Species Identification and Antibiotic Resistance Prediction by Analysis of Whole-Genome Sequence Data by Use of ARESdb: an Analysis of Isolates from the Unyvero Lower Respiratory Tract Infection Trial

Keeping up with the pathogens: improved antimicrobial resistance detection and prediction from Pseudomonas aeruginosa genomes

BACKGROUND

Antimicrobial resistance (AMR) is an intensifying threat that requires urgent mitigation to avoid a post-antibiotic era. Pseudomonas aeruginosa represents one of the greatest AMR concerns due to increasing multi- and pan-drug resistance rates. Shotgun sequencing is gaining traction for in silico AMR profiling due to its unambiguity and transferability; however, accurate and comprehensive AMR prediction from P. aeruginosa genomes remains an unsolved problem.

METHODS

We first curated the most comprehensive database yet of known P. aeruginosa AMR variants. Next, we performed comparative genomics and microbial genome-wide association study analysis across a Global isolate Dataset (n = 1877) with paired antimicrobial phenotype and genomic data to identify novel AMR variants. Finally, the performance of our P. aeruginosa AMR database, implemented in our AMR detection and prediction tool, ARDaP, was compared with three previously published in silico AMR gene detection or phenotype prediction tools-abritAMR, AMRFinderPlus, ResFinder-across both the Global Dataset and an analysis-naïve Validation Dataset (n = 102).

RESULTS

Our AMR database comprises 3639 mobile AMR genes and 728 chromosomal variants, including 75 previously unreported chromosomal AMR variants, 10 variants associated with unusual antimicrobial susceptibility, and 281 chromosomal variants that we show are unlikely to confer AMR. Our pipeline achieved a genotype-phenotype balanced accuracy (bACC) of 85% and 81% across 10 clinically relevant antibiotics when tested against the Global and Validation Datasets, respectively, vs. just 56% and 54% with abritAMR, 58% and 54% with AMRFinderPlus, and 60% and 53% with ResFinder. ARDaP's superior performance was predominantly due to the inclusion of chromosomal AMR variants, which are generally not identified with most AMR identification tools.

CONCLUSIONS

Our ARDaP software and associated AMR variant database provides an accurate tool for predicting AMR phenotypes in P. aeruginosa, far surpassing the performance of current tools. Implementation of ARDaP for routine AMR prediction from P. aeruginosa genomes and metagenomes will improve AMR identification, addressing a critical facet in combatting this treatment-refractory pathogen. However, knowledge gaps remain in our understanding of the P. aeruginosa resistome, particularly the basis of colistin AMR.

Is Carbapenem Therapy Necessary for the Treatment of Non-CTX-M Extended-Spectrum β-Lactamase-Producing Enterobacterales Bloodstream Infections?

BACKGROUND

Investigations into antibiotics for extended-spectrum β-lactamase-producing Enterobacterales (ESBL-E) bloodstream infections (BSIs) have focused on blaCTX-M genes. Patient outcomes from non-CTX-M-producing ESBL-E BSIs and optimal treatment are unknown.

METHODS

A multicenter observational study investigating 500 consecutive patients with ceftriaxone-resistant Enterobacterales BSIs during 2018-2022 was conducted. Broth microdilution and whole-genome sequencing confirmed antibiotic susceptibilities and ESBL gene presence, respectively. Inverse probability weighting (IPW) using propensity scores ensured patients with non-CTX-M and CTX-M ESBL-E BSIs were similar before outcome evaluation.

RESULTS

396 patients (79.2%) were confirmed to have an ESBL-E BSI. ESBL gene family prevalence was as follows: blaCTX-M (n = 370), blaSHV (n = 16), blaOXY (n = 12), and blaVEB (n = 5). ESBL gene identification was not limited to Escherichia coli and Klebsiella species. In the IPW cohort, there was no difference in 30-day mortality or ESBL-E infection recurrence between the non-CTX-M and CTX-M groups (odds ratio [OR], 0.99; 95% confidence interval [CI], .87-1.11; P = .83 and OR, 1.10; 95% CI, .85-1.42; P = .47, respectively). In an exploratory analysis limited to the non-CTX-M group, 86% of the 21 patients who received meropenem were alive on day 30; none of the 5 patients who received piperacillin-tazobactam were alive on day 30.

CONCLUSIONS

Our findings suggest that non-CTX-M and CTX-M ESBL-E BSIs are equally concerning and associated with similar clinical outcomes. Meropenem may be associated with improved survival in patients with non-CTX-M ESBL-E BSIs, underscoring the potential benefit of comprehensive molecular diagnostics to enable early antibiotic optimization for ESBL-E BSIs beyond just blaCTX-M genes.

Navigating the network: a narrative overview of AMR surveillance and data flow in the United States

The antimicrobial resistance (AMR) surveillance landscape in the United States consists of a data flow that starts in the clinical setting and is maintained by a network of national and state public health laboratories. These organizations are well established, with robust methodologies to test and confirm antimicrobial susceptibility. Still, the bridge that guides the flow of data is often one directional and caught in a constant state of rush hour that can only be refined with improvements to infrastructure and automation in the data flow. Moreover, there is an absence of information in the literature explaining the processes clinical laboratories use to coalesce and share susceptibility test data for AMR surveillance, further complicated by variability in testing procedures. This knowledge gap limits our understanding of what is needed to improve and streamline data sharing from clinical to public health laboratories. Successful models of AMR surveillance display attributes like 2-way communication between clinical and public health laboratories, centralized databases, standardized data, and the use of electronic health records or data systems, highlighting areas of opportunity and improvement. This article explores the roles and processes of the organizations involved in AMR surveillance in the United States and identifies current knowledge gaps and opportunities to improve communication between them through standardization, communication, and modernization of data flow.

Rapid nanopore sequencing and predictive susceptibility testing of positive blood cultures from intensive care patients with sepsis

We aimed to evaluate the performance of Oxford Nanopore Technologies (ONT) sequencing from positive blood culture (BC) broths for bacterial identification and antimicrobial susceptibility prediction. Patients with suspected sepsis in four intensive care units were prospectively enrolled. Human-depleted DNA was extracted from positive BC broths and sequenced using ONT (MinION). Species abundance was estimated using Kraken2, and a cloud-based system (AREScloud) provided in silico predictive antimicrobial susceptibility testing (AST) from assembled contigs. Results were compared to conventional identification and phenotypic AST. Species-level agreement between conventional methods and AST predicted from sequencing was 94.2% (49/52), increasing to 100% in monomicrobial infections. In 262 high-quality AREScloud AST predictions across 24 samples, categorical agreement (CA) was 89.3%, with major error (ME) and very major error (VME) rates of 10.5% and 12.1%, respectively. Over 90% CA was achieved for some taxa (e.g., Staphylococcus aureus) but was suboptimal for Pseudomonas aeruginosa. In 470 AST predictions across 42 samples, with both high quality and exploratory-only predictions, overall CA, ME, and VME rates were 87.7%, 8.3%, and 28.4%. VME rates were inflated by false susceptibility calls in a small number of species/antibiotic combinations with few representative resistant isolates. Time to reporting from sequencing could be achieved within 8-16 h from BC positivity. Direct sequencing from positive BC broths is feasible and can provide accurate predictive AST for some species. ONT-based approaches may be faster but significant improvements in accuracy are required before it can be considered for clinical use.IMPORTANCESepsis and bloodstream infections carry a high risk of morbidity and mortality. Rapid identification and susceptibility prediction of causative pathogens, using Nanopore sequencing direct from blood cultures, may offer clinical benefit. We assessed this approach in comparison to conventional phenotypic methods and determined the accuracy of species identification and susceptibility prediction from genomic data. While this workflow holds promise, and performed well for some common bacterial species, improvements in sequencing accuracy and more robust predictive algorithms across a diverse range of organisms are required before this can be considered for clinical use. However, results could be achieved in timeframes that are faster than conventional phenotypic methods.

