Ensuring the reliability, standardization, and international comparability of antimicrobial resistance (AMR) surveillance data critically depends on the implementation of robust quality assurance frameworks. South Korea established the Korea Global Antimicrobial Resistance Surveillance System (Kor-GLASS), supported by a centralized quality control center (QCC). As Kor-GLASS transitioned from Phase II to Phase III, new bacterial species and antimicrobial agents were incorporated, underscoring the need to evaluate whether quality assurance performance could be sustained during system expansion. We analyzed interlaboratory proficiency testing (IPT) and external quality assessment (EQA) outcomes generated by the QCC between 2020 and 2024, covering Phases II and III of Kor-GLASS. Clinical isolates were collected at participating hospitals and transferred to organism-specialized analysis centers for standardized antimicrobial susceptibility testing (AST), while the QCC independently oversees data quality through IPT and EQA. IPT was conducted by comparing AST results between analysis centers and the QCC using subsets of routine clinical isolates, with acceptance criteria defined as categorical agreement (CA) ≥90% and major error rates <3%. EQA involved quarterly distribution of pre-characterized strains to participating centers. Additional evaluations addressed the performance of newly introduced ceftazidime-avibactam susceptibility testing and interlaboratory validation for Haemophilus spp. Across the study period, overall CA consistently exceeded 97% in IPT, and no EQA failures observed among participating centers. While major errors during Phase II were primarily attributable to AST reading and near-breakpoint discrepancies, their frequency markedly decreased in Phase III following targeted corrective actions and educational interventions. Susceptibility testing for ceftazidime-avibactam showed high concordance between centers, with rare discrepancies limited to near-breakpoint measurements. Interlaboratory validation confirmed acceptable performance for AST of Haemophilus spp., supporting its formal inclusion in Phase III. These findings demonstrate that a centralized, QCC-led quality assurance framework can maintain stable and reliable AMR surveillance performance during periods of system expansion. Beyond routine oversight, coordinated quality assurance activities function as an evidence-based evaluation of how standardized laboratory data are generated and validated, reinforcing their essential role in sustaining the credibility and future development of AMR surveillance systems.
We consecutively obtained four blaCTX-M-27-positive Klebsiella pneumoniae isolates (named KE411, KE412, KE413, and KE414) from a patient with acute leukemia who underwent sequential antimicrobial therapy. These isolates exhibited a stepwise increase in β-lactam resistance alongside changes in the antibiotic treatment and ultimately acquired resistance to all β-lactams examined, including carbapenems. This study aimed to identify the genetic changes between the isolates that led to increased antimicrobial resistance. Complete genome sequences of the four isolates were compared to identify genetic differences between them. To characterize the genetic alterations, we profiled outer membrane proteins and quantified mRNA expression. Only one single-nucleotide polymorphism was found between the isolates, indicating their clonality. However, multiple genomic structural alterations associated with mobile genetic elements were detected, many of which contributed to differences in antibiotic susceptibility. There were two key genetic events in the evolutionary process: First, insertion of an ISEcp1-mediated transposition unit carrying blaCTX-M-27 into the 23S rRNA gene. This insertion was first detected in isolate KE412, obtained during cefepime therapy, and was shared by the later isolates, resulting in increased blaCTX-M-27 expression. Second, a 48-bp deletion that inactivated the porin-encoding ompK36 gene, detected in isolates KE413 and KE414 (obtained after shifting to doripenem therapy). Because the K. pneumoniae clone was deficient in OmpK35 production, this deletion led to the acquisition of carbapenem resistance due to dual deficiency of OmpK35 and OmpK36. We reveal stepwise within-host evolution of a K. pneumoniae clone enabling its survival under changing antimicrobial pressures.
