Antibiotics are among the most prescribed drugs in the world and help us recover from bacterial infection without much concern. However, as with most things, antibiotics come with a caveat. This idea of antibiotics as a double-edged sword has, for many years, focused solely on problems with antibiotic resistance and how it hinders our battle against bacterial infection. However, more recent work suggests that the other edge of this sword may be more complex than we had previously believed, with potentially significant impacts on general immunity. A better understanding of these effects could be the key to ensuring the safe and uncomplicated use of antibiotics.
A brief history of antibiotic use
The use of antibiotics to treat disease predates Alexander Fleming’s work on penicillin. Historically, the application of mouldy bread on open wounds was considered one of the more effective treatments for infection, with literature about its benefits coming from Ancient Rome, Greece, and Egypt. This concept of using compounds produced by fungi or other bacteria to treat infection carried on into the modern era.
Although not the first, Fleming’s contribution to the field of antibiotics is not to be undermined. His discovery in 1928 that penicillin, produced by the fungus Penicillium chrysogenum, could kill many types of bacteria kickstarted the golden age of antibiotics. The most commonly used classes of antibiotics were discovered during this era, which peaked in the 1950s. There was a massive effort to identify the different compounds produced by fungi and bacteria. In fact, Christian missionaries would send back soil samples from exotic places they had visited so they could be analysed for new antibiotic classes. Vancomycin, an antibiotic that targets the bacterial cell wall, was isolated in this manner from a soil sample from Borneo.
It is entirely possible that, had humans been more preoccupied by the burden of bacterial infection, advancements in cancer therapy or cardiovascular health may not have been as rapid.
During this time, the contribution of bacterial infection to global mortality plummeted. Between 1930 and 1970, human life expectancy increased by an average of 11 years. This allowed humans to shift global healthcare efforts away from infectious disease and towards more non-communicable conditions, such as cardiovascular disease. It is entirely possible that, had humans been more preoccupied by the burden of bacterial infection, advancements in fields such as vaccine development, cancer therapy, and cardiovascular health may not have been as rapid.
Over time, we began to learn that our aggressive and often irrational use of antibiotics has consequences. In 2014, more than 25% of antibiotic prescriptions in ambulatory care in the USA were for conditions where antibiotics were unnecessary. Antibiotics are commonly prescribed without laboratory testing when symptoms resemble a bacterial infection. The problem is that many viral infections also present similarly, but antibiotics are ineffective against these pathogens. As a result, many antibiotics are prescribed unnecessarily. Alexander Fleming warned in his 1945 Nobel lecture that ‘it is not difficult to make microbes resistant to penicillin in the laboratory… and the same thing has occasionally happened in the body’. Shortly after, his concerns were realised.
Selection pressure refers to anything that threatens the survival and continuity of a species, forcing them to evolve or adapt. Antibiotics are a major source of selection pressure for bacteria. The widespread introduction of antibiotics into healthcare introduced a massive selection pressure on these organisms, and as a result, resistance became a significant healthcare challenge. Antimicrobial resistance has become the most prominent concern surrounding antibiotic use and is listed as one of the World Health Organisation’s (WHO) major challenges to global health.
Antibiotic resistance may not be our only problem
In 2022, Michael Pichichero and colleagues published a study investigating the link between antibiotic use and childhood vaccine immunogenicity. Vaccine immunogenicity refers to the ability of a vaccine to create an immune response against a specific pathogen. This was a retrospective study analysing already-available data gathered for another study focusing on respiratory infections in primary care. It followed over 500 children from Rochester, New York, from the ages of 24 months to 6 years. Investigators explored the relationship between the children’s vaccination records and antibody levels. Antibodies are proteins that counteract specific pathogens. They “tag” these pathogens for destruction by immune cells and serve as “adapters” between pathogens and immune cells, allowing immune cells to recognise the pathogen and destroy it. The presence of antibodies against a specific pathogen is a marker for immunity against that pathogen. Thus, antibody levels are a reliable indicator of vaccine outcomes.
