The World Health Organization (WHO) has declared tuberculosis (TB) as the infectious disease with the highest mortality worldwide, resulting in 1.8 million deaths in 2015 despite development of potent antibiotics and vaccine [1, 2]. TB places its heaviest burden on the poorest and most vulnerable populations globally, aggravating existing inequity . Management of the disease has been continually challenged by the emergence of antimicrobial resistance. Multidrug-resistant TB (MDR-TB) defined as resistance to two most commonly administered drugs (rifampicin and isoniazid) remains a public health crisis as only one in four patients are diagnosed , and only one in five receive treatment  of whom only half are cured. In 2015, roughly 480,000 new MDR-TB patients and additional 100,000 rifampicin resistant TB (RR-TB) patients were reported  and an estimated 190,000 human deaths have been recorded worldwide in 2014 . These data reflect unaddressed gaps in detection and treatment of MDR and RR-TB, which require the same treatment regimen to achieve a cure. Recognizing the immediate need to rapidly detect drug resistant TB, WHO has recommended implementation of molecular diagnostic assays, including GeneXpert MTB/RIF (soon to be replaced by the GeneXpert MTB/RIF Ultra) and GenoType MTBDRplus (Hain) [4, 5]. As initial diagnosis of RR-TB increasingly relies on molecular tests , large volumes of data are being produced. If not coupled with appropriate analysis, raw data from these assays will not be fully utilized and crucial clinical and public health information might be missed. Processing, mining, and analysing these data through molecular and epidemiological approaches can provide opportunities to not only serve the individual patient in receiving prompt effective treatment, but also public health objectives for TB elimination [6, 7]. The product of such analysis could also bring about prediction of molecular test results from whole genome sequences (WGS) of RR-TB strains akin to how classic strain typing of Mycobacterium tuberculosis can now be predicted from this data . This tool could then aid policy makers in selecting the most useful diagnostic tool to recommend for implementation in a particular setting. Further modelling of data derived from RR-TB diagnostic tests could help identify transmission hotspots and facilitate prediction of new outbreaks, allowing for prompt detection of additional patients who can undergo early treatment and prevent further transmission in the community. Through collaboration with National TB Control Program (NTP) managers, policy makers, and researchers from various fields of molecular biology and epidemiology, this PhD project aims to (1) develop a bioinformatics tool for prediction of molecular test results, (2) use both this tool and generated molecular test data to determine distributions of resistance mutations and associated molecular test patterns in each country and (3) design a RR-TB surveillance system taking into account needs and preferences of NTPs for rapid detection of RR-TB transmission hotspots and potentially implement and assess the system in the community.
|Effective start/end date||12/01/17 → 2/06/20|
IWETO expertise domain