Scoping review of risk-scoring tools for early prediction of bloodstream infections caused by carbapenem-resistant Enterobacterales: do we really have a reliable risk-scoring tool?

Background

Bloodstream infections (BSIs) caused by carbapenem-resistant Enterobacterales (CRE) are a global health concern. Rapid identification of CRE may improve patient outcomes and reduce inappropriate antibiotic prescription. The use of risk-scoring tools (RSTs) can be valuable for optimizing the decision-making process for empirical antibiotic therapy of suspected CRE bacteraemia. These tools can also be used to triage use of expensive rapid diagnostic methods.

Methods

We systematically reviewed the relevant literature in PubMed/MEDLINE, CINAHL, Cochrane, Web of Science, Embase and Scopus up to 1 November 2022 to identify RSTs that predict CRE BSIs. The literature review and analysis of the articles were performed by two researchers; any inconsistencies were resolved through discussion.

Results

We identified 9 RSTs developed for early prediction of CRE BSIs and only logistic regression was used for most studies. These RSTs were quite different from each other in terms of their performance and the variables they included. They also had notable limitations and very few of them were externally validated.

Conclusions

RSTs for early prediction of CRE BSIs have limitations and lack of external validity outside the local setting in which they were developed. Future studies to identify optimal RSTs in high and low CRE-endemic settings are warranted. Approaches based on rapid diagnostics and RSTs should be compared with a treatment approach using both methods in a randomized controlled trial.

Clinical Diagnostics of Bacterial Infections and Their Resistance to Antibiotics-Current State and Whole Genome Sequencing Implementation Perspectives

Antimicrobial resistance (AMR), defined as the ability of microorganisms to withstand antimicrobial treatment, is responsible for millions of deaths annually. The rapid spread of AMR across continents warrants systematic changes in healthcare routines and protocols. One of the fundamental issues with AMR spread is the lack of rapid diagnostic tools for pathogen identification and AMR detection. Resistance profile identification often depends on pathogen culturing and thus may last up to several days. This contributes to the misuse of antibiotics for viral infection, the use of inappropriate antibiotics, the overuse of broad-spectrum antibiotics, or delayed infection treatment. Current DNA sequencing technologies offer the potential to develop rapid infection and AMR diagnostic tools that can provide information in a few hours rather than days. However, these techniques commonly require advanced bioinformatics knowledge and, at present, are not suited for routine lab use. In this review, we give an overview of the AMR burden on healthcare, describe current pathogen identification and AMR screening methods, and provide perspectives on how DNA sequencing may be used for rapid diagnostics. Additionally, we discuss the common steps used for DNA data analysis, currently available pipelines, and tools for analysis. Direct, culture-independent sequencing has the potential to complement current culture-based methods in routine clinical settings. However, there is a need for a minimum set of standards in terms of evaluating the results generated. Additionally, we discuss the use of machine learning algorithms regarding pathogen phenotype detection (resistance/susceptibility to an antibiotic).

Machine-Learning Model for Prediction of Cefepime Susceptibility in Escherichia coli from Whole-Genome Sequencing Data

The declining cost of performing bacterial whole-genome sequencing (WGS) coupled with the availability of large libraries of sequence data for well-characterized isolates have enabled the application of machine-learning (ML) methods to the development of nonlinear sequence-based predictive models. We tested the ML-based model developed by Next Gen Diagnostics for prediction of cefepime phenotypic susceptibility results in Escherichia coli. A cohort of 100 isolates of E. coli recovered from urine (n = 77) and blood (n = 23) cultures were used. The cefepime MIC was determined in triplicate by reference broth microdilution and classified as susceptible (MIC of ≤2 μg/mL) or not susceptible (MIC of ≥4 μg/mL) using the 2022 Clinical and Laboratory Standards Institute breakpoints. Five isolates generated both susceptible and not susceptible MIC results, yielding categorical agreement of 95% for the reference method to itself. Categorical agreement of ML to MIC interpretations was 97%, with 2 very major (false, susceptible) and 1 major (false, not susceptible) errors. One very major error occurred for an isolate with blaCTX-M-27 (MIC mode, ≥32 μg/mL) and one for an isolate with blaTEM-34 for which the MIC cefepime mode was 4 μg/mL. One major error was for an isolate with blaCTX-M-27 but with a MIC mode of 2 μg/mL. These preliminary data demonstrated performance of ML for a clinically important antimicrobial-species pair at a caliber similar to phenotypic methods, encouraging wider development of sequence-based susceptibility prediction and its validation and use in clinical practice.

Artificial Intelligence for Antimicrobial Resistance Prediction: Challenges and Opportunities towards Practical Implementation

Antimicrobial resistance (AMR) is emerging as a potential threat to many lives worldwide. It is very important to understand and apply effective strategies to counter the impact of AMR and its mutation from a medical treatment point of view. The intersection of artificial intelligence (AI), especially deep learning/machine learning, has led to a new direction in antimicrobial identification. Furthermore, presently, the availability of huge amounts of data from multiple sources has made it more effective to use these artificial intelligence techniques to identify interesting insights into AMR genes such as new genes, mutations, drug identification, conditions favorable to spread, and so on. Therefore, this paper presents a review of state-of-the-art challenges and opportunities. These include interesting input features posing challenges in use, state-of-the-art deep-learning/machine-learning models for robustness and high accuracy, challenges, and prospects to apply these techniques for practical purposes. The paper concludes with the encouragement to apply AI to the AMR sector with the intention of practical diagnosis and treatment, since presently most studies are at early stages with minimal application in the practice of diagnosis and treatment of disease.

