Antimicrobial resistance (AMR) is one of the leading public health threats of our time. For the past century or so, antibiotics like penicillin have turned previously dangerous infections into manageable conditions. However, as microbes become resistant to the drugs we use against them, diseases that once plagued humanity could become a danger once again. Tuberculosis (TB) is a particularly stark example. In 2017, more than half a million people developed multi-drug resistant TB. The disease couldn’t be treated with the two most powerful drugs, and the resulting mortality rate among those patients was nearly 50%. In sub-Saharan Africa, where there are 230 million cases of the mosquito-borne disease malaria, there are growing fears of a drug resistant strain spreading. Even in HIV – a condition considered to be manageable nowadays – some antiretrovirals used in treatment are becoming less effective.

Other examples of antimicrobial resistance include the so-called ‘super gonorrhea’ (drug-resistant gonorrhea) presenting a new sexual health threat, and the several ‘superbugs’ in hospitals already considered a notorious menace.

Today, 700,000 people die a year from drug-resistant diseases – and that figure is almost certain to climb. The UN estimates that, by 2050, AMR could contribute to ten million deaths a year, the bulk of which will be in lower-income countries. Against this backdrop, it is abundantly clear that we shouldn’t be overusing antibiotics. The more a drug is used, the greater the odds that the pathogen being targeted – along with other bacteria in the body – will develop drug-resistant properties. According to the US Centers for Disease Control and Prevention (CDC), between one-third and half of all antibiotic use in humans is unnecessary or inappropriate – adding fuel to a fast-growing fire.

If we want to dampen the blaze, we need widespread antimicrobial susceptibility testing (AST), a laboratory procedure designed to identify which drug is appropriate for which patient. “Antibiotic susceptibility testing allows you to ensure that you’re using the most appropriate antibiotic for that infection,” says Professor Matthew Avison, senior lecturer in microbiology at the University of Bristol. “At the moment, clinicians tend to err on the side of newer antibiotics, which are where rates of resistance are likely to be lower. Unfortunately, the more they use them, the more rates of resistance will go up. We want to encourage them to use the older ones where possible.” Although these older antibiotics might prove ineffective, AST should give you a definite answer either way. If they are in fact going to work, you should be able to prescribe them with confidence, keeping the newer ones in reserve. “Alternatively, when it’s not going to work, you should then use the thing that will work,” says Avison. “Those are the two main reasons why we would use a susceptibility test – to ensure the patient gets something that’s going to work and to ensure that the latest antibiotics aren’t wasted when they aren’t necessary.”

Time is of the essence

Professor Alex van Belkum, global director of microbiology research at diagnostics company bioMérieux, points out that resistance flourishes the moment that you treat your patients with the wrong combination of drugs. “If you use antimicrobials that are too broad spectrum, then you may be wasting them by applying them in large numbers to patients who could profit from some of the more focused therapeutic approaches,” he says. “Aside from that, AST can also be used to do epidemiological studies and monitor the resistance of bugs.” AST, then, is an important tool in our armoury against AMR. But unfortunately, the procedure hasn’t yet lived up to its billing.

The AST process takes around 24–48 hours on average, meaning the patient will often have begun a course of antibiotics – and not always the right ones – before the result comes back. “The systems that are currently available do what they should do, but they take way too long,” says van Belkum.

“If you could do an antimicrobial susceptibility test like you do a urine test or haemoglobin test, in half an hour, you would be able to provide your patient with exactly the right antimicrobial targeted towards their specific infection.”

It’s for that reason that researchers have been turning their focus to rapid AST – tests that could be performed almost at the point of care, or at the very least, in time to guide that patient’s course of treatment. If the lab could return results that same day, rather than a day or two later, GPs would be able to take advantage of ‘delayed prescribing’ and quash inappropriate prescriptions.

“By the time the patient goes to get the prescription filled that evening, or maybe the next morning, the chemist will have been given the information through the GP’s system,” says Avison. “The patient will receive the drug that actually is necessary, rather than just the best guess.” Further down the line, the test might be performed by the GP practice itself, with no need to send the patient’s sample to a lab. “Potentially, the patient could sit in the waiting room and then leave with the correct prescription,” says Avison. “But that's a little way away, for all sorts of commercial and other reasons.”

