Potential Information Loss Due to Categorization of Minimum Inhibitory Concentration Frequency Distributions

Longitudinal antimicrobial susceptibility trends of canine Staphylococcus pseudintermedius

Antimicrobial resistance within Staphylococcus pseudintermedius poses a significant risk for the treatment of canine pyoderma and as a reservoir for resistance and potential zoonoses, but few studies examine long-term temporal trends of resistance. This study assesses the antimicrobial resistance prevalence and minimum inhibitory concentration (MIC) trends in S. pseudintermedius (n=1804) isolated from canine skin samples at the Cornell University Animal Health Diagnostic Center (AHDC) between 2007 and 2020. Not susceptible (NS) prevalence, Cochran-Armitage tests, logrank tests, MIC50 and MIC90 quantiles, and survival analysis models were used to evaluate resistance prevalence and temporal trends to 23 antimicrobials. We use splines as predictors in accelerated failure time (AFT) models to model non-linear temporal trends in MICs. Multidrug resistance was common among isolates (47%), and isolates had moderate to high NS prevalence to the beta-lactams, chloramphenicol, the fluoroquinolones, gentamicin, the macrolides/lincosamides, the tetracyclines, and trimethoprim-sulfamethoxazole. However, low levels of NS to amikacin, rifampin, and vancomycin were observed. Around one third of isolates (38%) were found to be methicillin resistant S. pseudintermedius (MRSP), and these isolates had a higher prevalence of NS to all tested antimicrobials than methicillin susceptible isolates. Amongst the MRSP isolates, one phenotypically vancomycin resistant isolate (MIC >16 µg/mL) was identified, but genomic sequence data was unavailable. AFT models showed increasing MICs across time to the beta-lactams, chloramphenicol, the fluoroquinolones, gentamicin, and the macrolides/lincosamides, and decreasing temporal resistance (decreasing MICs) to doxycycline was observed amongst isolates. Notably, ATF modeling showed changes in MIC distributions that were not identified using Cochran-Armitage tests on prevalence, MIC quantiles, and logrank tests. Increasing resistance amongst these S. pseudintermedius isolates highlights the need for rational, empirical prescribing practices and increased antimicrobial resistance (AMR) surveillance to maintain the efficacy of current therapeutic agents. AFT models with non-linear predictors may be a useful, breakpoint-independent, surveillance tool alongside other modeling methods and antibiograms.

Association between Tulathromycin Treatment for Bovine Respiratory Disease and Antimicrobial Resistance Profiles among Gut Commensals and Foodborne Bacterial Pathogens Isolated from Feces of Beef Steers

ABSTRACT

This study was conducted to evaluate the association between a therapeutic dose of tulathromycin for bovine respiratory disease in beef steers and the antimicrobial and multidrug resistance profiles of the gastrointestinal tract commensals Escherichia coli and Enterococcus spp. and the foodborne pathogens Salmonella enterica and Campylobacter spp. isolated from fecal samples. Individual fecal samples were collected on days 0, 14, and 28 from 70 beef steers that were housed in a single pen and had been treated or not treated with tulathromycin. Samples were cultured for bacterial isolation, and isolates were tested for antimicrobial susceptibility with the broth microdilution method to determine the MICs of clinically relevant antimicrobials used in both human and veterinary medicine. Generalized linear mixed effects models were fitted to estimate the prevalence of the bacterial species and the prevalence of resistant isolates over time and between treated and nontreated cattle and of multidrug-resistant isolates. Model-adjusted mean prevalences of E. coli, Enterococcus spp., S. enterica, and Campylobacter spp. were 99.5, 85.9, 1.5, and 17.7%, respectively. The prevalence of erythromycin-resistant Enterococcus spp. was significantly higher on day 14 (59.7%) than on day 28 (22.2%). A higher prevalence of erythromycin-resistant Enterococcus spp. was found in samples from treated (59.3%) than in samples from nontreated (27.6%) animals. Multidrug resistance (three or more antimicrobial classes) was observed in 8.4% of E. coli isolates and 62.7% of Enterococcus isolates. The administration of tulathromycin was significantly associated with an increased prevalence of erythromycin-resistant Enterococcus spp. isolates.