Metagenomic Antimicrobial Susceptibility Testing from Simulated Native Patient Samples

Genomic antimicrobial susceptibility testing (AST) has been shown to be accurate for many pathogens and antimicrobials. However, these methods have not been systematically evaluated for clinical metagenomic data. We investigate the performance of in-silico AST from clinical metagenomes (MG-AST). Using isolate sequencing data from a multi-center study on antimicrobial resistance (AMR) as well as shotgun-sequenced septic urine samples, we simulate over 2000 complicated urinary tract infection (cUTI) metagenomes with known resistance phenotype to 5 antimicrobials. Applying rule-based and machine learning-based genomic AST classifiers, we explore the impact of sequencing depth and technology, metagenome complexity, and bioinformatics processing approaches on AST accuracy. By using an optimized metagenomics assembly and binning workflow, MG-AST achieved balanced accuracy within 5.1% of isolate-derived genomic AST. For poly-microbial infections, taxonomic sample complexity and relatedness of taxa in the sample is a key factor influencing metagenomic binning and downstream MG-AST accuracy. We show that the reassignment of putative plasmid contigs by their predicted host range and investigation of whole resistome capabilities improved MG-AST performance on poly-microbial samples. We further demonstrate that machine learning-based methods enable MG-AST with superior accuracy compared to rule-based approaches on simulated native patient samples.

Validation and Application of Long-Read Whole-Genome Sequencing for Antimicrobial Resistance Gene Detection and Antimicrobial Susceptibility Testing

Next-generation sequencing applications are increasingly used for detection and characterization of antimicrobial-resistant pathogens in clinical settings. Oxford Nanopore Technologies (ONT) sequencing offers advantages for clinical use compared with other sequencing methodologies because it enables real-time basecalling, produces long sequencing reads that increase the ability to correctly assemble DNA fragments, provides short turnaround times, and requires relatively uncomplicated sample preparation. A drawback of ONT sequencing, however, is its lower per-read accuracy than short-read sequencing. We sought to identify best practices in ONT sequencing protocols. As some variability in sequencing results may be introduced by the DNA extraction methodology, we tested three DNA extraction kits across three independent laboratories using a representative set of six bacterial isolates to investigate accuracy and reproducibility of ONT technology. All DNA extraction techniques showed comparable performance; however, the DNeasy PowerSoil Pro kit had the highest sequencing yield. This kit was subsequently applied to 42 sequentially collected bacterial isolates from blood cultures to assess Ares Genetics's pipelines for predictive whole-genome sequencing antimicrobial susceptibility testing (WGS-AST) performance compared to phenotypic triplicate broth microdilution results. WGS-AST results ranged across the organisms and resulted in an overall categorical agreement of 95% for penicillins, 82.4% for cephalosporins, 76.7% for carbapenems, 86.9% for fluoroquinolones, and 96.2% for aminoglycosides. Very major errors/major errors were 0%/16.7% (penicillins), 11.7%/3.6% (cephalosporins), 0%/24.4% (carbapenems), 2.5%/7.7% (fluoroquinolones), and 0%/4.1% (aminoglycosides), respectively. This work showed that, although additional refinements are necessary, ONT sequencing demonstrates potential as a method to perform WGS-AST on cultured isolates for patient care.

Core Genome Multilocus Sequence Typing and Antibiotic Susceptibility Prediction from Whole-Genome Sequence Data of Multidrug-Resistant Pseudomonas aeruginosa Isolates

Over the past decade, whole-genome sequencing (WGS) has overtaken traditional bacterial typing methods for studies of genetic relatedness. Further, WGS data generated during epidemiologic studies can be used in other clinically relevant bioinformatic applications, such as antibiotic resistance prediction. Using commercially available software tools, the relatedness of 38 clinical isolates of multidrug-resistant Pseudomonas aeruginosa was defined by two core genome multilocus sequence typing (cgMLST) methods, and the WGS data of each isolate was analyzed to predict antibiotic susceptibility to nine antibacterial agents. The WGS typing and resistance prediction data were compared with pulsed-field gel electrophoresis (PFGE) and phenotypic antibiotic susceptibility results, respectively. Simpson's Diversity Index and adjusted Wallace pairwise assessments of the three typing methods showed nearly identical discriminatory power. Antibiotic resistance prediction using a trained analytical pipeline examined 342 bacterial-drug combinations with an overall categorical agreement of 92.4% and very major, major, and minor error rates of 3.6, 4.1, and 4.1%, respectively. IMPORTANCE Multidrug-resistant Pseudomonas aeruginosa isolates are a serious public health concern due to their resistance to nearly all or all of the available antibiotics, including carbapenems. Utilizing molecular approaches in conjunction with antibiotic susceptibility prediction software warrants investigation for use in the clinical laboratory workflow. These molecular tools coupled with antibiotic resistance prediction tools offer the opportunity to overcome the extended turnaround time and technical challenges of phenotypic susceptibility testing.

Optimized Method for Bacterial Nucleic Acid Extraction from Positive Blood Culture Broth for Whole-Genome Sequencing, Resistance Phenotype Prediction, and Downstream Molecular Applications

The application of direct metagenomic sequencing from positive blood culture broth may solve the challenges of sequencing from low-bacterial-load blood samples in patients with sepsis. Forty prospectively collected blood culture broth samples growing Gram-negative bacteria were extracted using commercially available kits to achieve high-quality DNA.

Genomic Analysis of a mcr-9.1-Harbouring IncHI2-ST1 Plasmid from Enterobacter ludwigii Isolated in Fish Farming

This study analyzed the resistome, virulome and mobilome of an MCR-9-producing Enterobacter sp. identified in a muscle sample of seabream (Sparus aurata), collected in a land tank from multitrophic fish farming production. Average Nucleotide Identity analysis identified INSAq77 at the species level as an Enterobacter ludwigii INSAq77 strain that was resistant to chloramphenicol, florfenicol and fosfomycin and was susceptible to all other antibiotics tested. In silico antimicrobial resistance analyses revealed genes conferring in silico resistance to β-lactams (bla_ACT-88), chloramphenicol (catA4-type), fosfomycin (fosA2-type) and colistin (mcr-9.1), as well as several efflux pumps (e.g., oqxAB-type and mar operon). Further bioinformatics analysis revealed five plasmid replicon types, including the IncHI2/HI2A, which are linked to the worldwide dissemination of the mcr-9 gene in different antibiotic resistance reservoirs. The conserved nickel/copper operon rcnR-rcnA-pcoE-ISSgsp1-pcoS-IS903-mcr-9-wbuC was present, which may play a key role in copper tolerance under anaerobic growth and nickel homeostasis. These results highlight that antibiotic resistance in aquaculture are spreading through food, the environment and humans, which places this research in a One Health context. In fact, colistin is used as a last resort for the treatment of serious infections in clinical settings, thus mcr genes may represent a serious threat to human health.

Acinetobacter baumannii Genomic Sequence-Based Core Genome Multilocus Sequence Typing Using Ridom SeqSphere+ and Antimicrobial Susceptibility Prediction in ARESdb

Whole-genome sequencing (WGS) is rapidly replacing traditional typing methods for the investigation of infectious disease outbreaks. Additionally, WGS data are being used to predict phenotypic antimicrobial susceptibility. Acinetobacter baumannii, which is often multidrug-resistant, is a significant culprit in outbreaks in health care settings. A well-characterized collection of A. baumannii was studied using core genome multilocus sequence typing (cgMLST). Seventy-two isolates previously typed by PCR-electrospray ionization mass spectrometry (PCR/ESI-MS) provided by the Antimicrobial Resistance Leadership Group (ARLG) were analyzed using a clinical microbiology laboratory developed workflow for cgMLST with genomic susceptibility prediction performed using the ARESdb platform. Previously performed PCR/ESI-MS correlated with cgMLST using relatedness thresholds of allelic differences of ≤9 and ≤200 allelic differences in 78 and 94% of isolates, respectively. Categorical agreement between genotypic and phenotypic antimicrobial susceptibility across a panel of 11 commonly used drugs was 89%, with minor, major, and very major error rates of 8%, 11%, and 1%, respectively.

