Treatment optimisation of multidrug-resistant tuberculosis

Abstract

A successful treatment outcome is seen in only 60% of persons treated for multidrugresistant tuberculosis (MDR-TB) worldwide, defined as resistance to both rifampicin and isoniazid. To improve these disturbingly low numbers, treatment optimisation is highly needed. Therefore, this thesis will evaluate how to optimise a treatment regimen using both older and repurposed drugs in studies on regimen composition, resistance detection, target attainment for efficacy, and reduction of adverse drug reactions.

In the first retrospective observational study (study I), we evaluated the effect of pyrazinamide treatment on end-of-treatment outcomes in a cohort (n=508) of persons affected by MDR-TB in Karakalpakstan, Uzbekistan. We found no evidence (aOR 0.86, 95% CI 0.51-1.44, p=0.6) that pyrazinamide treatment was associated with end-oftreatment outcomes. In study II, pyrazinamide treatment was evaluated using time to sputum culture conversion in a historical Swedish MDR-TB cohort (n=157). We found strong evidence that no pyrazinamide treatment compared to receiving pyrazinamide treatment was associated with a longer time to sputum culture conversion (aHR 0.49, 95% CI 0.29-0.82, p=0.007), when accounting for genotypic drug susceptibility testing (DST).

In study III, we assessed the total exposure of moxifloxacin and levofloxacin over the minimum inhibitory concentration of the infecting Mycobacterium tuberculosis strain in persons with MDR-TB in Xiamen (n=32), China. In this prospective observational study, we showed that no participants treated with levofloxacin, and 60-73% receiving moxifloxacin, reached the proposed efficacy targets when dosed according to the Chinese national guidelines. In the last retrospective observational study (study IV), we evaluated risk factors for adverse drug reactions associated with linezolid treatment (n=132) for MDR-TB in Sweden. We found strong evidence that a daily linezolid dose of ≥12 mg/kg was associated with a higher risk of peripheral neuropathy (aHR 2.92, 95% CI 1.09-7.84, p=0.033), anaemia, or leukopenia. Moreover, in an exploratory analysis, a linezolid trough concentration of ≥2 mg/L was associated with a higher risk of anaemia and thrombocytopenia.

In conclusion, treatment with pyrazinamide seems to have a role in MDR-TB, at least in terms of improving interim outcomes. The use of genotypic DST is highly promising and may simplify and shorten the time to resistance testing. Adequate dosing of fluoroquinolones is important as underdosing could reduce treatment effects. Linezolid dose adjustment based on weight, or a high trough level might avoid adverse drug reactions. Importantly, dose adjustment needs to consider both efficacy and risk of adverse drug reactions, therefore, therapeutic drug monitoring can be a useful tool in the quest to personalise treatment.