HIGHLIGHTS

A Bayesian approach to modeling antimicrobial multidrug resistance

Multidrug resistance (MDR) has been a significant threat to public health and effective treatment of bacterial infections. Current identification of MDR is primarily based upon the large proportions of isolates resistant to multiple antibiotics simultaneously, and therefore is a belated evaluation. For bacteria with MDR, we expect to see strong correlations in both the quantitative minimum inhibitory concentration (MIC) and the binary susceptibility as classified by the pre-determined breakpoints. Being able to detect correlations from these two perspectives allows us to find multidrug resistant bacteria proactively. In this paper, we provide a Bayesian framework that estimates the resistance level jointly for antibiotics belonging to different classes with a Gaussian mixture model, where the correlation in the latent MIC can be inferred from the Gaussian parameters and the correlation in binary susceptibility can be inferred from the mixing weights. By augmenting the laboratory measurement with the latent MIC variable to account for the censored data, and by adopting the latent class variable to represent the MIC components, our model was shown to be accurate and robust compared with the current assessment of correlations. Applying the model to Salmonella heidelberg samples isolated from human participants in National Antimicrobial Resistance Monitoring System (NARMS) provides us with signs of joint resistance to Amoxicillin-clavulanic acid & Cephalothin and joint resistance to Ampicillin & Cephalothin. Large correlations estimated from our model could serve as a timely tool for early detection of MDR, and hence a signal for clinical intervention.

A Bayesian latent class mixture model with censoring for correlation analysis in antimicrobial resistance across populations

Background

The emergence of antimicrobial resistance across populations is a global threat to public health. Surveillance programs often monitor human and animal populations to evaluate trends of emergence in these populations. Many national level antibiotic resistance surveillance programs quantify the proportion of resistant bacteria as a means of monitoring emergence and control measures. The reason for monitoring these different populations are many, including interest in similar changes in resistance which might provide insight into emergence and control options.

Methods

In this research, we developed a method to quantify the correlation in antimicrobial resistance across populations, for the conventionally unnoticed mean shift of the susceptible bacteria. With the proposed Bayesian latent class mixture model with censoring and multivariate normal hierarchy, we address several challenges associated with analyzing the minimum inhibitory concentration data.

Results

Application of this approach to the surveillance data from National Antimicrobial Resistance Monitoring System led to a detection of positive correlation in the central tendency of azithromycin resistance of the susceptible populations from Salmonella serotype Typhimurium across food animal and human populations.

Conclusions

Our proposed approach has been shown to be accurate and superior to the commonly used naïve estimation by simulation studies. Further implementation of this Bayesian model could serve as a useful tool to indicate the co-existence of antimicrobial resistance, and potentially a need of clinical intervention.

Temporal Trends in Antimicrobial Resistance of Fecal Escherichia coli from Deer

The changing epidemiologic role of wildlife as reservoirs of antimicrobial-resistant bacteria (ARB) is poorly understood. In this study, we characterize the phenotypic resistance of commensal Escherichia coli from fecal samples of 879 individual white-tailed (Odocoileus virginianus; WTD) over a ten-year period and analyze resistance patterns. Our results show commensal E. coli from WTD had significant linear increases in reduced susceptibility to 5 of 12 antimicrobials, including broad-spectrum cephalosporins and fluoroquinolones, from 2006 to 2016. In addition, the relative frequency distribution of minimal inhibitory concentrations of two additional antimicrobials shifted towards higher values from across the study period. The prevalence of multidrug-resistant commensal E. coli increased over the study period with a prevalence of 0%, 2.2%, and 3.7% in 2006, 2012, and 2016, respectively. WTD may be persistently and increasingly exposed to antibiotics or their residues, ARB, and/or antimicrobial resistance genes via contaminated environments like surface water receiving treated wastewater effluent.

Prevalence and trend analysis of antimicrobial resistance in clinical Escherichia coli isolates collected from diseased pigs in the USA between 2006 and 2016