The complexity of age-dependent cerebrospinal fluid (CSF) dynamics and the insufficient quantification of disease-induced physiological alterations remain significant barriers to precisely assessing central nervous system (CNS) antimicrobial exposure in pediatric meningitis patients. This study aims to develop a CNS physiologically-based pharmacokinetic (PBPK) model to characterize CNS drug exposure and inform clinical dosing regimens. The model describes the general physiological processes of the CNS, assumptions, controlling equations and system parameters to support the description of multi-pathway dosing regimens and incorporates disease effects. The model was applied to the pediatric meningitis population across all age subgroups by incorporating developmental characteristics. PBPK models were validated using 44 datasets from the literature and real-world data, 98.7% of predictions fell within a 2-fold error margin. For vancomycin, 10-20 mg/kg q12h/q24h achieved efficacy with low nephrotoxicity risk at MIC ≤ 0.5 mg/L. Higher MICs necessitated age-stratified daily doses: 40-45 (adults), 45-60 (adolescents), 60-80 (children), 80 (infants), 40-60 (term), and 30-40 mg/kg (preterm). When MIC > 1 mg/L, the narrow therapeutic window warrants alternatives to balance efficacy and nephrotoxicity. For meropenem, 40-60 mg/kg q6h/q8h (3-h infusion) achieved 50% fT>MIC target at MIC ≤ 2 mg/L in patients >1 month, while <1 month required 10-20 mg/kg q6h/q8h. Achieving 100% fT>MIC remained challenging. The model further characterized meningitis-induced barrier permeability changes and route-dependent brain exposure. This CNS PBPK model offers a robust in silico framework to guide and optimize the selection of antibacterial dosing regimens for the treatment of CNS infections in clinical practice.
To systematically characterise the global distribution of bacterial pathogens, antimicrobial resistance patterns, and resistance determinants in non-bloodstream infections among adults with cirrhosis. We performed a systematic review and proportional meta-analysis of studies reporting AMR in non-bloodstream bacterial infections among patients with cirrhosis. PubMed, Embase, Scopus and Web of Science were searched up to 15 September 2025. Pooled proportion estimates of multidrug-resistance (MDR) and the etiological bacterial pathogens were calculated using random-effects models. Subgroup analyses were performed by country income, continent, and bacterial species. Thirty-one studies contributed data from 2781 clinical infections and 381 colonisation observations. Spontaneous bacterial peritonitis predominated (89%), followed by urinary tract infections (7%). Gram-negative bacteria accounted for 60% of isolates (Escherichia coli 29%, Klebsiella pneumoniae 11%), while Gram-positive pathogens represented 39% (Enterococcus spp. 14%, Staphylococcus aureus 6%). The overall pooled MDR proportion was 23% (95% CI: 13-37), with higher burdens in lower-middle-income countries (MDR 56% vs. 15%-41%; ESBL 24% vs. 8%-10%; VRE 23% vs. 2%; CRE 32% vs. 1%). Genotypic data identified 436 resistance genes with marked continental differences. In cirrhosis-associated non-bloodstream infections, Gram-negative bacteria predominate and are associated with high levels of MDR, particularly in lower-middle-income countries. These findings highlight the need for integrated phenotypic and genomic surveillance of resistance patterns to inform empiric antimicrobial therapy.
In response, to spreading resistance-associated mutations in Plasmodium falciparum, several malaria-endemic countries are evaluating the introduction of triple artemisinin combination therapies (TACTs), and alternative first-line ACTs alongside or in place of existing regimens such as artemether-lumefantrine (AL). Such combinations were associated with higher post treatment gametocyte count and prevalence compared with AL. We aim to mechanistically explain the effects of using ACTs associated with higher gametocytemia and the spread of drug resistance. We developed a deterministic model that describes the effects of drugs on idealized dynamics of asexual parasites (merozoites) and early- and late-stage gametocytes. The model incorporates differential drug activity against partially resistant and sensitive merozoites, as well as against early- and late-stage gametocytes. Model inputs are parameters describing the metabolic costs, delayed clearance associated with resistance mutations, and the onset of treatment timing. The model output is the evolutionary fitness, defined as the area under the curve (AUC) of late-stage gametocytes, of resistant and sensitive parasites. These fitness measures are combined with a population-genetic model tailored to the spread of antimalarial resistance to illustrate how differences in gametocyte dynamics influence resistance spread. We compared the evolutionary fitness of three drug regimens with equivalent efficacy against asexual parasites, but varying gametocytocidal activity: (i) no effect, (ii) activity against early-stage gametocytes only, and (iii) activity against both early- and late-stage gametocytes. Early-stage gametocytocidal activity substantially decelerated resistance spread. Late-stage gametocytocidal activity was effective only if sustained beyond asexual parasite clearance. As our model is mechanistic, these results should be interpreted qualitatively rather than qualitatively. Nevertheless. Our findings highlight that drug-induced gametocyte dynamics play a central role in shaping the evolutionary fitness of resistant parasites. In particular, treatment strategies associated with higher post-treatment gametocyte prevalence may increase the transmission advantage of resistant parasites. These results highlight the importance of considering gametocyte dynamics when evaluating strategies to limit the spread of antimalarial resistance, and suggest that interventions targeting gametocytes, such as 8-aminoquinolines, may help mitigate this effect.
