Research forecasts global antimicrobial resistance threats for next two decades
Research led by King’s College London has analysed antimicrobial resistance (AMR) on a global scale to predict how resistance patterns could evolve by the year 2050, identifying around 210 resistance traits that could pose the greatest future risk.
AMR is considered one of the most urgent global public health threats, with experts predicting AMR could cause 39 million deaths between 2025 and 2050. AMR is not a single problem, and instead involves many different genes, pathogens, hosts and environmental factors.
The new study, published in Cell Genomics, used machine learning and forecasting to identify which genetic factors or ‘traits’ linked to AMR in pathogens are most likely to become future threats, and which may decline over time. Understanding which AMR traits are most likely to spread could help target future interventions and investment towards the highest-risk threats.
The researchers analysed 16 bacterial species identified by the World Health Organization (WHO) as critical priority pathogens, including Klebsiella, Acinetobacter and Escherichia coli, which are associated with high mortality and limited treatment options.
The research analysed data collected from across 127 countries and involved three main stages. First, researchers analysed over 45,000 genomes from the 16 bacterial species and AMR data using machine learning to identify which resistance genes were genuinely linked to antibiotic resistance.
Secondly, the team examined more than 1,000 environmental, health and socioeconomic indicators, including poverty, climate and healthcare trends, to understand how these global factors may change in the future.
Finally, the researchers linked these global trends with resistance patterns to identify which antimicrobial resistance traits are most strongly associated with future environmental and socioeconomic changes, allowing them to forecast the resistance threats most likely to grow by 2050.
Analyses revealed 210 pathogen-specific AMR traits projected to increase by 2050. These resistance traits were then narrowed down to 32 high-risk threats based on factors including their presence in WHO priority pathogens, their ability to move between bacteria, and their occurrence across multiple hosts. These 32 threats were strongly linked to future spread and associated with key drivers including mortality, poverty, healthcare access and population density.
Socioeconomic inequality and living conditions, such as overcrowding and access to sanitation, emerged as major predictors of increasing AMR. The study also highlights that the highest-risk resistance mechanisms are often highly mobile, allowing them to spread rapidly between bacteria, animals and humans.
Professor Tania Dottorini, Professor of Artificial Intelligence in Science at King’s College London and senior author of the paper, said: “Our research uses a multi-scale, multi-modal approach that has never been applied in this way before. By identifying which resistance traits are increasing, where they are spreading, and what is driving them, we can better target surveillance, policy and interventions towards the threats that are most likely to impact global health in the future.
“Reducing antibiotic use alone won't be enough. Tackling AMR requires structural interventions on inequality, sanitation, nutrition and health equity alongside stewardship. We believe that our findings provide a roadmap for targeted AMR interventions.”
The researchers hope that the findings will help guide future global surveillance strategies and support more targeted approaches to tackling AMR.
The study was supported by funding from UKRI through the Medical Research Council, Biotechnology and Biological Sciences Research Council and Innovate UK, in partnership with French National Research Academy (ANR), Swiss National Research Foundation (SNSF), Fondazione Regionale per la Ricerca Biomedica (FRRB) as part of the international Joint Programming Initiative on Antimicrobial Resistance (JPIAMR).
Sadhana Sharma, BBSRC AMR lead, said: “This important study demonstrates the value of a whole-systems approach, integrating genomics, advanced machine learning and global socioeconomic data to better understand the complexity of antimicrobial resistance patterns.
“By identifying the resistance traits most likely to emerge by 2050, it supports a shift from reactive to proactive action, helping to direct resources towards the highest-risk threats.
“By recognising where these risks are concentrated, we can better coordinate efforts to mitigate what remains one of the most serious threats to global public health.”
Journal
Cell Genomics
No comments:
Post a Comment