New approaches

Under existing methods, an antimicrobial is added to a living bacterial culture to see whether it continues to grow. Lab technicians work with a suspension culture that includes hundreds of thousands of cells and need to go through multiple replications – a time-consuming process. As a result, researchers are on the hunt for other approaches that don’t rely on detecting changes in bacterial populations.

“What has happened over the past decade or so is that there are a couple of single cell approaches that are very promising,” says van Belkum. “If you are capable of fishing out one cell from a clinical specimen, then you only need to visualise what happens to that cell in order to define whether or not it is susceptible to the antimicrobial.”

US company Accelerate Diagnostics was the first to commercialise such a technology. Its Accelerate Pheno system immobilises individual cells on a piece of glass, and traces them over time in the presence or absence of an antibiotic. Results arrive within about seven hours. But even this could be too long a wait in some cases, and several diagnostics startups are attempting to cut the time down to minutes, not hours. Swedish startup Astrego Diagnostics uses microfluidics and image analysis methods to perform AST in less than 30 minutes. Similarly, UK-based Vitamica, a spin-off from the University of Bristol, is developing a rapid AST that detects fluctuations within the individual bacterium. “Using a microscope, you look at bacteria in a sample and add an antibiotic,” says Avison, who sits on Vitamica’s scientific advisory board. “This microscope shines an evanescent field onto the bacteria, and you see the bacteria shimmer and wiggle. I call it a wiggle-ometer. If the antibiotics are having an effect on the bacteria, that wiggling slows down, and if the bacteria are resistant to the antibiotic, the wiggling carries on. You see that effect within just a few minutes of the antibiotic being present.”

There are a few more speculative technologies on the horizon too, involving next-generation sequencing, metabolomics and transcriptomics. With the costs of sequencing falling every year, these kinds of approaches are starting to become more feasible. “If you know an organism’s entire genome, then in principle you can deduce whether certain resistance genes or susceptibility pathways are present,” says van Belkum. “That will give you an opportunity to do a computer search for those resistance markers, which will then tell you what sort of antimicrobials can still be used and for which ones this bug is probably resistant.”

The bigger picture

Both Avison and van Belkum, while optimistic about the potential for rapid AST, see a number of obstacles to implementing them in healthcare systems. Avison points out that, while there is definitely a desire within the health service to speed up AST, it can be hard to get labs to agree on what that looks like in practice. Van Belkum adds that since the current systems are cheap, automated and high-throughput, labs need to be convinced that the newer systems are just as good. “There are these promising rapid systems that do their thing in about 20 minutes,” he says. “But if you want to put an instrument like that into a clinical laboratory, then it needs to be capable of what the current automated systems are capable of – which is tens of antimicrobials, multiple dilutions, for a vast array of different microbial species.” He warns too that many new technologies, such as sequencing, won’t be practicable in developing countries. AMR is a global problem, and as such, it will be important to focus on technologies that are suitable for resource-poor settings.

“What is technically feasible, and what we can afford here in Western countries, is important because it sets the stage of what can be done in the end,” he says. “But AMR is not a local issue. You should always put your technologies into an international framework, where that type of technology is available to all, because otherwise, it’s not going to solve the problems that we’re facing.”

Avison believes that, despite some of the challenges around implementation, we will start to see rapid AST using visualisation methods within the next five years, and genome sequencing approaches in the 2030s. “I think the next five to ten years is when we’re going to see a sea change in the way we do susceptibility testing,” he says. “At the moment, it’s pretty much the same as we would have done it 50 years ago, apart from a few more bells and whistles.” However, he remarks that the introduction of new technologies won’t automatically lead to a change in the way antibiotics are prescribed. It will fall to clinicians to use this rapid AST judiciously, with a view to making an impact on antibiotic resistance. “I think we need to perhaps do some work around encouraging practice, rather than just assuming the technology will dig us out of a hole,” he says.

50%

As much as half of all antibiotic use in humans is unnecessary or inappropriate.

CDC

1–2

Number of days the antimicrobial susceptibility testing process takes on average.

bioMérieux