Carbapenem-resistant Enterobacterales (CRE) are increasing in the UK, but the drivers of this rise and effective treatment options remain uncertain. Ninety-seven acute NHS Trusts, including 192 UK hospitals, submitted data on 8,840 consecutive, non-duplicate CRE (October 2023-September 2024) and local laboratory methods. Predominant CRE species were Klebsiella pneumoniae (30.4%, 2,709/8,840), E. coli (29.3%, 2,592/8,840), Enterobacter cloacae complex (22.2%, 1,966/8,840). 21.8% (1,927/8,840) of CRE were collected in outpatient settings. Overall, 85.5% (7,279/8,513) of CRE underwent carbapenemase detection testing, with significant variation depending on species, specimen type and baseline antibiogram. Carbapenemases detected were OXA-48-like (22.3%, 1,967/8,840), NDM (18.6%, 1,645/8,840), KPC (9.2%, 817/8,840), IMP (2.0%, 178/8,840), VIM (0.7%, 62/8,840), multi-carbapenemase producers (4.0%, 351/8,840). 51.2% (4,526/8,840) of CRE were tested against ≥1 novel agent. In-house antimicrobial susceptibility testing availability was common for ceftazidime-avibactam (90.7%, 88/97) and cefiderocol (73.2%, 71/97), while less common for meropenem-vaborbactam (40.2%, 40/97), imipenem-relebactam (21.6%, 21/97), aztreonam-avibactam (11.3%, 11/97). OXA-48-like- (97.6%, 1,138/1,166) and KPC-producers (96.2%, 429/446) remained susceptible to ceftazidime-avibactam. KPC-producers to meropenem-vaborbactam (99.4%, 160/161) and imipenem-relebactam (98.4%, 62/63). Cefiderocol resistance was 32.4% overall and higher among NDM- (58.4%) and multi-carbapenemase-producers (50.3%). Ceftazidime-avibactam plus aztreonam synergy was performed for 17.2% (286/1,658) of aztreonam-resistant metallo-β-lactamase-producers; synergy was observed in 78.7% (225/286) of cases. Colistin resistance was 10.6% (153/1,445) among non-intrinsically resistant species. Whilst most UK CRE are tested for carbapenemases, coverage gaps persist. The community burden of CRE is increasing. Cefiderocol resistance is concerningly high, particularly among metallo-β-lactamase-producers.
We reported stepwise evolution of a Methicillin-Resistant Staphylococcus epidermidis (MRSE) with reduced susceptibility to dalbavancin in a patient with recurrent episode of bloodstream infections (BSIs). SNPs (single nucleotide polymorphism) and insert/deletions analysis between genomes of dalbavancin -susceptible and -non-susceptible strains was performed. Dalbavancin exposure and PK/PD target attainment was determined by TDM. In vitro synergy testing was performed by evaluating the fractional inhibitory concentration indices. Four clonally related MRSE isolates belonging to ST23 were recovered during recurrent bacteraemic episodes in a single patient treated with dalbavancin-based therapy over a 3-year period. Progressive increases of the dalbavancin MIC was observed and resistome analysis showed a conserved antimicrobial resistance genes among isolates. First dalbavancin non-susceptible strain carried an A414T substitution within walK, whereas the second non-susceptible MRSE strain harboured L957F and G470D mutations in rpoB and vraG, respectively. TDM analysis indicated optimal plasma exposure and prolonged PK/PD target attainment by considering clinical breakpoint and MIC of dalbavancin of the susceptible MRSE strains. Synergy testing demonstrated that dalbavancin combined with fosfomycin exhibited synergistic activity against 75% of isolates, whereas combinations with β-lactams were mostly indifferent. We described in vivo evolution of dalbavancin reduced susceptibility in MRSE during long-term dalbavancin therapy highlighting multiple genetic trajectories involving different genetic determinants. These findings underscore the risk of resistance selection despite adequate systemic exposure and support the need for optimized dosing strategies, source control in preventing recurrence in prosthetic infections, and combination regimens to prevent resistance during dalbavancin treatment of MRSE infections.
