Improving diagnostic approaches in extensively and multidrug resistant tuberculosis: new diagnostic tools and repurposing drugs against X/MDRTB

Abstract

Tuberculosis remains a major killer worldwide. Accurate, fast, and accessible diagnostics for both the disease and drug resistance remain challenges for tuberculosis (TB) control. In four different translational projects linked by the fact they dealt with laboratory diagnosis of drug resistance, I evaluated repurposed compounds and diagnosis of tuberculosis using resources already available in most microbiology laboratories in high-income countries and rapidly expanding to low and middle-income countries. Our aim was to utilise resources already available in most microbiology laboratories in high-income countries wherever possible with the belief that these systems would subsequently expand to low and middle-income countries. In Chapter 1, I found that the MGIT instrument missed true rifampicin resistance associated with specific rpoB mutations. This has now been overcome by the introduction of whole genome sequencing. In Chapter 2, carbapenems were tested against M. tuberculosis using different methodologies, showing that their role in the treatment of tuberculosis remains unclear. Testing in a more physiological model of the granuloma showed similar unclear results but confirmed the excellent activity of clofazimine, a drug that until recently could not be reliably tested in vitro. In Chapter 3, a MALDI-ToF MS lipidomic approach was used to identify M. tuberculosis and NTMs at species and some sub-species level. This approach could be easily introduced in UK microbiology laboratories today. In Chapter 4, urine was explored as a sample for the diagnosis of pulmonary tuberculosis using a lipidomic-MALDI-ToF MS approach and a PCR-based approach aiming to detect M. tuberculosis trans-renal DNA. Over the period of my PhD, I was part of the seminal projects that produced the evidence to introduce WGS for susceptibility testing and which recommended stopping routine phenotypic testing for these drugs. I was involved in the provision of phenotypic data, validation of comparative UK data, writing the standard operating procedure for DST and training technical staff to set up and read commercially available microtiter plates for slow growing mycobacteria. This included MGIT-based phenotypic test methodology which was used to pair with genotypic results. For first line drugs results were very good. Implementation of whole genome sequencing in the UK for the diagnosis, susceptibility testing and surveillance promised to overcome some of the issues mentioned above. However, 7 years later, a reliable phenotypic drug susceptibility testing methodology is desperately needed for both new compounds and to inform on those drugs for which genomic information is unclear or sparse. New understanding of pharmacokinetics-pharmacodynamics should also be incorporated in these models.