Investigation of novel biomarkers of tuberculosis (TB) treatment response in clinical trials and clinical practice

Abstract

Tuberculosis remains a major public healthcare issue. One main challenge in effectively managing the disease and novel drug development is the lack of predictive biomarkers. Current methods of determining the numbers of viable bacteria are inaccurate. Moreover, ‘persister’ bacteria living in a reduced metabolic state affect treatment response and may be responsible for relapse. New short regimens will need to eradicate both actively replicating and persister cells. The capacity to accurately measure the burden of bacteria in multiple cell states will assist the development of shorter and better TB regimens. Serial sputum samples from 5- week high dose rifampicin II (HIGHRIF2) Phase 2B clinical trial were used to: 1) evaluate two quantitative sputum culture techniques compared with measurements of TB-MBLA, 2) monitor phenotypic changes of mycobacteria during therapy with fluorescence microscopy, 3) investigate pharmacodynamic responses during treatment using a newly developed RT-qPCR based cell viability assay. Study results highlighted contamination causing a large proportion (30.4%) of missing data points in culture, while TB-MBLA produced complete data sets. TB-MBLAmeasured bacterial load correlated with solid culture SSCC (p<0.0001) and MGIT time-to-positivity (p<0.0001). Based on our pilot study, TB-MBLA identified identical result to the HIGHRIF2 study with fewer patients. Our novel RT-qPCR cell viability assay simultaneously measures the ratio of pre-16S rRNA to the total rRNA mass measured bacterial burden and cell physiology. Mathematical modelling showed the composition of TB bacilli changed rapidly from predominantly active replicators to metabolically suppressed growth-inhibited population in response to drug exposure. Fluorescence microscopy results indicated the proportion of two phenotypically distinct M. tuberculosis cell-types “lipid-rich” and “lipid poor” changed during drug therapy, suggesting the emergence of a persister phenotype. This work shows the potential value of TB-MBLA and RTqPCR based techniques as a monitoring tool and as a methodology to rapidly evaluate metabolic state of M. tuberculosis ex vivo.

Date of Award	30 Nov 2021
Original language	English
Awarding Institution	University of St Andrews
Supervisor	Stephen Gillespie (Supervisor) & Derek Sloan (Supervisor)