Logo: to the web site of Uppsala University

Background

Heteroresistance (HR) is a significant type of antibiotic resistance observed for several bacterial species and antibiotic classes where a susceptible main population contains small subpopulations of resistant cells. Mathematical models, animal experiments and clinical studies associate HR with treatment failure. Currently used susceptibility tests do not detect heteroresistance reliably, which can result in misclassification of heteroresistant isolates as susceptible which might lead to treatment failure. Here we examined if whole genome sequence (WGS) data and machine learning (ML) can be used to detect bacterial HR.

Methods

We classified 467 Escherichia coli clinical isolates as HR or non-HR to the often used β-lactam/inhibitor combination piperacillin-tazobactam using pre-screening and Population Analysis Profiling tests. We sequenced the isolates, assembled the whole genomes and created a set of predictors based on current knowledge of HR mechanisms. Then we trained several machine learning models on 80% of this data set aiming to detect HR isolates. We compared performance of the best ML models on the remaining 20% of the data set with a baseline model based solely on the presence of β-lactamase genes. Furthermore, we sequenced the resistant sub-populations in order to analyse the genetic mechanisms underlying HR.

Findings

The best ML model achieved 100% sensitivity and 84.6% specificity, outperforming the baseline model. The strongest predictors of HR were the total number of β-lactamase genes, β-lactamase gene variants and presence of IS elements flanking them. Genetic analysis of HR strains confirmed that HR is caused by an increased copy number of resistance genes via gene amplification or plasmid copy number increase. This aligns with the ML model's findings, reinforcing the hypothesis that this mechanism underlies HR in Gram-negative bacteria.

Interpretation

We demonstrate that a combination of WGS and ML can identify HR in bacteria with perfect sensitivity and high specificity. This improved detection would allow for better-informed treatment decisions and potentially reduce the occurrence of treatment failures associated with HR.

Background

Antibiotic heteroresistance is a common bacterial phenotype characterised by the presence of small resistant subpopulations within a susceptible population. During antibiotic exposure, these resistant subpopulations can be enriched and potentially lead to treatment failure. In this study, we examined the prevalence, misclassification, and clinical effect of heteroresistance in Escherichia coli bloodstream infections for the clinically important antibiotics cefotaxime, gentamicin, and piperacillin–tazobactam.

Methods

We conducted a retrospective cohort analysis of patients (n=255) admitted to in-patient care and treated for E coli bloodstream infections within the Uppsala region in Sweden between Jan 1, 2014, and Dec 31, 2015. Patient inclusion criteria were admission to hospital on suspicion of infection, starting systemic antibiotics at the time of admission, positive blood cultures for the growth of E coli upon admission, and residency in the Uppsala health-care region at the time of admission. Exclusion criteria were growth of an additional pathogen than E coli in blood cultures taken at admission or previous inclusion of the patients in the study for another bloodstream infection. Antibiotic susceptibility of preserved blood culture isolates of E coli was assessed for cefotaxime, gentamicin, and piperacillin–tazobactam by disk diffusion and breakpoint crossing heteroresistance (BCHR) was identified using population analysis profiling. The clinical outcome parameters were obtained from patient records. The primary outcome variable was length of hospital stay due to the E coli bloodstream infection, defined as the time between admission and discharge from inpatient care as noted on the physician’s notes. Secondary outcomes were time to fever resolution, admission to intermediary care unit or intensive care unit during time in hospital, switching or adding another intravenous antibiotic treatment, re-admission to hospital within 30 days of original admission, recurrent E coli infection within 30 days of admission to hospital, and all-cause mortality within 90 days of admission.

Findings

A total of 255 participants with a corresponding E coli isolate (out of 500 screened for eligibility) met the inclusion criteria, with 135 female patients and 120 male patients. One (<1%) of 255 strains was BCHR for cefotaxime, 109 (43%) of 255 strains were BCHR for gentamicin, and 22 (9%) of 255 strains were BCHR for piperacillin–tazobactam. Clinical susceptibility testing misclassified 120 (96%) of 125 heteroresistant bacterial strains as susceptible. The BCHR phenotypes had no correlation to length of hospital stay due to the E coli bloodstream infection. However, patients with piperacillin–tazobactam BCHR strains who received piperacillin–tazobactam had 3·1 times higher odds for admittance to the intermediate care unit (95% CI 1·1–9·6, p=0·041) than the remainder of the cohort, excluding those treated with gentamicin. Similarly, those infected with gentamicin BCHR who received gentamicin showed higher odds for admittance to the intensive care unit (5·6 [1·1–42·0, p=0·043]) and mortality (7·1 [1·2–49·2, p=0·030]) than patients treated with gentamicin who were infected with non-gentamicin BCHR E coli.

