In recognition of World Tuberculosis (TB) Day on March 24th, this science story explores current work underway at the NML using bioinformatics to predict drug resistance in Tuberculosis. Innovative analysis tools have the potential to unlock important pathogen information, such as drug resistance, that can support timely clinical decision-making. Applied research to improve diagnostic assays for this priority public health threat is one way that NML is contributing to international efforts to end the global TB epidemic.
What was known about this area prior to your work, and why was the research done?
Tuberculosis (TB) is an airborne infectious disease caused by Mycobacterium tuberculosis (M. tuberculosis). This bacterium is responsible for more deaths than any other single infectious agent worldwide. In 2017, almost 1,800 Canadian TB cases were reported. As with many other pathogens, drug resistance poses a serious threat to the health of infected individuals and makes treatment or control efforts much more difficult. Drug-resistant TB strains usually take much longer to treat and require higher priced drugs that have more severe side effects. In 2016, it was reported that drug-susceptible TB treatment cost $18,000 USD while treatment for the most drug-resistant form of the disease, known as extensively drug resistant (XDR) TB, cost over $500,000 USD.
The gold standard for drug resistance testing is a slow process that can take several weeks. Predicting drug resistance by detecting specific mutations in the bacterial DNA through whole genome sequencing (WGS) takes much less time and has the potential to guide treatment options in a more timely manner.
What are your most significant findings from this work?
Two well-known publicly available software tools were evaluated for their accuracy to predict drug resistance profiles of M. tuberculosis using WGS data from strains received at the NML over the past two years. Of the two applications, Mykrobe performed best in that it found more mutations in resistant strains, including 34% more mutations than previous methods of DNA sequencing. Fortunately, most M. tuberculosis strains received at the NML are sensitive to all four of the most important drugs (Isoniazid, Rifampin, Ethambutol and Pyrazinamide). The software tool correctly predicted strains to be sensitive for all four drugs with an accuracy of 98.5%. When 18,700 publically available genomes were analyzed, drug resistance was correctly predicted with a sensitivity of 75-95% for the four drugs and a specificity of 93-98%. Accuracy of tests are often gauged by their sensitivity and specificity, which describe results as true positives and negatives, respectively.
What are the implications or impact of the research?
This research showed that genomic prediction of antimicrobial susceptibility in M. tuberculosis has high sensitivity and specificity for most drugs, but falls short of the World Health Organization’s targets of 90% sensitivity and 95% specificity. Additional research to understand drug resistance mechanisms and identify other DNA mutations is necessary to achieve a higher correlation to the existing gold standard test.
A significant advantage of using WGS data to predict drug resistance is that it can quickly detect all possible drug resistance genes at once. The predicted resistance pattern could inform clinical treatment options and provide alternate drug choices, if side effects occur. Predicting drug resistance with WGS data has the potential to decrease the time for diagnostic results to 5-15 days from the traditional 2-6 weeks. The faster drug resistance is identified, the sooner patients can receive appropriate treatment to clear the infection and reduce the chance of spreading the disease.
Additional References of Significance:
- CRyPTIC consortium and the 100,000 Genomes Project, et al. Prediction of Susceptibility to First-Line Tuberculosis Drugs by DNA Sequencing. N Engl J Med 2018 Oct 11;379(15):1403-15. doi: https://doi.org/10.1056/NEJMoa1800474
- World Health Organization. Global tuberculosis report 2019. Geneva (CH): World Health Organization; 2019. https://www.who.int/tb/publications/global_report/en/
- LaFreniere M, Dam D, Canadian Tuberculosis Laboratory Technical Network, et al. Tuberculosis drug resistance in Canada: 2018. Can Commun Dis Rep 2020 Jan 2; 46(1):9 15. doi: https://doi.org/10.14745/ccdr.v46i01a02