Automated antimicrobial susceptibility testing and antimicrobial resistance genotyping using Illumina and Oxford Nanopore Technologies sequencing data among Enterobacteriaceae

Whole-genome sequencing (WGS) enables the molecular characterization of bacterial pathogens. We compared the accuracy of the Illumina and Oxford Nanopore Technologies (ONT) sequencing platforms for the determination of AMR classes and antimicrobial susceptibility testing (AST) among 181 clinical Enterobacteriaceae isolates. Sequencing reads for each isolate were uploaded to AREScloud (Ares Genetics) to determine the presence of AMR markers and the predicted WGS-AST profile. The profiles of both sequencing platforms were compared to broth microdilution (BMD) AST. Isolates were delineated by resistance to third-generation cephalosporins and carbapenems as well as the presence of AMR markers to determine clinically relevant AMR classes. The overall categorical agreement (CA) was 90% (Illumina) and 88% (ONT) across all antimicrobials, 96% for the prediction of resistance to third-generation cephalosporins for both platforms, and 94% (Illumina) and 91% (ONT) for the prediction of resistance to carbapenems. Carbapenem resistance was overestimated on ONT with a major error of 16%. Sensitivity for the detection of carbapenemases, extended-spectrum β-lactamases, and plasmid-mediated ampC genes was 98, 95, and 70% by ONT compared to the Illumina dataset as the reference. Our results highlight the potential of the ONT platform’s use in clinical microbiology laboratories. When combined with robust bioinformatics methods, WGS-AST predictions may be a future approach to guide effective antimicrobial decision-making.

Proceedings of the Clinical Microbiology Open 2018 and 2019 - a Discussion about Emerging Trends, Challenges, and the Future of Clinical Microbiology

Clinical Microbiology Open (CMO), a meeting supported by the American Society for Microbiology's Clinical and Public Health Microbiology Committee (CPHMC) and Corporate Council, provides a unique interactive platform for leaders from diagnostic microbiology laboratories, industry, and federal agencies to discuss the current and future state of the clinical microbiology laboratory. The purpose is to leverage the group's diverse views and expertise to address critical challenges, and discuss potential collaborative opportunities for diagnostic microbiology, through the utilization of varied resources. The first and second CMO meetings were held in 2018 and 2019, respectively. Discussions were focused on the diagnostic potential of innovative technologies and laboratory diagnostic stewardship, including expansion of next-generation sequencing into clinical diagnostics, improvement and advancement of molecular diagnostics, emerging diagnostics, including rapid antimicrobial susceptibility and point of care testing (POCT), harnessing big data through artificial intelligence, and staffing in the clinical microbiology laboratory. Shortly after CMO 2019, the coronavirus disease 2019 (COVID-19) pandemic further highlighted the need for the diagnostic microbiology community to work together to utilize and expand on resources to respond to the pandemic. The issues, challenges, and potential collaborative efforts discussed during the past two CMO meetings proved critical in addressing the COVID-19 response by diagnostic laboratories, industry partners, and federal organizations. Planning for a third CMO (CMO 2022) is underway and will transition from a discussion-based meeting to an action-based meeting. The primary focus will be to reflect on the lessons learned from the COVID-19 pandemic and better prepare for future pandemics.

Submarine Outfalls of Treated Wastewater Effluents are Sources of Extensively- and Multidrug-Resistant KPC- and OXA-48-Producing Enterobacteriaceae in Coastal Marine Environment

The rapid and ongoing spread of carbapenemase-producing Enterobacteriaceae has led to a global health threat. However, a limited number of studies have addressed this problem in the marine environment. We investigated their emergence in the coastal waters of the central Adriatic Sea (Croatia), which are recipients of submarine effluents from two wastewater treatment plants. Fifteen KPC-producing Enterobacteriaceae (nine Escherichia coli, four Klebsiella pneumoniae and two Citrobacter freundii) were recovered, and susceptibility testing to 14 antimicrobials from 10 classes showed that four isolates were extensively drug resistant (XDR) and two were resistant to colistin. After ERIC and BOX-PCR typing, eight isolates were selected for whole genome sequencing. The E. coli isolates belonged to serotype O21:H27 and sequence type (ST) 2795, while K. pneumoniae isolates were assigned to STs 37 and 534. Large-scale genome analysis revealed an arsenal of 137 genes conferring resistance to 19 antimicrobial drug classes, 35 genes associated with virulence, and 20 plasmid replicons. The isolates simultaneously carried 43–90 genes encoding for antibiotic resistance, while four isolates co-harbored carbapenemase genes blaKPC-2 and blaOXA-48. The blaOXA-48 was associated with IncL-type plasmids in E. coli and K. pneumoniae. Importantly, the blaKPC-2 in four E. coli isolates was located on ~40 kb IncP6 broad-host-range plasmids which recently emerged as blaKPC-2 vesicles, providing first report of these blaKPC-2-bearing resistance plasmids circulating in E. coli in Europe. This study also represents the first evidence of XDR and potentially virulent strains of KPC-producing E. coli in coastal waters and the co-occurrence of blaKPC-2 and blaOXA-48 carbapenemase genes in this species. The leakage of these strains through submarine effluents into coastal waters is of concern, indicating a reservoir of this infectious threat in the marine environment.

Comparative Analysis and Data Provenance for 1,113 Bacterial Genome Assemblies

The availability of public genomics data has become essential for modern life sciences research, yet the quality, traceability, and curation of these data have significant impacts on a broad range of microbial genomics research. While microbial genome databases such as NCBI’s RefSeq database leverage the scalability of crowd sourcing for growth, genomics data provenance and authenticity of the source materials used to produce data are not strict requirements. Here, we describe the de novo assembly of 1,113 bacterial genome references produced from authenticated materials sourced from the American Type Culture Collection (ATCC), each with full genomics data provenance relating to bioinformatics methods, quality control, and passage history. Comparative genomics analysis of ATCC standard reference genomes (ASRGs) revealed significant issues with regard to NCBI’s RefSeq bacterial genome assemblies related to completeness, mutations, structure, strain metadata, and gaps in traceability to the original biological source materials. Nearly half of RefSeq assemblies lack details on sample source information, sequencing technology, or bioinformatics methods. Deep curation of these records is not within the scope of NCBI’s core mission in supporting open science, which aims to collect sequence records that are submitted by the public. Nonetheless, we propose that gaps in metadata accuracy and data provenance represent an “elephant in the room” for microbial genomics research. Effectively addressing these issues will require raising the level of accountability for data depositors and acknowledging the need for higher expectations of quality among the researchers whose research depends on accurate and attributable reference genome data.