Interpretation

In a cohort of patients with E coli bloodstream infections, heteroresistance is common and frequently misidentified in routine clinical testing. Several negative effects on patient outcomes are associated with heteroresistant strains.

Plasmids are the primary vectors of horizontal transfer of antibiotic resistance genes among bacteria. Previous studies have shown that the spread and maintenance of plasmids among bacterial populations depend on the genetic makeup of both the plasmid and the host bacterium. Antibiotic resistance can also be acquired through mutations in the bacterial chromosome, which not only confer resistance but also result in changes in bacterial physiology and typically a reduction in fitness. However, it is unclear whether chromosomal resistance mutations affect the interaction between plasmids and the host bacteria. To address this question, we introduced 13 clinical plasmids into a susceptible Escherichia coli strain and three different congenic mutants that were resistant to nitrofurantoin (Delta nfsAB), ciprofloxacin (gyrA, S83L), and streptomycin (rpsL, K42N) and determined how the plasmids affected the exponential growth rates of the host in glucose minimal media. We find that though plasmids confer costs on the susceptible strains, those costs are fully mitigated in the three resistant mutants. In several cases, this results in a competitive advantage of the resistant strains over the susceptible strain when both carry the same plasmid and are grown in the absence of antibiotics. Our results suggest that bacteria carrying chromosomal mutations for antibiotic resistance could be a better reservoir for resistance plasmids, thereby driving the evolution of multi-drug resistance. IMPORTANCE Plasmids have led to the rampant spread of antibiotic resistance genes globally. Plasmids often carry antibiotic resistance genes and other genes needed for its maintenance and spread, which typically confer a fitness cost on the host cell observed as a reduced growth rate. Resistance is also acquired via chromosomal mutations, and similar to plasmids they also reduce bacterial fitness. However, we do not know whether resistance mutations affect the bacterial ability to carry plasmids. Here, we introduced 13 multi-resistant clinical plasmids into a susceptible and three different resistant E. coli strains and found that most of these plasmids do confer fitness cost on susceptible cells, but these costs disappear in the resistant strains which often lead to fitness advantage for the resistant strains in the absence of antibiotic selection. Our results imply that already resistant bacteria are a more favorable reservoir for multi-resistant plasmids, promoting the ascendance of multi-resistant bacteria. Plasmids have led to the rampant spread of antibiotic resistance genes globally. Plasmids often carry antibiotic resistance genes and other genes needed for its maintenance and spread, which typically confer a fitness cost on the host cell observed as a reduced growth rate. Resistance is also acquired via chromosomal mutations, and similar to plasmids they also reduce bacterial fitness. However, we do not know whether resistance mutations affect the bacterial ability to carry plasmids. Here, we introduced 13 multi-resistant clinical plasmids into a susceptible and three different resistant E. coli strains and found that most of these plasmids do confer fitness cost on susceptible cells, but these costs disappear in the resistant strains which often lead to fitness advantage for the resistant strains in the absence of antibiotic selection. Our results imply that already resistant bacteria are a more favorable reservoir for multi-resistant plasmids, promoting the ascendance of multi-resistant bacteria.

Antibiotic heteroresistance is a phenotype in which a susceptible bacterial population includes a small subpopulation of cells that are more resistant than the main population. Such resistance can arise by tandem amplification of DNA regions containing resistance genes that in single copy are not sufficient to confer resistance. However, tandem amplifications often carry fitness costs, manifested as reduced growth rates. Here, we investigated if and how these fitness costs can be genetically ameliorated. We evolved four clinical isolates of three bacterial species that show heteroresistance to tobramycin, gentamicin and tetracyclines at increasing antibiotic concentrations above the minimal inhibitory concentration (MIC) of the main susceptible population. This led to a rapid enrichment of resistant cells with up to an 80-fold increase in the resistance gene copy number, an increased MIC, and severely reduced growth rates. When further evolved in the presence of antibiotic, these strains acquired compensatory resistance mutations and showed a reduction in copy number while maintaining high-level resistance. A deterministic model indicated that the loss of amplified units was driven mainly by their fitness costs and that the compensatory mutations did not affect the loss rate of the gene amplifications. Our findings suggest that heteroresistance mediated by copy number changes can facilitate and precede the evolution towards stable resistance. Antibiotic heteroresistance, in which a susceptible bacterial population includes a small resistant subpopulation, can arise by tandem amplification of resistance genes, which often carry fitness costs. Here, Pal and Andersson show that these fitness costs can be ameliorated by the acquisition of compensatory mutations and a reduction in copy number of the resistance genes.