Antimicrobial resistance (AMR) is an emerging threat to both human and animal health. Antimicrobial use and resistance in food animal production, including swine, has received increased scrutiny as a source of resistant foodborne pathogens. Continuous surveillance of AMR in bacterial isolates of swine origin can guide in conservation of antimicrobials used in both human and swine medicine. The objective of this study was to evaluate the prevalence and trends of the phenotypic AMR in Escherichia coli of swine origin isolated from clinical samples at the Minnesota Veterinary Diagnostic laboratory between 2006 and 2016. The prevalence of resistance to ampicillin, tetracyclines and sulphadimethoxine remained greater than 50% throughout the period. There was a drastic change in enrofloxacin resistance, increasing from less than 1% to more than 20% between 2006 and 2016 (annual relative increase of 57% between 2006 and 2013 and 16% between 2013 and 2016). The prevalence of resistance to other antimicrobials remained constant (ceftiofur, oxytetracycline and chlortetracycline) or changed significantly (annual relative changes of less than 10%) for at least some time-period between 2006 and 2016 (ampicillin, florfenicol, gentamicin, neomycin, sulphadimethoxine, trimethoprim-sulphamethoxazole and spectinomycin). Rarefaction analysis revealed an increase in the number of unique combinations of AMRs per year. Network analysis was performed by estimating and plotting partial correlations between minimum inhibitory concentrations (MICs) of various antimicrobials. An increase in strength of these networks was observed, particularly in networks created after 2010, which can be indicative of increased multiple AMR in these isolates. These results provide valuable insight into the trends in AMR in E. coli of swine origin in the USA and act as supplementary information to the existing active AMR surveillance systems.

Overview of Quantitative Methodologies to Understand Antimicrobial Resistance via Minimum Inhibitory Concentration

Simple Summary

An emerging threat to human and food animal health is the development of antimicrobial resistance in bacteria associated with food animals. One of the primary tools for assessing resistance levels and monitoring for changes in expressed resistance is the use of minimum inhibitory concentration tests, which expose bacterial isolates to a series of dilutions of an antimicrobial agent to identify the lowest concentration of the antimicrobial that effectively prevents bacterial growth. These tests produce a minimum inhibitory value that falls within a range of concentrations instead of an exact value, a process known as censoring. Analysis of censored data is complex and careful consideration of methods of analysis is necessary. The use of regression methods such as logistic regression that divide the data into two or three categories is relatively easy to implement but may not detect important changes in the distributions of data that occur within the categories. Models that do not simplify the data may be more complex but may detect potentially relevant changes missed when the data is categorized. As a result, the analysis of minimum inhibitory concentration data requires careful consideration to identify the appropriate model for the purpose of the study.

Abstract

The development of antimicrobial resistance (AMR) represents a significant threat to humans and food animals. The use of antimicrobials in human and veterinary medicine may select for resistant bacteria, resulting in increased levels of AMR in these populations. As the threat presented by AMR increases, it becomes critically important to find methods for effectively interpreting minimum inhibitory concentration (MIC) tests. Currently, a wide array of techniques for analyzing these data can be found in the literature, but few guidelines for choosing among them exist. Here, we examine several quantitative techniques for analyzing the results of MIC tests and discuss and summarize various ways to model MIC data. The goal of this review is to propose important considerations for appropriate model selection given the purpose and context of the study. Approaches reviewed include mixture models, logistic regression, cumulative logistic regression, and accelerated failure time–frailty models. Important considerations in model selection include the objective of the study (e.g., modeling MIC creep vs. clinical resistance), degree of censoring in the data (e.g., heavily left/right censored vs. primarily interval censored), and consistency of testing parameters (e.g., same range of concentrations tested for a given antibiotic).

A hierarchical Bayesian latent class mixture model with censorship for detection of linear temporal changes in antibiotic resistance

Identifying and controlling the emergence of antimicrobial resistance (AMR) is a high priority for researchers and public health officials. One critical component of this control effort is timely detection of emerging or increasing resistance using surveillance programs. Currently, detection of temporal changes in AMR relies mainly on analysis of the proportion of resistant isolates based on the dichotomization of minimum inhibitory concentration (MIC) values. In our work, we developed a hierarchical Bayesian latent class mixture model that incorporates a linear trend for the mean log2MIC of the non-resistant population. By introducing latent variables, our model addressed the challenges associated with the AMR MIC values, compensating for the censored nature of the MIC observations as well as the mixed components indicated by the censored MIC distributions. Inclusion of linear regression with time as a covariate in the hierarchical structure allowed modelling of the linear creep of the mean log2MIC in the non-resistant population. The hierarchical Bayesian model was accurate and robust as assessed in simulation studies. The proposed approach was illustrated using Salmonella enterica I,4,[5],12:i:- treated with chloramphenicol and ceftiofur in human and veterinary samples, revealing some significant linearly increasing patterns from the applications. Implementation of our approach to the analysis of an AMR MIC dataset would provide surveillance programs with a more complete picture of the changes in AMR over years by exploring the patterns of the mean resistance level in the non-resistant population. Our model could therefore serve as a timely indicator of a need for antibiotic intervention before an outbreak of resistance, highlighting the relevance of this work for public health. Currently, however, due to extreme right censoring on the MIC data, this approach has limited utility for tracking changes in the resistant population.