Salmonella enterica is a major public health concern due to its widespread contamination of food products and the increasing prevalence of multidrug-resistant strains. Bacteriophage-based biocontrol has emerged as a safe and effective alternative to conventional antibiotics. Here, we isolated and characterized the broad-spectrum lytic phage vB_SgulP_SP124, which lysed 81.08% of 148 tested Salmonella strains across five clinically important serotypes: S. Enteritidis, S. Typhimurium, S. Pullorum, S. Choleraesuis, and S. Dublin. The phage exhibited high viability across a pH range of 2-10 and temperatures ranging from 4°C to 60°C. Its 40,499 bp dsDNA genome (G + C content 49.60%) was classified within the class Caudoviricetes, subfamily Guernseyvirinae, and genus Jerseyvirus. Genomic analysis confirmed the absence of virulence, lysogeny, or antibiotic resistance genes, supporting its safety profile for biocontrol applications. In planktonic assays conducted at the optimal multiplicity of infection of 0.01, the phage significantly inhibited the growth of four Salmonella strains, with viable counts of two strains reduced within 2 h. Moreover, the phage significantly reduced S. Enteritidis levels on chicken breast, milk, liquid egg, and lettuce during 36 h of storage at 4°C. Pharmacokinetic analysis in specific pathogen-free broilers showed that intraperitoneal (IP) administration provided superior phage persistence and biodistribution compared with oral and intramuscular routes. Both early and delayed IP phage treatment significantly lowered bacterial burdens in blood and organs, although early intervention showed a consistent (but non-significant) trend toward greater efficacy. These results establish phage vB_SgulP_SP124 as a promising biocontrol agent for Salmonella contamination across the farm-to-table continuum.
This study gathered public data regarding the in vitro activity of three novel antifungal agents, ibrexafungerp, manogepix, and rezafungin, against the fungi listed in the WHO fungal priority pathogens list. Epidemiological cut-off values (ECVs/ECOFFs) were established for species with sufficient data (≥100 MIC values) and when distributions met further essential criteria. For pathogens in which data did not fulfil one or more essential criteria, tentative ECVs (T-ECVs/T-ECOFFs) were proposed. The values were determined by a comprehensive analysis that included visual inspection of histograms and the calculation of ECOFFinder-based values (95% and 97.5%), as well as the Mode + 2 and MIC50 + 2 dilutions. A total of 34,736 MIC values were collected from studies using standardised CLSI (25,027) or EUCAST (9,709) methodologies: ibrexafungerp (5,388), manogepix (10,610), and rezafungin (18,738). The analysis successfully defined ECVs for a core group of clinically relevant yeasts, including Candida albicans, Candidozyma auris, and Nakaseomyces glabratus. Tentative ECVs were proposed for data-limited pathogens such as Cryptococcus neoformans, Lomentospora prolificans, and Talaromyces marneffei, among others. For Aspergillus fumigatus, a CLSI ECV was also established with rezafungin. The ECVs and T-ECVs derived herein provide crucial reference points to guide future resistance surveillance and inform the foundational epidemiology needed for clinical breakpoint setting. Importantly, the data gaps identified for dimorphic and filamentous fungi underscore the urgent need for standardised susceptibility testing methods and targeted research efforts for these neglected priority pathogens.
Therapeutic drug monitoring (TDM) individualises antibiotic dosing, aiming for therapeutic exposures and minimising toxicity. However, TDM cannot always be performed due to resource requirements, and clinicians often dose according to guidelines. This study aimed to quantify the achievement of therapeutic exposures (concentrations) in a heterogeneous cohort of critically ill adults using nomogram-based antibiotic dosing. This was a prospective, single-centre, cohort study of nomogram-based dosing of beta-lactams in critically ill adults. The primary outcome was the proportion of patients with therapeutic exposures, that is those above the efficacy threshold but below the toxic threshold. Secondary endpoints were exposures associated with toxicity, and the relationship between exposure and clinical outcomes. In 79 patients, therapeutic beta-lactam exposure was achieved in 82% (63/77) and 81% (35/43) of patients at days 1 to 3 and days 4 to 6, respectively. Piperacillin yielded the lowest therapeutic attainment, with 75% (24/32) and 71% (12/17) at days 1 to 3 and days 4 to 6, respectively, and the highest proportion of sub-therapeutic exposures: 25% (8/32) and 29% (5/17), respectively. Therapeutic tazobactam trough concentrations were achieved in 66% (21/32) and 65% (11/17) of patients at days 1 to 3 and days 4 to 6, respectively. In a single-centre heterogeneous population of critically ill adults with infection, initial dosing informed by a dosing nomogram for beta-lactams resulted in a high proportion of patients achieving therapeutic exposures.
