Test to Treat TB: Impact of Sputum Sequencing-guided Individualised Therapy on Outcomes in Drug-resistant Tuberculosis

Resistance to anti-tuberculosis drugs is a continually growing problem. Multidrug-resistant tuberculosis (MDRTB) is resistance to at least rifampicin and isoniazid, and extensively drug-resistant TB is additional resistance to a fluoroquinolone and a second injectable line drug. Methods currently employed in testing for resistance are inadequate and a contributing factor to the 40-50% MDR-TB treatment success rate. Current drug susceptibility testing methods are slow for most drugs, taking weeks. Rapid molecular methods such as the line probe assays, e.g. Hain GenoType MDRTBplus and sl, provide resistant calls to only a limited number of drugs, and are often less useful in smear negative patients. Molecular technologies such as sequencing can provide a comprehensive readout of drug resistance and are able to detect resistant populations at very low levels (≤1%), thus enabling individualized therapy. This can be done directly from sputum. Targeted sequencing amplifies regions of genomic DNA associated with resistance prior to sequencing. Rapid analytic software is used to process the raw sequence data, identify resistance causing mutations and provide a readout of clinically relevant information. However, the feasibility, and more importantly the impact, of this approach has not been evaluated in a clinical trial to establish proof of concept. Aim 1: To conduct a randomised controlled trial to determine the impact of sputum-based targeted sequencing in detecting resistance to second-line TB drugs compared to the current programmatic standard of care (Hain MDRTBplus/sl and adjunct phenotypic drug susceptibility testing) when used to inform of treatment for MDR-TB. Aim 2: To compare currently available drug resistant sequencing pipelines for diagnostic accuracy, sensitivity, specificity and predictive value as compared to culture based phenotypic drug susceptibility testing. Aim 3: To compare the feasibility, accuracy, turn-aroundtime, and cost implications of the