Variation in fluoroquinolone pharmacodynamic parameter values among isolates of two bacterial pathogens of bovine respiratory disease

To design an antimicrobial treatment regimen for a bacterial disease, data on the drug pharmacodynamics (PD) against selected drug-susceptible strains of the pathogen are used. The regimen is applied across such strains in the field, assuming the PD parameter values remain the same. We used time-kill experiments and PD modeling to investigate the fluoroquinolone enrofloxacin PD against different isolates of two bovine respiratory disease pathogens: four Mannheimia haemolytica and three Pasteurella multocida isolates. The models were fitted as mixed-effects non-linear regression; the fixed-effects PD parameter values were estimated after accounting for random variation among experimental replicates. There was both inter- and intra- bacterial species variability in the PD parameters Hill-coefficient and E_max (maximal decline of bacterial growth rate), with more variable PD responses among M. haemolytica than among P. multocida isolates. Moreover, the Hill-coefficient was correlated to the isolate’s maximal population growth rate in the absence of antimicrobial exposure (a.k.a. specific growth rate; Spearman’s ρ = 0.98, p-value = 0.003, n = 6 isolates excluding one outlier). Thus, the strain’s properties such as growth potential may impact its PD responses. This variability can have clinical implications. Modifying the treatment regimen depending on phenotypic properties of the pathogen strain causing disease may be a precision medicine approach.

Establishing Statistical Equivalence of Data from Different Sampling Approaches for Assessment of Bacterial Phenotypic Antimicrobial Resistance

To assess phenotypic bacterial antimicrobial resistance (AMR) in different strata (e.g., host populations, environmental areas, manure, or sewage effluents) for epidemiological purposes, isolates of target bacteria can be obtained from a stratum using various sample types. Also, different sample processing methods can be applied. The MIC of each target antimicrobial drug for each isolate is measured. Statistical equivalence testing of the MIC data for the isolates allows evaluation of whether different sample types or sample processing methods yield equivalent estimates of the bacterial antimicrobial susceptibility in the stratum. We demonstrate this approach on the antimicrobial susceptibility estimates for (i) nontyphoidal Salmonella spp. from ground or trimmed meat versus cecal content samples of cattle in processing plants in 2013-2014 and (ii) nontyphoidal Salmonella spp. from urine, fecal, and blood human samples in 2015 (U.S. National Antimicrobial Resistance Monitoring System data). We found that the sample types for cattle yielded nonequivalent susceptibility estimates for several antimicrobial drug classes and thus may gauge distinct subpopulations of salmonellae. The quinolone and fluoroquinolone susceptibility estimates for nontyphoidal salmonellae from human blood are nonequivalent to those from urine or feces, conjecturally due to the fluoroquinolone (ciprofloxacin) use to treat infections caused by nontyphoidal salmonellae. We also demonstrate statistical equivalence testing for comparing sample processing methods for fecal samples (culturing one versus multiple aliquots per sample) to assess AMR in fecal Escherichia coli These methods yield equivalent results, except for tetracyclines. Importantly, statistical equivalence testing provides the MIC difference at which the data from two sample types or sample processing methods differ statistically. Data users (e.g., microbiologists and epidemiologists) may then interpret practical relevance of the difference.IMPORTANCE Bacterial antimicrobial resistance (AMR) needs to be assessed in different populations or strata for the purposes of surveillance and determination of the efficacy of interventions to halt AMR dissemination. To assess phenotypic antimicrobial susceptibility, isolates of target bacteria can be obtained from a stratum using different sample types or employing different sample processing methods in the laboratory. The MIC of each target antimicrobial drug for each of the isolates is measured, yielding the MIC distribution across the isolates from each sample type or sample processing method. We describe statistical equivalence testing for the MIC data for evaluating whether two sample types or sample processing methods yield equivalent estimates of the bacterial phenotypic antimicrobial susceptibility in the stratum. This includes estimating the MIC difference at which the data from the two approaches differ statistically. Data users (e.g., microbiologists, epidemiologists, and public health professionals) can then interpret whether that present difference is practically relevant.