The global rise of pathogens co-resistant to carbapenems and tigecycline represents a critical public health threat. This study aimed to characterize clinical K. pneumoniae co-harbouring carbapenem and tigecycline resistance genes, focusing on their co-occurrence and transmission mechanisms. Clinical K. pneumoniae isolates co-carrying carbapenemase and tigecycline resistance genes were screened by PCR. Isolates were further characterized using antimicrobial susceptibility testing, conjugation experiments, whole-genome sequencing, and bioinformatics analysis. Publicly available genomes of ST1545 K. pneumoniae and K. pneumoniae co-harbouring tmexCD-toprJ and blaNDM/blaKPC were retrieved for comparative analysis. Seven multidrug-resistant ST1545 K. pneumoniae isolates co-harbouring blaNDM-1 and blaKPC-2 were identified, with three additionally carrying tmexCD2-toprJ2. In seven isolates, blaNDM-1 genes were located on non-conjugative multi-replicon plasmids, three of which co-harboured tmexCD2-toprJ2, while blaKPC-2 resided on conjugative IncFII plasmids. Genetic environment analysis indicated that ISKpn14, ISKpn28, IS26, and class 1 integron facilitate mobilization of blaNDM-1, tmexCD2-toprJ2 and blaKPC-2. Analysis across 13 ST1545 K. pneumoniae isolates revealed SNP changes associated with plasmid and resistance genes acquisition. Analysis of 84 K. pneumoniae strains co-harbouring tmexCD-toprJ and blaNDM/blaKPC showed diverse STs, a primary human origin in China, and a high burden of resistance genes, with blaNDM-1, blaKPC-2, and tmexCD1-toprJ1/tmexCD2-toprJ2 as predominant variants. This study first reported the convergence of blaNDM-1, blaKPC-2, and tmexCD2-toprJ2 in ST1545 K. pneumoniae, highlighting the pivotal role of diverse mobile genetic elements in the accumulation and dissemination of carbapenem and tigecycline resistance. ST1545 K. pneumoniae may represent a potential reservoir for these critical resistance determinants, warranting continuous surveillance.
Azole resistance in Candida tropicalis has increased worldwide and is largely driven by an emerging population previously reported Cluster AZR/X. This cluster exhibits high-level azole resistance driven by ERG11 A395T/W mutations accompanied by amplification of mutant ERG11 alleles, underscoring the clinical need for a quantitative, mechanism-informed molecular assay. We developed and validated a droplet digital PCR (ddPCR)-based assay enabling simultaneous detection of the ERG11 mutation and quantitative assessment of ERG11 copy number variations (CNVs). The assay employs allele-specific probes to detect wild-type and mutant ERG11 alleles, and normalised against endogenous reference genes for allele-resolved quantification. Validation of the ddPCR assay was carried out in 93 clinical isolates, and the results were compared with whole-genome sequencing data. The assay demonstrated excellent concordance for both ERG11 A395T/W mutation detection (κ = 1.00) and CNV estimation (R² = 0.98), and exhibited high repeatability and specificity. When further applied to a multicentre collection of 98 invasive C. tropicalis isolates, the ddPCR assay predicted fluconazole resistance with 91.4% sensitivity and 100% specificity compared with phenotypic antifungal susceptibility results. Furthermore, our proof-of-concept evaluation demonstrated the feasibility of direct ddPCR analysis of clinical specimens. This study established a mechanism-informed ddPCR assay that enables accurate detection of mutation and CNVs of azole target, providing a practical tool to address the growing clinical and public health challenges posed by the emerging azole-resistant C. tropicalis cluster.
