Implementing molecular tuberculosis diagnostic methods in limited-resource and high-burden countries

Existing infrastructure

A major challenge in sequencing methods is that they generate vast amounts of data that require strong expertise and powerful machines to extract and analyse. The required infrastructure to effectively obtain results using WGS is not available in many laboratories in LICs. The data obtained from sequencing methods need to be properly stored in data warehouses that are yet to be developed in most LICs. Cloud storage represents another storage solution; however, the existing internet capacity needs to be improved before this option can be realistically accessible in LICs.

Cost of the machines and reagents

The cost of WGS varies by the type of assays, equipment, and software used. The commercial cost of sequencing an MTB genome is between USD 150 and USD 300 [6]. This cost includes the process of resequencing, where the MTB genome sequence is compared to a reference genome.

The costs are increased when establishing a sequencing facility in low-income settings due to the costs of shipments, customs, and the eventual profit margin for local companies. Lessons learnt from implementing WGS in Kyrgyzstan show that the main cost drivers were not using library preparation and sequencing kits to their full capacity during initial training, and partly by prices charged by the regional distributor exceeding those charged to laboratories in Europe and North America [7]. In addition to hardware costs, data manipulation and analysis are done using software licensed and commercialised by companies, representing an additional cost to institutions in LICs.

Training and capacity building

Local capacity needs to be built in both the laboratory and bioinformatics aspects of MTB WGS as many institutions continue to suffer from an insufficient number of experienced personnel that can perform, supervise, and train others in bioinformatics and sequencing research [8]. The fields of sequencing and novel TB diagnostic methods are interdisciplinary, requiring knowledge and training in biology, chemistry, and bioinformatics. In addition to training laboratory staff, there is a need to build capacity for clinicians to interpret the results of sequencing and adequately incorporate them into their routine clinical practice.

Adapting infrastructure capacity to needs

Decision makers at the country level need to be aware of and adapt the sequencing capacity to the country's needs. National programmes should avoid obtaining the sequencing infrastructure and not using it, because when not in use the maintenance of the machines is even higher, with implicit cost consequences. Therefore, LICs could opt for a national referral hub where all samples would be processed and then returned to health centres. Moreover, several countries could share a regional referral hub for sequencing TB strains, with shared infrastructure and human resources. Another effective strategy would be to mutualise resources with other pathologies (e.g. coronavirus disease 2019 (COVID-19), malaria, or even cancer) for surveillance, antimicrobial resistance control and research [9–11]. The lessons learnt from the COVID-19 pandemic response could inform TB sequencing technologies and inform local and regional stakeholders on the best strategies to implement [9, 12, 13].

Advocating for lowering costs

When sequencing methods are widespread and used at a large scale, there are compelling arguments to advocate with companies to lower the cost of machines and reagents. Using next-generation sequencing (NGS) (USD 50–100) instead of WGS (USD 150–300) as a diagnostic tool would help reduce costs without reducing the advantages of these innovative tools [6]. Mutualising costs with other countries for third-generation sequencing techniques like nanopore might address some of the existing challenges regarding infrastructure or cost, and could be a competitor for current sequencing companies in low-income and high-burden countries [14].

Building bioinformatics and laboratory capacity

An increasing number of universities are offering training in bioinformatics in low-resource settings. Bioinformatics education in Africa is currently being shaped through initiatives such as H3ABioNet, a project funded through the US National Institutes of Health [15]. One of the network's main remits is to build human capacity in bioinformatics in Africa. Proposed solutions include continuous mentorship and support, train-the-trainer programmes, internships, knowledge transfer programmes and e-learning platforms [16]. While being a much-needed initiative in the current landscape, the long-term sustainability of this kind of project remains of concern, with countries needing to take ownership of bioinformatics training and capacity building. In the future, most of the technology could be automated, leaving less of the processing burden to laboratory staff.

Data interpretation and sharing

Data analysis and visualisations should be available in real-time, and be simplified and easy to interpret by laboratory staff and clinicians to facilitate clinical decisions. Access to knowledge and results sharing should be available to researchers and decision-makers in resource-limited settings through open access publications and open license software.