Improving tuberculosis surveillance by detecting international transmission using publicly available whole genome sequencing data

Abstract

Improving the surveillance of tuberculosis (TB) is one of the eight core activities identified by the World Health Organization (WHO) and the European Respiratory Society to achieve TB elimination, defined as less than one incident case per million [1]. Monitoring transmission is especially important for multidrug-resistant (MDR) Mycobacterium tuberculosis isolates – defined as being resistant to rifampicin and isoniazid – and for extensively drug-resistant (XDR) M. tuberculosis isolates – defined as MDR isolates with additional resistance to at least one of the fluoroquinolones and at least one of the second-line injectable drugs. In 2017, the WHO estimated that worldwide more than 450,000 people fell ill with MDR-TB and among these, more than 38,000 fell ill with XDR-TB [2].

The rapid advance in molecular typing technology – especially the availability of whole genome sequencing (WGS) to identify and characterise pathogens – gives us the chance to integrate this information into disease surveillance. For TB surveillance, it is possible to combine the results of molecular typing of isolates from the M. tuberculosis complex with traditional epidemiological information to infer or to exclude TB transmission [3,4]. This is of particular relevance if transmission occurs among multiple countries, where epidemiological data such as social contacts are more difficult to get and where data exchange is more difficult to organise. The European Centre for Disease Prevention and Control (ECDC) reported 44 events of international transmission (international clusters) of MDR-TB in different European countries between 2012 and 2015 [5]. In that report, the authors inferred TB transmission using the mycobacterial interspersed repetitive units variable number of tandem repeats (MIRU-VNTR) typing method. However, this method has limitations such as low correlation with epidemiological information in outbreak settings and low discriminatory power [3,6]. In comparison, WGS analysis offers a much higher discriminatory power and allows inferring (or excluding) TB transmission at a higher resolution [4]. In a recent systematic review, van der Werf et al. identified three studies that used WGS to investigate the international transmission of TB [7].

In recent years, the amount of available WGS data is increasing, especially because sequencing has become cheaper [8]. In addition, more and more authors deposit the raw data of their projects in open access public repositories such as the Sequence Read Archive (SRA) of the National Center for Biotechnology Information (NCBI) [9]. These publicly available raw WGS data for thousands of isolates enable the re-use and the additional analyses at a large and global scale [10]. For example, it is possible to compare genomic data among different studies or countries since the data are available in a single place. Moreover, new software tools can be tested using the same raw WGS data [11]. However, standards in bioinformatics analysis and interpretation of these WGS data for surveillance purposes are not yet fully established [12].

We aimed to assess the usefulness of raw WGS data of global MDR/XDR M. tuberculosis isolates available in public repositories to improve TB surveillance. Specifically, we wanted to identify potential international events of TB transmission and to compare the international isolates with a collection of M. tuberculosis isolates collected in Germany in 2012 and 2013.