Sulfaguanidine (SG), a sulfonamide-based synthetic antimicrobial agent, is widely used in livestock and aquaculture. Due to the presence of N1-guanidinium substitutions in its molecular structure, SG exhibits not only high ecotoxicity but also remarkable stability, making it difficult to degrade in the environment. Consequently, residues of SG can persist in water and soil for extended periods, posing risks of bioaccumulation and biomagnification through the food chain. Therefore, establishing a rapid and sensitive on-site detection method to monitor SG residues is of great significance. However, the unique structure of SG limits the availability of highly sensitive and specific haptens suitable for immunoassays, which hinders the development of rapid immunodetection technologies. To address this challenge, two novel haptens were designed via electrostatic potential, molecular orbitals and immunodeterminant moiety in this study, to optimize the exposure of the unique guanidino group. This approach laid the foundation for establishing a highly sensitive immunoassay for SG. Subsequently, a high-affinity monoclonal antibody, 3G5 (IC50 = 0.11 ng/mL), was obtained via mouse immunization. Based on these findings, a colloidal gold-based visual immunochromatographic assay for SG detection was pioneered. Following simple extraction and dilution procedures, the visual detection limits were 0.2 ng/mL in milk and 1 ng/mL in fish. The method was further validated using real samples, yielding recoveries ranging from 89.1% to 107.5%, confirming the method's applicability for detecting SG residues in complex matrix samples. The present study introduces a novel approach for the sensitive and precise detection of persistent organic pollutants.
Antimicrobial resistance is a global health problem, with methicillin-resistant Staphylococcus aureus (MRSA) causing significant morbidity and mortality. As antibiotic options dwindle, (bacterio) phage therapy is re-emerging as a promising alternative. We characterised a novel lytic S. aureus phage, PBSA08, and evaluated its host range using a reference panel of 39 MRSA isolates selected from the 2021 Australian Group on Antimicrobial Resistance repository. Phage susceptibility was tested by efficiency of plating (EOP) and growth kinetics assays (GKA). Factors associated with phage susceptibility were evaluated. Phage PBSA08, classified under the genus Silviavirus, revealed an icosahedral head and long tail structure using transmission electron microscopy. Adsorption assays indicated efficient bacterial binding, with a burst size of approximately 120 plaque forming units (PFU) per infected cell. EOP testing revealed 51.3% of the MRSA reference panel isolates were susceptible (EOP ≥10%). GKA results correlated strongly with EOP. In univariate analysis, phage PBSA08 susceptibility was associated with multilocus sequence type (ST), ciprofloxacin susceptibility, and the presence of the Panton-Valentine leucocidin-associated genes, with all ST30-IV and ST93-IV isolates susceptible to PBSA08 and all ST45-V isolates resistant to PBSA08. Firth logistic regression showed that ST30/93 isolates had significantly higher odds of susceptibility (aOR 68.9, 95% CI 2.0-2378.5; P = 0.0191), as did ciprofloxacin-susceptible strains (aOR 15.4, 95% CI 1.7-143.0; P = 0.0160). PBSA08 demonstrates potential as a therapeutic phage, particularly against community-associated MRSA lineages such as ST93. Its inclusion in phage cocktails targeting prevalent MRSA clones warrants further investigation.
Artificial intelligence (AI) tools are increasingly used to support antimicrobial prescribing, but most published literature focuses on accuracy rather than reproducibility across identical inputs. Reproducibility is a key requirement for safe clinical decision support. We conducted a comparative experimental study assessing the reproducibility of outputs and the clinical correctness of antibiotic and treatment-duration recommendations generated by a general-purpose large language model (ChatGPT-5.2) and a literature-grounded clinical decision-support system (OpenEvidence). Twelve standardized infectious disease cases (four Gram-positive, four Gram-negative non-fermenters, and four Gram-negative fermenters) were submitted 100 times to each tool using official web interfaces under default settings. Variability was quantified using Shannon entropy, normalized entropy, and the Gini-Simpson index with bootstrap confidence intervals. Clinical correctness was assessed by three blinded infectious diseases specialists. Across 2,400 independent model interactions, corresponding to 12 clinical scenarios submitted 100 times to each of the two AI tools, substantial variability in response reproducibility was observed. For antibiotic selection, entropy ranged from complete agreement (H=0) to high heterogeneity (H up to 2.53). ChatGPT showed complete agreement in five of 12 cases, compared with one of 12 for OpenEvidence. In some scenarios, the probability that two responses were different reached 81%. Despite lower reproducibility, OpenEvidence generated more correct antibiotic recommendations (54.8% vs 35.7%; p<0.001) and more fully correct combined responses (33.4% vs 19.0%; p<0.001). In mixed-effects analysis, OpenEvidence had higher odds of correct antibiotic recommendations (OR 12.4, 95% CI 8.1-18.7). Substantial variability in outputs generated from identical inputs was observed, and reproducibility did not align with clinical correctness. Such variability may lead to inconsistent therapeutic decisions, raising concerns about the use of these tools for direct clinical decision-making. Reproducibility should be considered a key requirement, alongside accuracy, in the evaluation and governance of AI-based clinical decision-support tools.