Background

Piperacillin/tazobactam is a β-lactam/β-lactamase inhibitor combination with a broad spectrum of activity that is often used as empirical and/or targeted therapy among hospitalized patients. Heteroresistance (HR) is a form of antibiotic resistance in which a minority population of resistant cells coexists with a majority susceptible population that has been found to be a cause of antibiotic treatment failure in murine models.

Objectives

To determine the prevalence of HR and mechanisms of HR to piperacillin/tazobactam among Klebsiella pneumoniae bloodstream infection (BSI) isolates.

Materials

From July 2018 to June 2021, K. pneumoniae piperacillin/tazobactam-susceptible BSI isolates were collected from two tertiary hospitals in Atlanta, GA, USA. Only first isolates from each patient per calendar year were included. Population analysis profiling (PAP) and WGS were performed to identify HR and its mechanisms.

Results

Among 423 K. pneumoniae BSI isolates collected during the study period, 6% (25/423) were found to be HR with a subpopulation surviving above the breakpoint. WGS of HR isolates grown in the presence of piperacillin/tazobactam at concentrations 8-fold that of the MIC revealed copy number changes of plasmid-located β-lactamase genes bla_CTX-M-15, bla_SHV33, bla_OXA-1 and bla_TEM-1 by tandem gene amplification or plasmid copy number increase.

Conclusions

Prevalence of HR to piperacillin/tazobactam among bloodstream isolates was substantial. The HR phenotype appears to be caused by tandem amplification of β-lactamase genes found on plasmids or plasmid copy number increase. This raises the possibility of dissemination of HR through horizontal gene transfer and requires further study.

Heteroresistance (HR) is an enigmatic phenotype where, in a main population of susceptible cells, small subpopulations of resistant cells exist. This is a cause for concern, as this small subpopulation is difficult to detect by standard antibiotic susceptibility tests, and upon antibiotic exposure the resistant subpopulation may increase in frequency and potentially lead to treatment complications or failure. Here, we determined the prevalence and mechanisms of HR for 40 clinical Staphylococcus aureus isolates, against 6 clinically important antibiotics: daptomycin, gentamicin, linezolid, oxacillin, teicoplanin, and vancomycin. High frequencies of HR were observed for gentamicin (69.2%), oxacillin (27%), daptomycin (25.6%), and teicoplanin (15.4%) while none of the isolates showed HR toward linezolid or vancomycin. Point mutations in various chromosomal core genes, including those involved in membrane and peptidoglycan/teichoic acid biosynthesis and transport, tRNA charging, menaquinone and chorismite biosynthesis and cyclic-di-AMP biosynthesis, were the mechanisms responsible for generating the resistant subpopulations. This finding is in contrast to gram-negative bacteria, where increased copy number of bona fide resistance genes via tandem gene amplification is the most prevalent mechanism. This difference can be explained by the observation that S. aureus has a low content of resistance genes and absence of the repeat sequences that allow tandem gene amplification of these genes as compared to gram-negative species.

Co-administering a low dose of colistin (CST) with ciprofloxacin (CIP) may improve the antibacterial effect against resistant Escherichia coli, offering an acceptable benefit-risk balance. This study aimed to quantify the interaction between ciprofloxacin and colistin in an in silico pharmacokinetic-pharmacodynamic model from in vitro static time-kill experiments (using strains with minimum inhibitory concentrations, MIC_CIP 0.023–1 mg/L and MIC_CST 0.5–0.75 mg/L). It was also sought to demonstrate an approach of simulating concentrations at the site of infection with population pharmacokinetic and whole-body physiologically based pharmacokinetic models to explore the clinical value of the combination when facing more resistant strains (using extrapolated strains with lower susceptibility). The combined effect in the final model was described as the sum of individual drug effects with a change in drug potency: for ciprofloxacin, concentration at half maximum killing rate (EC₅₀) in combination was 160% of the EC₅₀ in monodrug experiments, while for colistin, the change in EC₅₀ was strain-dependent from 54.1% to 119%. The benefit of co-administrating a lower-than-commonly-administrated colistin dose with ciprofloxacin in terms of drug effect in comparison to either monotherapy was predicted in simulated bloodstream infections and pyelonephritis. The study illustrates the value of pharmacokinetic-pharmacodynamic modelling and simulation in streamlining rational development of antibiotic combinations.