resistant tuberculosis. Conclusion: Exposure to either LFX or MXF may be inadequate, but LFX appears to provide better treatment outcomes. Patient clinical and behavioral variables can predict drug exposure, drug resistance and the overall treatment outcome. Even though the line probe assay (LPA) showed a moderate level of resistance in Ethiopian patients, genotypic test results do not always correlate well with phenotypic results. The actual drug resistance level could be higher than anticipated. Higher MIC values of clinical isolates correlate with a higher risk of treatment failure. In such scenarios, dosage optimization may improve outcome. GyrA mutation is associated with FQ-resistance in Ethiopian patients. Further controlled clinical studies are needed to establish optimal doses and exposure of FQs.en-USFluoroquinolonesmulti-drug resistant tuberculosispopulation pharmacokineticstreatment outcomegyrA sequencingThesis