Smart combinations of antibiotics can slow down resistance
Antibiotic resistance
When a bacterium becomes resistant to one antibiotic, it may sometimes become more sensitive to another. This biological side-effect offers an unexpected opportunity in the fight against antibiotic resistance.
‘The idea is that we can use this phenomenon to develop better dosing schedules for stubborn infections,’ explains researcher Coen van Hasselt. ‘By switching wisely between antibiotics or combining them, you can use this biological effect – known as collateral sensitivity – to reduce the chance that bacteria become resistant, and prevent treatments from failing.’
From laboratory to clinical data: how researchers mapped antibiotics and bacteria
The phenomenon has long been known from laboratory work, but there has been little clinical research so far. ‘Many studies rely on bacteria grown in the lab,’ says Van Hasselt. ‘Those differ greatly from the bacteria you find in patients.’ This led the Leiden team to a fundamental question: do we also see these effects in real pathogens taken from patients? PhD candidate Sebastian Tandar is exploring that question in his PhD research.
What hundreds of thousands of bacterial strains reveal about antibiotic resistance
Since 2008, the Dutch National Institute for Public Health and the Environment (RIVM) has been collecting data on bacteria that frequently cause infections, such as E. coli, S. aureus and S. pneumoniae. The data come from almost every medical microbiology laboratory in the Netherlands. For this study, the Leiden researchers worked closely with the RIVM’s Centre for Epidemiology and Surveillance of Infectious Diseases, which supplied data from the Infectious Diseases Surveillance Information System – Antibiotic Resistance (ISIS-AR).
‘We can track trends in resistance and identify new developments.’
The database contains test results from samples taken across all layers of the healthcare system – from GPs to university hospitals – and from all kinds of bodily materials. This creates a complete picture of how sensitive different bacteria are to a wide range of antibiotics. Annelot Schoffelen of the RIVM, who contributed to the study, explains: ‘Because these data are collected in the same way every year, we can track trends in resistance and identify new developments.’
‘We combined the RIVM data with an international dataset,’ says Van Hasselt. Using a statistical method developed by Laura Zwep (LACDR), the team analysed data from hundreds of thousands of bacterial strains to find patterns in which resistance to one drug goes hand in hand with increased sensitivity to another.
‘This study is a strong starting point for investigating the hypothesis of collateral sensitivity further.’
Can new antibiotic combinations slow down resistance?
The analysis yielded a promising picture: some combinations appeared across different bacterial species, suggesting a broadly applicable effect. ‘A few combinations had already been seen in the lab, but we also found new pairs,’ says Van Hasselt.
To make the results accessible, the team developed a web application that allows researchers to explore the data themselves. The RIVM not only provided the data but also supported the interpretation and possible clinical relevance. Schoffelen: ‘This study is a strong starting point for investigating the hypothesis of collateral sensitivity further.’
From pattern to practice: what these insights could mean for treatments
The RIVM stresses the importance of this kind of research in the global fight against antibiotic resistance. ‘In the Netherlands, the problem is still relatively limited thanks to infection prevention and policy,’ the institute says. ‘But certain forms of resistance are slowly increasing, and difficult-to-treat bacteria are being imported. Research into new treatment strategies is therefore essential.’
‘You often work from lab to patient. This time, we went the other way round – and it turned out to be surprisingly fruitful.’
The next step is to investigate how such combinations can be used in practice. ‘We are working with computer models and infection models in the lab to understand how to fine-tune dosing,’ says Van Hasselt. ‘Ultimately, we want to know whether this approach can truly slow down resistance in patients.’
The study, recently published [link], shows that patient data are useful not only for tracking resistance but also for finding new treatment strategies. ‘You often work from lab to patient,’ Van Hasselt concludes. ‘This time, we went the other way round – and it turned out to be surprisingly fruitful.’