IMPORTANCE The traceability of microbial genomics data to authenticated physical biological materials is not a requirement for depositing these data into public genome databases. This creates significant risks for the reliability and data provenance of these important genomics research resources, the impact of which is not well understood. We sought to investigate this by carrying out a comparative genomics study of 1,113 ATCC standard reference genomes (ASRGs) produced by ATCC from authenticated and traceable materials using the latest sequencing technologies. We found widespread discrepancies in genome assembly quality, genetic variability, and the quality and completeness of the associated metadata among hundreds of reference genomes for ATCC strains found in NCBI’s RefSeq database. We present a comparative analysis of de novo-assembled ASRGs, their respective metadata, and variant analysis using RefSeq genomes as a reference. Although assembly quality in RefSeq has generally improved over time, we found that significant quality issues remain, especially as related to genomic data and metadata provenance. Our work highlights the importance of data authentication and provenance for the microbial genomics community, and underscores the risks of ignoring this issue in the future.

Defining Baseline Mechanisms of Cefiderocol Resistance in the Enterobacterales

The objective of this study was to identify putative mechanisms contributing to baseline cefiderocol resistance among carbapenem-resistant Enterobacterales (CRE). We evaluated 56 clinical CRE isolates with no previous exposure to cefiderocol. Cefiderocol and comparator agent minimum inhibitory concentrations (MICs) were determined by broth microdilution. Short-read and/or long-read whole genome sequencing was pursued. Cefiderocol nonwild type (NWT; i.e., MICs ≥4 mg/L) CRE were compared with species-specific reference genomes and with cefiderocol wild type (WT) CRE isolates to identify genes or missense mutations, potentially contributing to elevated cefiderocol MICs. A total of 14 (25%) CRE isolates met cefiderocol NWT criteria. Of the 14 NWT isolates, various β-lactamases (e.g., carbapenemases in Klebsiella pneumoniae and AmpC β-lactamases in Enterobacter cloacae complex) in combination with permeability defects were associated with a ≥ 80% positive predictive value in identifying NWT isolates. Unique mutations in the sensor kinase gene baeS were identified among NWT isolates. Cefiderocol NWT isolates were more likely to be resistant to colistin than WT isolates (29% vs. 0%). Our findings suggest that no consistent antimicrobial resistance markers contribute to baseline cefiderocol resistance in CRE isolates and, rather, cefiderocol resistance results from a combination of heterogeneous mechanisms.

Strengths and caveats of identifying resistance genes from whole genome sequencing data

INTRODUCTION

Antimicrobial resistance (AMR) continues to present major challenges to modern healthcare. Recent advances in whole-genome sequencing (WGS) have made the rapid molecular characterization of AMR a realistic possibility for diagnostic laboratories; yet major barriers to clinical implementation exist.

AREAS COVERED

We describe and compare short- and long-read sequencing platforms, typical components of bioinformatics pipelines, tools for AMR gene detection and the relative merits of read- or assembly-based approaches. The challenges of characterizing mobile genetic elements from genomic data are outlined, as well as the complexities inherent to the prediction of phenotypic resistance from WGS. Practical obstacles to implementation in diagnostic laboratories, the critical role of quality control and external quality assurance, as well as standardized reporting standards are also discussed. Future directions, such as the application of machine-learning and artificial intelligence algorithms, linked to clinically meaningful outcomes, may offer a new paradigm for the clinical application of AMR prediction.

EXPERT OPINION

AMR prediction from WGS data presents an exciting opportunity to advance our capacity to comprehensively characterize infectious pathogens in a rapid manner, ultimately aiming to improve patient outcomes. Collaborative efforts between clinicians, scientists, regulatory bodies and healthcare administrators will be critical to achieve the full promise of this approach.

Genome-Wide Mutation Scoring for Machine-Learning-Based Antimicrobial Resistance Prediction

The prediction of antimicrobial resistance (AMR) based on genomic information can improve patient outcomes. Genetic mechanisms have been shown to explain AMR with accuracies in line with standard microbiology laboratory testing. To translate genetic mechanisms into phenotypic AMR, machine learning has been successfully applied. AMR machine learning models typically use nucleotide k-mer counts to represent genomic sequences. While k-mer representation efficiently captures sequence variation, it also results in high-dimensional and sparse data. With limited training data available, achieving acceptable model performance or model interpretability is challenging. In this study, we explore the utility of feature engineering with several biologically relevant signals. We propose to predict the functional impact of observed mutations with PROVEAN to use the predicted impact as a new feature for each protein in an organism’s proteome. The addition of the new features was tested on a total of 19,521 isolates across nine clinically relevant pathogens and 30 different antibiotics. The new features significantly improved the predictive performance of trained AMR models for Pseudomonas aeruginosa, Citrobacter freundii, and Escherichia coli. The balanced accuracy of the respective models of those three pathogens improved by 6.0% on average.

Artificial Intelligence and Mapping a New Direction in Laboratory Medicine: A Review

BACKGROUND

Modern artificial intelligence (AI) and machine learning (ML) methods are now capable of completing tasks with performance characteristics that are comparable to those of expert human operators. As a result, many areas throughout healthcare are incorporating these technologies, including in vitro diagnostics and, more broadly, laboratory medicine. However, there are limited literature reviews of the landscape, likely future, and challenges of the application of AI/ML in laboratory medicine.

CONTENT

In this review, we begin with a brief introduction to AI and its subfield of ML. The ensuing sections describe ML systems that are currently in clinical laboratory practice or are being proposed for such use in recent literature, ML systems that use laboratory data outside the clinical laboratory, challenges to the adoption of ML, and future opportunities for ML in laboratory medicine.

SUMMARY

AI and ML have and will continue to influence the practice and scope of laboratory medicine dramatically. This has been made possible by advancements in modern computing and the widespread digitization of health information. These technologies are being rapidly developed and described, but in comparison, their implementation thus far has been modest. To spur the implementation of reliable and sophisticated ML-based technologies, we need to establish best practices further and improve our information system and communication infrastructure. The participation of the clinical laboratory community is essential to ensure that laboratory data are sufficiently available and incorporated conscientiously into robust, safe, and clinically effective ML-supported clinical diagnostics.

A personalised approach to antibiotic pharmacokinetics and pharmacodynamics in critically ill patients

Critically ill patients admitted to intensive care unit (ICU) with severe infections, or those who develop nosocomial infections, have poor outcomes with substantial morbidity and mortality. Such patients commonly have suboptimal antibiotic exposures at routinely used antibiotic doses related to an increased volume of distribution and altered clearance due to their underlying altered physiology. Furthermore, the use of extracorporeal devices such as renal replacement therapy and extracorporeal membrane oxygenation in these group of patients also has the potential to alter in vivo drug concentrations. Moreover, ICU patients are likely to be infected with less-susceptible pathogens. Therefore, one potential contributing cause to the poor outcomes observed in critically ill patients may be related to subtherapeutic antibiotic exposures. Newer concepts include the clinician considering optimised dosing based on a blood antibiotic exposure defined by pharmacokinetic modelling and therapeutic drug monitoring, combined with a knowledge of the antibiotic penetration into the site of infection, thereby achieving optimal bacterial killing. Such optimised dosing is likely to improve patient outcomes. The aim of this review is to highlight key aspects of antibiotic pharmacokinetics and pharmacodynamics (PK/PD) in critically ill patients and provide a PK/PD approach to tailor antibiotic dosing to the individual patient.

