Data modelling breakthrough in the fight against antimicrobial resistance
Antimicrobial resistance (AMR) is one of the leading global public health threats, and was directly responsible for 1.27 million deaths in 2019.
The impact of AMR is far-reaching and puts many of the gains of modern medicine in jeopardy, including increasing the risk of procedures and treatments such as surgery, caesarean sections and cancer chemotherapy.
AMR occurs when “microorganisms” such as bacteria, fungi and parasites change over time and no longer respond to medicines.
Microorganisms that develop AMR to several of the commonly-used antimicrobial medicines are often referred to as “superbugs”, which can cause infections that are harder to treat, increasing the risk of disease spread, severe illness and death.
At the same time, the development of antibiotics to treat bacterial infections cannot keep up with the rapid rise in AMR. While there are concerted efforts in play to tackle the issue – both at Monash University and around the world – the World Health Organisation states the scope of research is “inadequate in the face of rising levels of resistance”.
Calling on data science
At the Monash Institute of Pharmaceutical Sciences (MIPS), researchers are harnessing advanced data science to help address the problem.
In a world first, the team has published a study that shows how they’re using computer modelling to describe and predict when resistance to antibiotics will emerge during treatment for a bacterial infection.
“There’s a pressing need for new ways to predict the emergence of antibiotic resistance during treatment to ensure the dose is right and the drug is working as effectively as possible,” says Associate Professor Cornelia Landersdorfer, co-lead author from MIPS.
Read more: The rise and fall of antibiotics. What would a post-antibiotic world look like?
She adds that the need for such modelling is not short-term, but rather something that will be required well into the future.
“While the development of new antibiotics is occurring, resistance to these drugs typically emerges quickly. This means that unfortunately we can’t rely on these medicines alone, which is where technology steps in.
“By using modelling to predict when antibiotic resistance will emerge during treatment, we can ideally stay a step ahead of the bugs and, in turn, make sure the dosing regimen is optimised to suppress resistance and increase the likelihood of successful treatment of infection.”
QSP model to predict resistance
Focusing on the antibiotic meropenem, which is administered intravenously to treat a range of serious bacterial infections such as meningitis, intra-abdominal infection, pneumonia and sepsis, the team has designed the first quantitative systems pharmacology (QSP) model to describe and predict meropenem resistance across seven strains of the deadly bacterium Pseudomonas aeruginosa, a pathogen associated with antibiotic resistance and high mortality in immunocompromised patients.
QSP uses computational models to describe interactions between a medicine and disease. In this case, the QSP model describes and predicts the full time-course of bacterial growth, bacterial killing, and emergence of antibiotic resistance across each of the seven Pseudomonas aeruginosa strains that had various pre-existing bacterial characteristics, including resistance mutations.
Read more: Antimicrobial resistance: Is phage therapy the key to unlocking the superbug crisis?
The study’s joint first author and MIPS PhD candidate Dominika Fuhs said the relationship between antibiotic, bacterial characteristics and resistance emergence during treatment is complicated, which means getting dosing regimens right can be tricky.
“Our QSP model represents a substantial advance compared with the way scientists currently attempt to measure the varying activity of antibiotics over time following administration.
“Until now, no other models have investigated meropenem regimens in the context of different resistance mechanisms using a panel of bacterial-resistant mutants,” says Fuhs.
Pseudomonas aeruginosa possesses an exceptional ability for resistance emergence during antibiotic treatment, and is among the leading pathogens causing deaths associated with antimicrobial resistance worldwide.
“The antibiotic meropenem is commonly used against Pseudomonas aeruginosa, but this is a particularly challenging organism to treat due to its extremely large armamentarium [the medicines, equipment, and techniques available to a medical practitioner] of resistance mechanisms and high capability to become resistant to all available antibiotics,” Fuhs adds.
“Traditionally, so-called pharmacokinetic/pharmacodynamic indices have been used in an attempt to link concentrations of antibiotics such as meropenem after initiating treatment with bacterial response to inform dosing.”
Traditional methods fall short
Associate Professor Landersdorfer says this traditional method does not, however, capture the full time-courses of exposure to the antibiotic, and the subsequent bacterial response.
“Furthermore, previous models don’t account for important pre-existing bacterial characteristics, including mutations, that can influence resistance emergence – particularly in bacteria such as Pseudomonas aeruginosa,” she says.
“In contrast, QSP models describe and predict the full time-courses of bacterial growth and resistance emergence.”
In the face of skyrocketing AMR around the world, employing technology to help mitigate the risk and, ultimately, decrease the exceptionally high death rates will play an increasingly important role in the ongoing management of this epic global health concern.
The study’s co-lead author, Associate Professor Antonio Oliver, from the Instituto de Investigación Sanitaria Illes Balears (IdISBa) and Hospital Son Espases, Palma de Mallorca, Spain, worked closely alongside the MIPS team, and made a significant contribution to the study by providing the bacterial strains and conducting the bioinformatics analysis, two essential components in the development of the QSP model.
This study was published in Clinical Microbiology and Infection.