Invasive fungal infections, such as those caused by fluconazole-resistant Candida tropicalis, pose an important threat to immunocompromised patients, with mortality rates ranging from 55 to 60%. The rise of clinically resistant C. tropicalis has compounded the difficulties in treatment, highlighting the urgent need for the development of novel antifungal agents. This study builds on previous research to evaluate the antifungal efficacy of dihydroartemisinin-loaded chitosan nanoparticles (DHA-CS NPs) against fluconazole-resistant C. tropicalis and elucidates their mechanisms of action. In vitro, antifungal activity was assessed via MIC/MFC determination, time‑kill kinetics, Scanning electron microscopy observation, and multi‑omics analyses, with filipin staining and RT‑qPCR used to verify ergosterol pathway genes (ERG3, HAP1, and UPC2). In vivo, efficacy was evaluated in a murine model by measuring fungal burden, immune markers (IL‑6, TNF‑α, and TLR2), and tissue damage. DHA-CS NPs demonstrated potent antifungal activity, with minimum inhibitory concentration (MIC) and minimum fungicidal concentration (MFC) values of 7.5 and 15 µg/mL, respectively, achieving 95.75% inhibition at the MIC. The antifungal effect was concentration-dependent, with sterilization achieved in 4 h at 4 × MIC. Scanning electron microscopy revealed that DHA-CS NPs compromised the integrity of the fungal cell membrane. Metabolomic and transcriptomic analyses revealed that DHA-CS NPs significantly disrupted ergosterol biosynthesis, a key component for maintaining fungal cell membrane integrity. Filipin fluorescent staining and RT-qPCR further confirmed this mechanism, showing that DHA-CS NPs inhibited ergosterol synthesis by downregulating critical biosynthetic genes, including ERG3, HAP1, and UPC2 (P < 0.05). In a murine infection model, DHA-CS NPs reduced fungal burden, modulated immune responses (increasing IL-6, TNF-α, and TLR2), and alleviated tissue damage. These findings underscore the potential of DHA-CS NPs as an effective therapeutic approach for drug-resistant candidiasis. In summary, DHA-CS NPs present a promising strategy against fluconazole-resistant C. tropicalis, providing a solid scientific foundation for the development of novel anti-resistance therapies.
Carbapenem-resistant Acinetobacter baumannii (CRAB) is a persistent nosocomial pathogen, posing a major global health threat owing to limited treatment options. Although tigecycline is still effective against CRAB, resistance emergence has become a critical concern. This study aimed to elucidate the in vivo evolutionary mechanisms underlying tigecycline resistance in CRAB. A total of 11 STpas2-SToxf540 CRAB strains were recovered from rectal swabs, sputum samples, and the surrounding environmental specimens of an intensive care unit (ICU) hospitalized patient who had received tigecycline treatment. CgSNP analysis confirmed environment-to-patient transmission among these CRAB isolates. Comparative genomic analysis indicated that mutations in adeS, pgaA, and gbsA may be associated with tigecycline resistance. In situ mutagenesis and antimicrobial susceptibility testing verified that the adeS N125S mutation mediates tigecycline resistance and collateral sensitivity to cefoperazone/sulbactam. Growth curve assays demonstrated that the adeS N125S mutation imposes a fitness cost on CRAB. Transcriptional analysis showed that the adeS N125S mutation drives the development of tigecycline resistance by upregulating the expression of the AdeABC efflux pump. By examining 45,377 publicly available global A. baumannii genome sequences, we characterized the molecular epidemiology of the STpas2-SToxf540 lineage and the distribution of adeS mutations. Our findings underscore the need to strengthen surveillance and rational antimicrobial use to prevent CRAB dissemination and antimicrobial resistance evolution, particularly in ICUs.
