Bacterial RNA profiling to predict tuberculosis treatment failure

At present there is no way of accurately predicting tuberculosis (TB) treatment failure and relapse to active disease. This results in many patients who might safely stop the toxic six-month therapy earlier and their treatment is unnecessarily extended. It also extends the length and cost of clinical trials to assess new therapies, creating a bottleneck to progress.

This project will test whether the transcriptional response of TB isolated from patient sputa during the first few days of treatment can predict relapse months later. Building my study demonstrating that sputa TB mRNA signatures differ between patients and reflect clinical/microbiological measures of disease, this proof-of-concept prospective patient study will for the first time define markers of relapse. TB RNA from patient sputa will be assayed by targeted and genome-wide profiling techniques at diagnosis and at multiple time points during drug therapy. RNA yields and gene expression signatures will be analysed to determine key methodological parameters concerning the collection and detection of sputa TB RNA and to assess applicability in the UK and Cameroon. TB mRNA signatures from patients who have relapsed and those who have been successfully treated will be compared using this approach to identify biomarkers that predict relapse.

The findings of this study could transform the management of TB.