CCRB-LMIC: Community Collection of Resistant Bacteria from LMICs

The ability to track the emergence and evolution of new strains of antibiotic-resistant bacteria is dependent on testing bacteria obtained from different environments. However, many LMICS are not equipped to collect, store and analyse these samples. Our solution will expand a pilot program that has successfully sourced >550 LMIC isolates.

Prof Mark Blaskovich

Director of Translation

Institute for Molecular Bioscience

The University of Queensland

To reduce or eliminate the impact of antimicrobial resistance on health systems worldwide, it is critical to monitor the emergence of new resistant bacterial strains, to identify and understand and the types of resistance mechanisms, and to test how well existing and newly antibiotics work against them. While several large international collections of clinical isolates exist for antibiotic evaluation, such as a commercial set available from International Health Management Associates, isolates from LMICs are significantly under-represented. Organisations in LMICs lack the infrastructure and resources to store and fully characterise clinical or agricultural/environmental isolates, with patient samples generally discarded after testing and agricultural/environmental samples often never tested.  This is particularly problematic as these LMICs often have high levels of AMR and can be the undetected source of new resistance.

Our solution is designed to enable access to these LMIC samples by collaborating with LMICs to facilitate isolate collection, storage and full characterisation. We will also provide infrastructure support to help collaborators characterise their bacteria. The validated isolates and associated phenotypic and genotypic data are assembled into a publicly accessible biobank and online database to facilitate global collaboration, supporting the development of new antibiotics and vaccines.

Our solution seeks to address the poor availability of well-characterised resistant isolates sourced from the clinic, livestock and the environment in LMICs. Our target audience is therefore divided into two categories:

1) researchers in LMIC organisations that will be empowered and supported to collect and analyse isolates and engage in collaborative research, and

2) the drug and vaccine development community, the end users of the isolates and their associated meta, phenotypic and genotypic data.

Our pilot program overcame significant challenges in identifying appropriate collaborators for isolate collection, and country-specific obstacles to provide them with financial support for the collection and shipment of isolates. Significant flexibility is needed to tailor solutions to individual collaborators: for some financial compensation is not acceptable, but materials and supplies are suitable. Some require salary support to specifically conduct isolate collection and some are keen to build their own sequencing capability.

For the drug and vaccine development community, the combination of genetic and phenotypic data will support programs into ‘genotype to phenotype’ prediction, while the sequencing data of antigens are valuable to vaccine developers. Novel lineage isolates are essential for researchers conducting fundamental studies into antimicrobial resistance, distribution of resistance, and to antibiotic developers. 

Our solution will provide an open access public repository of isolate data from clinical, environmental and agricultural sources in LMICs. It contributes to the goal of our crowd-sourcing antibiotic initiative CO-ADD, providing a public knowledgebase to stimulate research into antimicrobial drug discovery and antimicrobial resistance. As with the CO-ADD database (see db.co-add.org), isolate phenotypic and genotypic characterisation will be publicly available on a purpose build web-site and mirrored through established public databases (such as GenBank at NIH), making the data highly accessible to the global community. The public data will be available following FAIR principles.

Our solution will also create a biobank of isolates accessible to the global community of antimicrobial resistant researchers and antibiotic drug developers, delivering the option for them to access the physical isolate samples to perform new experiments. This service will be provided under a cost-recovery basis, and leverage the network of researchers established by CO-ADD.

Finally, we work closely with local LMIC researchers, finding bespoke solutions to enable them to participate in the program. This includes financial and logistical support, and in some cases capacity building through provision of sequencing capability via supply and training of MinION (ONT) equipment.

Our collaborative framework facilitates a cooperative approach to antimicrobial resistance research. As with our pilot program, all data generated at UQ (such as whole genome sequencing) will be shared with collaborators, providing them with robust, reliable, and trusted data from established laboratories and facilities. The exchange of knowledge through the provision of sequencing equipment and subsequent training will also empower LMIC researchers and support the advancement of their capabilities. This, along with the outputs of our collaboration will generate publications and contribute new knowledge and insights about AMR.

Through our funding model, LMIC researchers will also be provided with the practical resources they need to further their research. Since the model does not impose stipulations on how funds need to be spent, they are able to decide where the resourcing is allocated to best fit their needs. Importantly, this project will ensure that data from LMIC isolates are part of international databases and other resources used to develop preventive and therapeutic measures to address AMR. Representation within datasets will result in relevant solutions that address the challenges specific to LMIC

The proposed solution will use the first year to further existing collaborations with organisations in the eight LMICs from the pilot program, expanding the diversity of isolates to focus on environmental and agricultural samples that complement the clinical isolates already collected.