Core Genome Multilocus Sequence Typing and Prediction of Antimicrobial Susceptibility Using Whole-Genome Sequences of Escherichia coli Bloodstream Infection Isolates

In total, 50 Escherichia coli bloodstream isolates from the clinical laboratory and 12 E. coli isolates referred for pulsed-field gel electrophoresis (PFGE) were sequenced, assessed for clonality using core genome multilocus sequence typing (cgMLST), and evaluated for genomic susceptibility predictions using ARESdb. Results of sequence typing using whole-genome sequencing (WGS)-based MLST and sequence type (ST)-specific PCR were identical. Overall categorical agreement between genotypic (ARESdb) and phenotypic susceptibility testing for 62 isolates and 11 antimicrobial agents was 91%. Among the referred isolates, high major error rates were found for ceftazidime, cefepime, and piperacillin-tazobactam.

A Combined Phenotypic-Genotypic Predictive Algorithm for In Vitro Detection of Bicarbonate: β-Lactam Sensitization among Methicillin-Resistant Staphylococcus aureus (MRSA)

Antimicrobial susceptibility testing (AST) is routinely used to establish predictive antibiotic resistance metrics to guide the treatment of bacterial pathogens. Recently, a novel phenotype termed "bicarbonate (NaHCO₃)-responsiveness" was identified in a relatively high frequency of clinical MRSA strains, wherein isolates demonstrate in vitro "susceptibility" to standard β-lactams (oxacillin [OXA]; cefazolin [CFZ]) in the presence of NaHCO₃, and in vivo susceptibility to these β-lactams in experimental endocarditis models. We investigated whether a targeted phenotypic-genotypic screening of MRSA could rule in or rule out NaHCO₃ susceptibility upfront. We studied 30 well-characterized clinical MRSA bloodstream isolates, including 15 MIC-susceptible to CFZ and OXA in NaHCO₃-supplemented Mueller-Hinton Broth (MHB); and 15 MIC-resistant to both β-lactams in this media. Using a two-tiered strategy, isolates were first screened by standard disk diffusion for susceptibility to a combination of amoxicillin-clavulanate [AMC]. Isolates then underwent genomic sequence typing: MLST (clonal complex [CC]); agr; SCCmec; and mecA promoter and coding region. The combination of AMC disk susceptibility testing plus mecA and spa genotyping was able to predict MRSA strains that were more or less likely to be NaHCO₃-responsive in vitro, with a high degree of sensitivity and specificity. Validation of this screening algorithm was performed in six strains from the overall cohort using an ex vivo model of endocarditis. This ex vivo model recapitulated the in vitro predictions of NaHCO₃-responsiveness vs. nonresponsiveness above in five of the six strains.

Culture-Free Detection of Antibiotic Resistance Markers from Native Patient Samples by Hybridization Capture Sequencing

The increasing incidence of antimicrobial resistance (AMR) is a major global challenge. Routine techniques for molecular AMR marker detection are largely based on low-plex PCR and detect dozens to hundreds of AMR markers. To allow for comprehensive and sensitive profiling of AMR markers, we developed a capture-based next generation sequencing (NGS) workflow featuring a novel AMR marker panel based on the curated AMR database ARESdb. Our primary objective was to compare the sensitivity of target enrichment-based AMR marker detection to metagenomics sequencing. Therefore, we determined the limit of detection (LOD) in synovial fluid and urine samples across four key pathogens. We further demonstrated proof-of-concept for AMR marker profiling from septic samples using a selection of urine samples with confirmed monoinfection. The results showed that the capture-based workflow is more sensitive and requires lower sequencing depth compared with metagenomics sequencing, allowing for comprehensive AMR marker detection with an LOD of 1000 CFU/mL. Combining the ARESdb AMR panel with 16S rRNA gene sequencing allowed for the culture-free detection of bacterial taxa and AMR markers directly from septic patient samples at an average sensitivity of 99%. Summarizing, the newly developed ARESdb AMR panel may serve as a valuable tool for comprehensive and sensitive AMR marker detection.

Cefiderocol Activity Against Clinical Pseudomonas aeruginosa Isolates Exhibiting Ceftolozane-Tazobactam Resistance

BACKGROUND

Mutations in the AmpC-AmpR region are associated with treatment-emergent ceftolozane-tazobactam (TOL-TAZ) and ceftazidime-avibactam (CAZ-AVI) resistance. We sought to determine if these mutations impact susceptibility to the novel cephalosporin-siderophore compound cefiderocol.

METHODS

Thirty-two paired isolates from 16 patients with index P. aeruginosa isolates susceptible to TOL-TAZ and subsequent P. aeruginosa isolates available after TOL-TAZ exposure from January 2019 to December 2020 were included. TOL-TAZ, CAZ-AVI, imipenem-relebactam (IMI-REL), and cefiderocol minimum inhibitory concentrations (MICs) were determined using broth microdilution. Whole-genome sequencing of paired isolates was used to identify mechanisms of resistance to cefiderocol that emerged, focusing on putative mechanisms of resistance to cefiderocol or earlier siderophore-antibiotic conjugates based on the previously published literature.

RESULTS

Analyzing the 16 pairs of P. aeruginosa isolates, ≥4-fold increases in cefiderocol MICs occurred in 4 of 16 isolates. Cefiderocol nonsusceptibility criteria were met for only 1 of the 4 isolates, using Clinical and Laboratory Standards Institute criteria. Specific mechanisms identified included the following: AmpC E247K (2 isolates), MexR A66V and L57D (1 isolate each), and AmpD G116D (1 isolate) substitutions. For both isolates with AmpC E247K mutations, ≥4-fold MIC increases occurred for both TOL-TAZ and CAZ-AVI, while a ≥4-fold reduction in IMI-REL MICs was observed.

CONCLUSIONS

Our findings suggest that alterations in the target binding sites of P. aeruginosa-derived AmpC β-lactamases have the potential to reduce the activity of 3 of 4 novel β-lactams (ie, ceftolozane-tazobactam, ceftazidime-avibactam, and cefiderocol) and potentially increase susceptibility to imipenem-relebactam. These findings are in need of validation in a larger cohort.