Results from the Pichichero paper showed an average reduction of 6-10% in antibody production in children who were prescribed at least one course of antibiotics in all investigated vaccines. They also reported a 12-20% reduction for all vaccine types after receiving a booster dose. Booster doses are additional doses of a vaccine given to strengthen the recipient’s immune response. A reduction of antibody levels by 12-20% is concerning as it may bring antibody levels below the protective threshold. These findings suggest the highly concerning possibility that early prescription of antibiotics in children may be driving adverse vaccine outcomes, manifesting as poorer immune protection.
We begin to learn that our aggressive and often irrational use of antibiotics has consequences.
These findings could have significant implications for public health, as poor vaccine outcomes can increase children’s susceptibility to infection throughout life. This could in turn impair herd immunity and create a risk of passing on infections to older, more vulnerable people. Herd immunity arises when a sufficient portion of the population becomes immune to an infectious disease. This reduces the probability of people lacking immunity from getting infected, as the largely immune population serves as a “buffer” against the transmission.
In the broader context of immune responses, previous work on the same data found that many children prone to upper respiratory infections, such as acute otitis media, also presented with poorer vaccine outcomes. Otitis media is an infection of the middle ear. Investigators showed that these otitis media-prone children had been prescribed multiple courses of antibiotics. They also showed evidence of several other immune impairments in the first two years of life in these children, including poor B and T cell responses. These cells mediate adaptive immunity in the body.
The adaptive immune response is specific to a particular pathogen and is therefore required for the effective and targeted destruction of invaders. Owing to the vital role of these cells in the immune response, it is unsurprising that children with poor T and B cell responses also tended to be more prone to other infections, such as influenza and pneumonia. Overall, this study presented strong evidence to suggest a link between antibiotic use and reduced immunity. This could have clinically significant manifestations, such as low vaccine responsiveness and a higher susceptibility to infection.
A potential role for the microbiome?
The microbiome is the collection of microbes (including bacteria, fungi, and viruses) that naturally live in or on our body without causing disease. Its role as a mediator of immune responses has received more attention in recent years. In addition to targeting infectious bacteria, antibiotics may also destroy some bacteria of the microbiome, upsetting its balance. This is a potential mechanism by which antibiotics may be driving these poorer immune responses.
Secondary infections such as thrush, a fungal infection that affects areas such as the mouth and vagina, are common after a course of antibiotics. This is because antibiotics cause a loss of microbiome integrity, resulting in the loss of an important mechanism for keeping other infectious agents at bay. However, results from the Pichichero paper suggest that the impact of antibiotics on the microbiome may have epidemiologically significant effects that go beyond mild fungal infections. The otitis media-prone children from the study (all of whom were prescribed multiple antibiotic courses) had an altered bacterial composition in their nasal microbiome.
Whilst the mechanisms by which the microbiome can drive immunity are still poorly understood, one possible route was identified by Bali Pulendran and his colleagues at Stanford University. They discovered that a receptor present on some immune cells in the intestine, known as macrophages, was required for detecting a protein in flagella. Flagella are the tail-like structures found on many bacterial cells, including those of the microbiome. Further research showed that activation of this receptor on macrophages, via interactions with bacteria in the microbiome, could drive macrophages to stimulate plasma cell maturation. Plasma cells are mature B cells which produce large amounts of a specific antibody. Their maturation is thus necessary for robust antibody responses. Mice lacking this flagellin receptor showed reduced antibody levels and fewer plasma cells when vaccinated with a single dose of influenza vaccine. These low antibody levels after vaccination resemble the observations of the Pichichero study. However, whether the same mechanisms are at play is far from established.
This concept needs to be further validated. We require a better understanding of the mechanisms by which the microbiome may affect other arms of the immune response. However, understanding how we may selectively target bacterial infections without disturbing the microbiome is important. Notably, Pichichero and colleagues reported that of the antibiotics considered in their study, amoxicillin was not associated with poor antibody levels following vaccination. Amoxicillin is a narrow-spectrum antibiotic, meaning it is only effective against a selected group of bacteria. Thus, it is possible that the use of selective, narrow-spectrum antibiotics in the treatment of bacterial infections may spare the microbiome and reduce the adverse impact on general immunity.
