Stream 5

Addressing other important, but neglected, questions about antibiotic resistance using systematic reviews of the literature and primary studies

Systematic reviews needed

The problem: Several questions about antibiotic resistance have not yet been explored by researchers anywhere in the world, yet these issues are potentially very important and could influence national and international practice and policy. These include questions such as:

5.1 Does use of two antibiotics simultaneously result in less resistance? This is common practice in tuberculosis treatment, where antibiotic resistance has been a serious problem since 1948, and indeed was only resolved by the use of multiple antibiotics. However this practice has fallen out of use for common infections in modern times and evidence on whether this is beneficial to recommence is needed.

5.2 Which antibiotics are least likely to cause resistance in specific pathogens? It is common knowledge that some organisms remain sensitive to some antibiotics (e.g. Streptococcus pneumoniae to penicillin) while others (e.g. Staphylococcus aureus) soon develop resistance to many antibiotics. This has not been quantified in the literature, yet data on this would inform clinical decisions for common infections.

5.3 Which alternative (symptomatic) treatments to antibiotics are effective for specific infections? The importance of having synthesised evidence in this area is to assist clinicians to fill the ‘therapeutic vacuum’ left after not using antibiotics. The literature is insufficiently systematically reviewed to allow quantitative comparisons of these treatments with each other, and, especially with antibiotics.

The solution: We will undertake the systematic reviews that provide the best quantification of the evidence for each of the above questions. We have developed a large set of the scientific literature from an earlier systematic search which examines resistance persistence to antibiotics in primary care, which contains data that can address both questions 5.1 and 5.2 under the leadership of CIs Del Mar and Hoffmann. We have also completed the search and screening of a large literature set to address the questions within 5.3 under the leadership of CI Beller.

The problemWe have discovered that insufficient primary studies have been undertaken to inform a systematic review we are currently completing, (paper in preparation, PROSPERO registration ID 25499). This started as an update of a previous systematic review conducted in the UK, which addresses the very important question about how long resistance persists in the microbiome after a usual course of antibiotics in the community. We were unable to replicate the methods, and have discovered that there are major concerns about the usual interpretation of their results. Our results show that for many antibiotics there is insufficient certainty about this question. Past primary study methods have not used modern methods to reduce the uncertainty.

The solution: We will undertake a prospective cohort study to estimate the rate of decay of resistance in the nasal microbiome after a usual course of antibiotics in the community.

Our method will be to invite a convenience sample of GPs to provide kits to 120 patients for whom they prescribe a single course of antibiotics, and take the first swab prior to antibiotics. The kits will contain 5 further swabs, including:

1) postage replied envelopes (5) for sending self-administered nasal swabs to our laboratory at 3-monthly intervals for one year after finishing the course of antibiotics, and 1 at 2 years;

2) brief questionnaires (5) to accompany the swabs, to document which antibiotic used, any other antibiotics used, any hospital admissions.

(Hence swabs occur at 0, 3, 6, 9, 12, and 24 months; time 0 will give us a pre-antibiotic baseline)

Patients will be eligible if they have not used antibiotics, or had a hospital admission in the past year; and are able and willing to provide informed consent for swabbing the anterior nares and post them, together with the survey form every 3 months. We have successfully tested the transmission of nasal swabs through the mail in the GAPS study.37 SMS text reminders and phone calls will be used to remind patients to supply their next swab, and this will be done by the research assistant.

The swabs will be subject to whole gene sequencing. The primary outcomes are antibiotic gene probes using DNA amplification methods (PCR) for Staphylococcus aureus multi-resistance (Multiple Resistance Staph aureus, MRSA); beta-lactamase genes carried by Haemophilus influenzae, and by a pneumococcus. The analysis will include:

1) changes of antibiotic resistance with time, including any influences of new antibiotics, the effect of combinations of antibiotics used (if any), and hospital admission on the decay of antibiotic resistance;

2) changes in the microbiome with time as a proportion of common bacteria living as commensals.

The budget request in this application includes the above genes, however we will seek external funding to upgrade to a full 84-resistance-gene analysis. The study is aimed at quantifying antibiotic resistance genes (ARGs) sourced from the presence of bacterial communities using nasal swabs. The temporal assessment of genes will be undertaken following genomic DNA (gDNA) extraction of every swab using the DNA Tissue EZ1 kits (Qiagen). Before extraction of the DNA, samples will be pre-treated with NALC lysis of the microbes with the addition of lysozyme before adding proteinase K final digestion cell wall agents. The amplification runs will be performed using the EZ1 Corbett machine (Qiagen) and the respective EZ1 bacteria Card, both tools available in our centre. Following the nucleic acid extraction, each gDNA (per swab) will be subject to a quantitative Polymerase Chain Reaction (qPCR) using unique pairs of specific microbials of many ARGs primers capable of annealing to template DNA sample. In time, multiple swabs will be provided per individual and therefore will enable to characterize the temporal profile of changes of each ARG’s expression. The expression levels of each ARG will be made available and importantly these data would have been previously normalized against any differences of starting amounts of DNA (from swab to swab). This normalization is made possible using multiple housekeeping genes where expressions are directly proportional to sample input and therefore represent a reference amplification assay within each array and will be used principally to compare the relative gene profiling between swabs of the same individual. This method of relative quantification relies on cycle threshold (cycle value for which an exponential growth of the gene amplicon occurs) and is commonly referred to as ΔΔCT = ΔCTgroup 2 – ΔCTgroup 1. In addition, the fold change in expression can be subsequently calculated as 2^–ΔCTgroup 2 / 2^–ΔCTgroup 1 (either increased or decreased values are possible and a 5 to 10-fold difference between groups in relative abundance are considered significant). Of note, these arrays also harbour a pan-bacteria assay that detects a broad spectrum of bacterial species, and additionally a Positive PCR Control assay is embedded to determine the efficiency of DNA amplification from swab to swab (possible presence of PCR inhibitors in the extracted DNA material).