Applications of Machine Learning to the Problem of Antimicrobial Resistance: an Emerging Model for Translational Research

Antimicrobial resistance (AMR) remains one of the most challenging phenomena of modern medicine. Machine learning (ML) is a subfield of artificial intelligence that focuses on the development of algorithms that learn how to accurately predict outcome variables using large sets of predictor variables that are typically not hand selected and are minimally curated. Models are parameterized using a training data set and then applied to a test data set on which predictive performance is evaluated. The application of ML algorithms to the problem of AMR has garnered increasing interest in the past 5 years due to the exponential growth of experimental and clinical data, heavy investment in computational capacity, improvements in algorithm performance, and increasing urgency for innovative approaches to reducing the burden of disease. Here, we review the current state of research at the intersection of ML and AMR with an emphasis on three domains of work. The first is the prediction of AMR using genomic data. The second is the use of ML to gain insight into the cellular functions disrupted by antibiotics, which forms the basis for understanding mechanisms of action and developing novel anti-infectives. The third focuses on the application of ML for antimicrobial stewardship using data extracted from the electronic health record. Although the use of ML for understanding, diagnosing, treating, and preventing AMR is still in its infancy, the continued growth of data and interest ensures it will become an important tool for future translational research programs.

Recent advances in rapid antimicrobial susceptibility testing systems

INTRODUCTION

Until recently antimicrobial susceptibility testing (AST) methods based on the demonstration of phenotypic susceptibility in 16-24 h remained largely unchanged.

AREAS COVERED

Advances in rapid phenotypic and molecular-based AST systems.

EXPERT OPINION

AST has changed over the past decade, with many rapid phenotypic and molecular methods developed to demonstrate phenotypic or genotypic resistance, or biochemical markers of resistance such as β-lactamases associated with carbapenem resistance. Most methods still require isolation of bacteria from specimens before both legacy and newer methods can be used. Bacterial identification by MALDI-TOF mass spectroscopy is now widely used and is often key to the interpretation of rapid AST results. Several PCR arrays are available to detect the most frequent pathogens associated with bloodstream infections and their major antimicrobial resistance genes. Many advances in whole-genome sequencing of bacteria and fungi isolated by culture as well as directly from clinical specimens have been made but are not yet widely available. High cost and limited throughput are the major obstacles to uptake of rapid methods, but targeted use, continued development and decreasing costs are expected to result in more extensive use of these increasingly useful methods.

Genomic and Phenotypic Analysis of Linezolid-Resistant Staphylococcus epidermidis in a Tertiary Hospital in Innsbruck, Austria

Whole genome sequencing is a useful tool to monitor the spread of resistance mechanisms in bacteria. In this retrospective study, we investigated genetic resistance mechanisms, sequence types (ST) and respective phenotypes of linezolid-resistant Staphylococcus epidermidis (LRSE, n = 129) recovered from a cohort of patients receiving or not receiving linezolid within a tertiary hospital in Innsbruck, Austria. Hereby, the point mutation G2603U in the 23S rRNA (n = 91) was the major resistance mechanism followed by the presence of plasmid-derived cfr (n = 30). The majority of LRSE isolates were ST2 strains, followed by ST5. LRSE isolates expressed a high resistance level to linezolid with a minimal inhibitory concentration of ≥256 mg/L (n = 83) in most isolates, particularly in strains carrying the cfr gene (p < 0.001). Linezolid usage was the most prominent (but not the only) trigger for the development of linezolid resistance. However, administration of linezolid was not associated with a specific resistance mechanism. Restriction of linezolid usage and the monitoring of plasmid-derived cfr in LRSE are potential key steps to reduce linezolid resistance and its transmission to more pathogenic Gram-positive bacteria.

Genotypic Characterization of Clinical Klebsiella spp. Isolates Collected From Patients With Suspected Community-Onset Sepsis, Sweden

Klebsiella is a genus of Gram-negative bacteria known to be opportunistic pathogens that may cause a variety of infections in humans. Highly drug-resistant Klebsiella species, especially K. pneumoniae, have emerged rapidly and are becoming a major concern in clinical management. Although K. pneumoniae is considered the most important pathogen within the genus, the true clinical significance of the other species is likely underrecognized due to the inability of conventional microbiological methods to distinguish between the species leading to high rates of misidentification. Bacterial whole-genome sequencing (WGS) enables precise species identification and characterization that other technologies do not allow. Herein, we have characterized the diversity and traits of Klebsiella spp. in community-onset infections by WGS of clinical isolates (n = 105) collected during a prospective sepsis study in Sweden. The sequencing revealed that 32 of the 82 isolates (39.0%) initially identified as K. pneumoniae with routine microbiological methods based on cultures followed by matrix-assisted laser desorption-time of flight mass spectrometry (MALDI-TOF MS) had been misidentified. Of these, 23 were identified as Klebsiella variicola and nine as other members of the K. pneumoniae complex. Comparisons of the number of resistance genes showed that significantly fewer resistance genes were detected in Klebsiella oxytoca compared to K. pneumoniae and K. variicola (both values of p < 0.001). Moreover, a high proportion of the isolates within the K. pneumoniae complex were predicted to be genotypically multidrug-resistant (MDR; 79/84, 94.0%) in contrast to K. oxytoca (3/16, 18.8%) and Klebsiella michiganensis (0/4, 0.0%). All isolates predicted as genotypically MDR were found to harbor the combination of β-lactam, fosfomycin, and quinolone resistance markers. Multi-locus sequence typing (MLST) revealed a high diversity of sequence types among the Klebsiella spp. with ST14 (10.0%) and ST5429 (10.0%) as the most prevalent ones for K. pneumoniae, ST146 for K. variicola (12.0%), and ST176 for K. oxytoca (25.0%). In conclusion, the results from this study highlight the importance of using high-resolution genotypic methods for identification and characterization of clinical Klebsiella spp. isolates. Our findings indicate that infections caused by other members of the K. pneumoniae complex than K. pneumoniae are a more common clinical problem than previously described, mainly due to high rates of misidentifications.

Multicenter Evaluation of the Unyvero Platform for Testing Bronchoalveolar Lavage Fluid

Bronchoalveolar lavage (BAL) culture is a standard, though time-consuming, approach for identifying microorganisms in patients with severe lower respiratory tract (LRT) infections. The sensitivity of BAL culture is relatively low, and prior antimicrobial therapy decreases the sensitivity further, leading to overuse of empirical antibiotics. The Unyvero LRT BAL Application (Curetis GmbH, Germany) is a multiplex molecular panel that detects 19 bacteria, 10 antibiotic resistance markers, and a fungus, Pneumocystis jirovecii, in BAL fluid in ∼4.5 h. Its performance was evaluated using 1,016 prospectively collected and 392 archived specimens from 11 clinical trial sites in the United States. Overall positive and negative percent agreements with culture results for identification of bacteria that grow in routine cultures were 93.4% and 98.3%, respectively, with additional potential pathogens identified by Unyvero in 21.7% of prospectively collected specimens. For detection of P. jirovecii, the positive percent agreement with standard testing was 87.5%. Antibiotic resistance marker results were compared to standard antibiotic susceptibility test results to determine positive predictive values (PPVs). PPVs ranged from 80 to 100%, based on the microorganism and specific resistance marker(s). The Unyvero LRT BAL Application provides accurate detection of common agents of bacterial pneumonia and of P. jirovecii The sensitivity and rapidity of this panel suggest significant clinical value for choosing appropriate antibiotics and for antibiotic stewardship.