Rapid genome sequencing tools could prevent the unnecessary prescription of antibiotics and better inform which antibiotics should be used.
However, we currently lack widely accessible, rapid, and reliable diagnostic tools. Many physicians may prescribe broad-spectrum antibiotics, which are effective against a larger range of bacterial types, as a precaution while waiting for a patient’s bacterial culture results to come back. The idea of rapid diagnostics at the bedside is widely considered our best bet to combat this problem. Companies such as Oxford Nanopore have developed rapid genome sequencing tools in scalable formats that may be used in the future to allow more precise prescribing. In theory, this could allow doctors to diagnose infections during a consultation by sequencing the pathogen, identifying type of bacteria, and determining what the microbe is resistant to. This could prevent the unnecessary prescription of antibiotics, and better inform which antibiotics should be used to selectively target the pathogen. Implementing such changes on a global scale, however, will not be easy.
Are longer antibiotic courses actually better?
The common consensus is that a failure to complete a course or taking a shorter course of antibiotics, despite symptom improvement, is associated with the emergence of resistance. It is thought that shorter courses do not eradicate the entire bacterial load, and the bacteria that remain can multiply and potentially gain resistance. However, more recent work has suggested that the longer courses of antibiotics may in fact be the ones driving resistance. This is thought to be because of the severe and prolonged selection pressure they place on the bacteria. This debate has not been fully settled yet, and it may well be the case that the answer differs across bacterial strains and antibiotic types.
Evidence suggests that antimicrobial resistance may not be the only thing affected by course length. Investigators from the Pichichero study also compared the relationship between different course lengths and vaccine outcomes. They found that a 10-day course of amoxicillin and clavulanate was associated with antibody levels below protective thresholds within 30 days of course completion, while the 5-day course was not.
Evidently, there is an interesting argument that shorter courses may prove beneficial. Additional work is needed to truly establish the relationship, not only between course length and antibiotic resistance but also between course length and immune function. Should it be the case that shorter is, in fact, better, much work will have to be done to dispel current beliefs about sticking to longer courses.
Understanding the other edge of the sword
Overall, new evidence suggests that resistance may not be the only concern when it comes to antibiotics. The Pichichero paper is the first to report this relationship between antibiotics and vaccine responses in children, shedding light on some effects of potential epidemiological significance. However, it is also important to be aware of the limitations of this trial. The cohort was lacking in diversity, consisting of a predominantly white and more affluent demographic. This population is not representative of the diverse global population, and the relationship between antibiotic use and immunity may differ when other demographic factors are at play. Additionally, the authors reported a failure to collect follow-up blood samples from all the participants, meaning that they could not investigate the relationship between antibiotic use and vaccine outcomes per child, which would have provided more telling information about the study results. Clearly these findings need to be repeated in larger, more diverse cohorts with better follow-up.
Antibiotics may certainly be a double-edged sword, but with targeted research, we may be able to selectively sharpen the more beneficial edge of this sword.
Despite these shortcomings, the idea that antibiotic use may be impairing immunity and vaccine outcomes implies significant knock-on effects from a global health perspective and should not be taken lightly. Understanding the mechanisms underlying these effects and how we can create guidelines to minimise them could help us better reap the benefits of antibiotics whilst reducing the associated complications. Potential avenues of investigation include the development of rapid diagnostic tools, and a better understanding of course length. Additionally, we need to be more open to other adverse effects that any major class of drug, least of all antibiotics, can have despite their efficacy. This cautiousness is essential if we are to minimise our contribution to major global health issues later down the line. Antibiotics may certainly be a double-edged sword, but with targeted research to better understand their negative effects, we may be able to selectively sharpen the more beneficial edge of this sword.