The escalating antimicrobial resistance crisis poses a significant threat to global health. Infections caused by carbapenem-resistant Gram-negative bacteria, particularly those producing metallo-β-lactamases (MBLs), present a formidable therapeutic challenge. Aztreonam-avibactam is a novel combination of a monobactam and a β-lactamase inhibitor that leverages aztreonam's stability against MBLs and avibactam's inhibition of co-produced serine β-lactamases, thereby demonstrating activity against all four Ambler classes. This review systematically collates current evidence concerning the synergistic mechanism, microbiological spectrum, pharmacokinetics/pharmacodynamics (PK/PD), clinical efficacy, safety profiles, resistance mechanisms, and antimicrobial susceptibility testing of aztreonam-avibactam. Aztreonam-avibactam demonstrates potent in vitro activity against carbapenem-resistant Enterobacterales (CRE), including MBL-producing strains. Its efficacy and safety have been evaluated in Phase III trials for the treatment of complicated intra-abdominal infections (cIAIs), complicated urinary tract infections (cUTIs), and hospital-acquired pneumonia (HAP) in adults. However, clinical data specifically for microbiologically confirmed MBL-producing infections remain limited and exploratory because of very small subgroup sizes. PK/PD analyses indicate that the synergistic interaction ensures effective β-lactamase suppression. Aztreonam-avibactam resistance mechanisms, including penicillin-binding protein 3 (PBP3) modifications, β-lactamase variants, porin mutations, and efflux alterations, exhibit genus-specific variations. However, current susceptibility testing still faces limitations in clinical implementation. Overall, aztreonam-avibactam offers a promising option for managing severe infections caused by carbapenem-resistant Gram-negative pathogens, supported by strong in vitro and PK/PD evidence. Future efforts should focus on expanding real-world evidence and optimizing dosing strategies to maximize its clinical utility.
Tuberculosis (TB), an infectious disease caused by Mycobacterium tuberculosis (Mtb) infection, has remained a persistent threat to global public health. Chemotherapy remains the primary treatment for active TB. Under sustained selective drug pressure, Mtb continues to evolve, facilitating the emergence of acquired drug resistance. Heteroresistance, a unique and complex form of TB drug resistance, refers to the co-existence of drug-resistant and drug-susceptible subpopulations within a host. These subpopulations may originate from various sources, including mixed infections or clonal heterogeneity, and undergo dynamic shifts in their relative proportions. This characteristic makes heteroresistance a critical factor contributing to underdiagnosis and treatment failure in clinical practice. Traditional detection methods are not only associated with long culture turnaround times but also have limited capacity to identify diverse drug resistance patterns. In recent years, the continued development of novel detection technologies has provided new strategies for the clinical management of TB and the prevention and control of drug resistance. PubMed was searched through both key terms and subject headings. The literatures were screened, assessed for the quality and evidence synthesized. This review summarizes the molecular mechanisms, detection methods, and targeted therapeutic strategies associated with TB heteroresistance, with the aim of enhancing the understanding of drug resistance evolution in Mtb and facilitating the development of more effective approaches to combat drug-resistant TB.
Cefiderocol (FDC) is a siderophore cephalosporin active against carbapenem-resistant Klebsiella pneumoniae, but resistance has emerged, often involving iron-uptake pathways and β-lactamases. We investigated how an NDM-producing clinical isolate adapts to FDC under stepwise selective pressure. A ST147 K. pneumoniae isolate (Kp-1) carrying blaNDM-5 and blaOXA-181, susceptible to FDC but resistant to other agents, was propagated in iron-depleted Mueller-Hinton broth with progressive FDC exposure. Population-based whole-genome sequencing tracked mutation dynamics, and reverse transcription-quantitative PCR profiled 17 genes related to FDC resistance, iron acquisition, and virulence, comparing the parental population (G0) to the first exposed generation (G1 + FDC). The mean number of mutations detected in generation G1 under FDC exposure was 24.3 ± 12.2, and in generation G2, it increased to 31.7 ± 6.4. Two nonsynonymous substitutions in baeS (V295G and T299P) were recurrently selected, with V295G reaching ≥ 70% frequency in independent lineages, indicating convergent evolution. reverse transcription-quantitative PCR revealed downregulation of iron-uptake and porin genes (iroN, cirA, fepA, kfu, ompK35) and upregulation of entB, ompK36, blaNDM-5, and capsule-related genes (wzm, wbbM); baeS/baeR transcript levels were unchanged. Early FDC exposure enhanced biofilm formation without significantly affecting capsule production. Short-term FDC exposure reproducibly selects baeS variants and triggers a transcriptional program reducing siderophore-receptor entry and remodelling the outer membrane. These adaptations-together with increased blaNDM-5 expression and biofilm formation-contribute to reduced FDC susceptibility. The recurrent baeS mutations (V295G/T299P) and decreased cirA/fepA expression may serve as early surveillance markers of FDC adaptation in K. pneumoniae.