Learning From Limited Data: Towards Best Practice Techniques for Antimicrobial Resistance Prediction From Whole Genome Sequencing Data

Antimicrobial resistance prediction from whole genome sequencing data (WGS) is an emerging application of machine learning, promising to improve antimicrobial resistance surveillance and outbreak monitoring. Despite significant reductions in sequencing cost, the availability and sampling diversity of WGS data with matched antimicrobial susceptibility testing (AST) profiles required for training of WGS-AST prediction models remains limited. Best practice machine learning techniques are required to ensure trained models generalize to independent data for optimal predictive performance. Limited data restricts the choice of machine learning training and evaluation methods and can result in overestimation of model performance. We demonstrate that the widely used random k-fold cross-validation method is ill-suited for application to small bacterial genomics datasets and offer an alternative cross-validation method based on genomic distance. We benchmarked three machine learning architectures previously applied to the WGS-AST problem on a set of 8,704 genome assemblies from five clinically relevant pathogens across 77 species-compound combinations collated from public databases. We show that individual models can be effectively ensembled to improve model performance. By combining models via stacked generalization with cross-validation, a model ensembling technique suitable for small datasets, we improved average sensitivity and specificity of individual models by 1.77% and 3.20%, respectively. Furthermore, stacked models exhibited improved robustness and were thus less prone to outlier performance drops than individual component models. In this study, we highlight best practice techniques for antimicrobial resistance prediction from WGS data and introduce the combination of genome distance aware cross-validation and stacked generalization for robust and accurate WGS-AST.

Drug-Resistant Fungi: An Emerging Challenge Threatening Our Limited Antifungal Armamentarium

The high clinical mortality and economic burden posed by invasive fungal infections (IFIs), along with significant agricultural crop loss caused by various fungal species, has resulted in the widespread use of antifungal agents. Selective drug pressure, fungal attributes, and host- and drug-related factors have counteracted the efficacy of the limited systemic antifungal drugs and changed the epidemiological landscape of IFIs. Species belonging to Candida, Aspergillus, Cryptococcus, and Pneumocystis are among the fungal pathogens showing notable rates of antifungal resistance. Drug-resistant fungi from the environment are increasingly identified in clinical settings. Furthermore, we have a limited understanding of drug class-specific resistance mechanisms in emerging Candida species. The establishment of antifungal stewardship programs in both clinical and agricultural fields and the inclusion of species identification, antifungal susceptibility testing, and therapeutic drug monitoring practices in the clinic can minimize the emergence of drug-resistant fungi. New antifungal drugs featuring promising therapeutic profiles have great promise to treat drug-resistant fungi in the clinical setting. Mitigating antifungal tolerance, a prelude to the emergence of resistance, also requires the development of effective and fungal-specific adjuvants to be used in combination with systemic antifungals.

Modifiable Risk Factors for the Emergence of Ceftolozane-Tazobactam Resistance

BACKGROUND

Ceftolozane-tazobactam (TOL-TAZ) affords broad coverage against Pseudomonas aeruginosa. Regrettably, TOL-TAZ resistance has been reported. We sought to identify modifiable risk factors that may reduce the emergence of TOL-TAZ resistance.

METHODS

Twenty-eight patients infected with carbapenem-resistant P. aeruginosa isolates susceptible to TOL-TAZ and treated with ≥72 hours of TOL-TAZ between January 2018 and December 2019 in Baltimore, Maryland were included. The 28 patients had P. aeruginosa isolates available both before and after TOL-TAZ exposure. Cases were defined as patients with at least a four-fold increase in P. aeruginosa TOL-TAZ MICs after exposure to TOL-TAZ. Independent risk factors for the emergence of TOL-TAZ resistance comparing cases and controls were investigated using logistic regression. Whole genome sequencing of paired isolates was used to identify mechanisms of resistance that emerged during TOL-TAZ exposure.

RESULTS

Fourteen patients (50%) had P. aeruginosa isolates which developed high-level TOL-TAZ resistance (i.e., cases). Cases were more likely to have inadequate source control (29% vs. 0%, p=0.04) and were less likely to receive TOL-TAZ as an extended 3-hour infusion (0% vs. 29%; p=0.04). Eighty-six percent of index isolates susceptible to ceftazidime-avibactam (CAZ-AVI) had subsequent P. aeruginosa isolates with high-level resistance to CAZ-AVI, after TOL-TAZ exposure. Common mutations identified in TOL-TAZ resistant isolates involved AmpC, a known binding site for both ceftolozane and ceftazidime, and DNA polymerase.

CONCLUSION

Due to our small sample size, our results remain exploratory but forewarn of the potential emergence of TOL-TAZ resistance during therapy and suggest extending TOL-TAZ infusions may be protective. Larger studies are needed to investigate this association.

Analytical Performance Validation of Next-Generation Sequencing Based Clinical Microbiology Assays Using a K-mer Analysis Workflow

Next-generation sequencing (NGS) enables clinical microbiology assays such as molecular typing of bacterial isolates which is now routinely applied for infection control and epidemiology. Additionally, feasibility for NGS-based identification of antimicrobial resistance (AMR) markers as well as genetic prediction of antibiotic susceptibility testing results has been demonstrated. Various bioinformatics approaches enabling NGS-based clinical microbiology assays exist, but standardized, computationally efficient and scalable sample-to-results workflows including validated quality control parameters are still lacking. Bioinformatics analysis workflows based on k-mers have been shown to allow for fast and efficient analysis of large genomics data sets as obtained from microbial sequencing applications. We here demonstrate applicability of k-mer based clinical microbiology assays for whole-genome sequencing (WGS) including variant calling, taxonomic identification, bacterial typing as well as AMR marker detection. The wet-lab and dry-lab workflows were developed and validated in line with Clinical Laboratory Improvement Act (CLIA) guidelines for laboratory-developed tests (LDTs) on multi-drug resistant ESKAPE pathogens. The developed k-mer based workflow demonstrated ≥99.39% repeatability, ≥99.09% reproducibility and ≥99.76% accuracy for variant calling and applied assays as determined by intra-day and inter-day triplicate measurements. The limit of detection (LOD) across assays was found to be at 20× sequencing depth and 15× for AMR marker detection. Thorough benchmarking of the k-mer based workflow revealed analytical performance criteria are comparable to state-of-the-art alignment based workflows across clinical microbiology assays. Diagnostic sensitivity and specificity for multilocus sequence typing (MLST) and phylogenetic analysis were 100% for both approaches. For AMR marker detection, sensitivity and specificity were 95.29 and 99.78% for the k-mer based workflow as compared to 95.17 and 99.77% for the alignment-based approach. Summarizing, results illustrate that k-mer based analysis workflows enable a broad range of clinical microbiology assays, potentially not only for WGS-based typing and AMR gene detection but also genetic prediction of antibiotic susceptibility testing results.