We will also use the first year to initiate new collaborations with additional LMIC organisations.  One of the major obstacles in the pilot program was executing a collaborative research agreement that fully addressed the needs of the collaborators while navigating their complex institutional landscapes. The first year will be used to learn about the collaborators and develop plans tailored to their needs. Based on experience, we envisage that this will take approximately 6 months. 

The first year will also be used finalise the open-access online database for dissemination of the data from both the pilot program and the new isolates collected. Isolate processing will be optimised along with quality control and data generation workflows for better efficiency and scalability.

This groundwork will enable us to scale up in the following years during which we will obtain and analyse additional isolates, as well as improve the analysis, visualisation, and data retrieval features of the online database.

The primary outcomes of this project are the physical library of characterised isolates and the accompanying open-access database.  

To evaluate the impact of these outcomes, we will monitor and evaluate operational progress by measuring:

i) number of collected and characterised isolates (300 Yr1, 350 Yr2, 350 Yr3),

ii) number of different organisations and origin of isolates (balance across regions and environmental/agricultural/clinical source).

ii) diversity analysis of collected isolates: species, genotype, phenotype and resistant profile (high diversity desired).

The analysis of the diversity of phenotype and genotypes from the pilot program will be used as reference benchmark. A forthcoming publication will provide a suitable standard reference for the collected data and isolates.

To evaluate the impact of the database, we will monitor the usage by number of unique users and the number of downloads from the database. We will also track citations and number of publications that are enabled by the data.

Engagement with LMICs is also an integral part of this project, so the number of organisations participating will also be measured. Through sustained engagement with the collaborators, we will also measure outcomes beyond publication such as any IP generated that is supported by the LMIC data.

The pilot project demonstrated that the main barrier was establishing the appropriate collaborative research framework and compensation mechanism. While the customisation of the legal frameworks were only minor, the LMIC approval processes were more complex, often involving different layers of authorities, and sometimes requiring additional agreements to be executed. In addition, several countries have more stringent standards in place, requiring ethics approvals or export approvals prior to execution of the agreement.

Remuneration to the collaborating organisations in the pilot program also required very flexible mechanisms to ensure compensation adhered to institutional policies. Our current proposal includes a dedicated administrative coordinator to navigate the approval and reimbursement processes by assisting the LMIC collaborators with any documentation they need to get the approval through their system, while also complying with UQ policy.

A second barrier the pilot program encountered was the identification of LMIC collaborators with sufficient resources and time to collaborate. Relationship management with the collaborators is essential to accommodate the cultural and procedural differences. We will address this barrier by initially using the existing collaborators in eight LMICs, supplementing them with potential collaborators already identified in additional LMICs, with whom discussions were already initiated during the pilot project.

We are applying to the Trinity Challenge as we aspire to generate a high-quality dataset populated with phenotypic and genotypic characterisation of resistant isolates from LMICs where AMR is most prevalent. Our vision is that this database and accompanying isolate biobank will be a valuable resource to the global research community, contributing to the development of new treatment options and a better understanding of the spread of AMR between the environment, animals and humans.

The Trinity Challenge provides a timely opportunity to keep this initiative progressing, maintaining the momentum and experience that we have created, while expanding the approach to address AMR questions which are relevant to a wider research community. The main barrier that Trinity can help address, apart from the obvious difficulty in seeking funding for this initiative, is to provide access to a network of connections that can help to expand the initiative to a larger community in the LMICs, identifying potential collaborative partners in new LMICs.

The team will continue pursuing collaboration and input from its existing collaborative network with organisations such as CARB-X and GARD-P, though these entities are generally focused on developing new antimicrobial treatments and hence mots interested in the analysis of clinical isolates, rather than environmental and agricultural.

We would highly welcome any mentoring and suggestions from the collaborators which support the Trinity Challenge, especially for opportunities within the field of agriculture, livestock and environmental surveillance. There are a number of initiatives from organisations such as the UN, WHO, Wellcome Trust, Fleming Fund and Gates Foundation, that focus on surveillance of AMR in various LMICs. Such initiatives could provide valuable support for our project, by adding more diverse isolate collections to our data, by providing assistance for analysis of the data we have collected, or by providing guidance on which environments or species the collection of new isolates should focus on to maximise the outcome for the global AMR community.