Molecular epidemiology of tuberculosis in low and high incidence settings

Abstract

Tuberculosis (TB) surpassed HIV/AIDS in 2015 to become the greatest infectious disease killer on the planet, while the inexorable rise of drug-resistant TB adds another layer of complexity to TB control challenges in both high and low incidence settings. The advances in molecular genotyping of Mycobacterium tuberculosis has promised new but yet validated evidence and performance indicators to inform jurisdictional TB control programs. The overall aim of this thesis was to improve our understanding of the epidemiology of M. tuberculosis through integration of traditional genotyping and geospatial methods in low and high incidence settings, exemplified, respectively, by New South Wales (NSW), Australia, and Mongolia. It also explored the added value of WGS in improving our understanding of M. tuberculosis transmission dynamics in low incidence settings. The following hypotheses were tested: The application of high-resolution genotyping of M. tuberculosis can reveal the unique features and dynamic changes in molecular epidemiology patterns over short period of time, and the synthesis of epidemiological and high-resolution M. tuberculosis genotyping data can identify spatio-temporal hotspots of potential disease transmission of relevance to TB control programs. To explore these hypotheses we examined mycobacterial interspersed repetitive unit-variable number tandem repeat (MIRU-24) genotypes of M. tuberculosis isolates recovered from patients in NSW and Mongolia, respectively, and corresponding drug susceptibility and epidemiological data. Thirty strains from NSW were also subjected to whole genome sequencing (WGS). In NSW important recent changes were demonstrated in the M. tuberculosis population structure. Three pre-dominant phylogenetic lineages (Beijing, East African Indian (EAI) and Delhi/CAS) accounted for most cases, but between 2010 and 2012, EAI strains overtook Beijing as the most prevalent genotype. This shift in M. tuberculosis population structure reflects increased migration to Australia from the Indian Sub-continent. Twenty-four loci mycobacterial interspersed repetitive unit-variable number tandem repeat (MIRU-24) demonstrated superior resolution to MIRU-12, but the resolution of Beijing lineage strains remained sub-optimal. Relatively high rates of identical MIRU-24 profiles (12.8%) suggested significant local transmission. Significant MIRU clustering of Beijing lineage strains was resolved by WGS. Genome sequencing demonstrated that only 2 strains within a 21-member Beijing cluster represented a probable recent transmission chain. The EAI clusters identified by MIRU-24 were completely resolved by WGS, demonstrating the value of WGS to limit unnecessary cluster investigation and guide better-targeted public health responses. Synthesis of geospatial and genotypic data offered important insights into the local TB transmission in NSW and indicated that the predominant majority of human cases of TB caused by Beijing lineage strains were reactivation of latent infection acquired in the country of origin rather than recent transmission.

In Mongolia, programmatic management of multi-drug resistant TB (MDR-TB) led to a rapid increase in MDR-TB case finding. Spatio-temporal mapping demonstrated clustering of MDR-TB cases along the Trans-Siberian railway line, suggesting initial introduction along the railway line with subsequent spread to more remote provinces. Molecular analysis of MDR-TB isolates recovered from cases diagnosed in 2012, especially those classified as Category 1 treatment failure, revealed predominance of Beijing lineage strain, with a high percentage of MIRU-24 clustering. Combined analysis of genotypic and drug susceptibility data suggested that the MDR-TB epidemic in Mongolia has been mainly driven by Beijing lineage strains resistant to all first-line drugs. These findings have not been reported before and have significant implications for TB control efforts within Mongolia. Research objectives have been achieved and the methods and findings from the thesis could be valuable to not only to NSW and Mongolian TB control programs but could assist other countries to gain in-depth understanding of dynamics of the disease spread and to make the vision of TB free world one step closer.