Search

Background Genomic epidemiology studies of gonorrhea in the United States have primarily focused on national surveillance for antibiotic resistance, and patterns of local transmission between demographic groups of resistant and susceptible strains are unknown. Methods We analyzed a convenience sample of genome sequences, antibiotic susceptibility, and patient data from 897 gonococcal isolates cultured at the New York City (NYC) Public Health Laboratory from NYC Department of Health and Mental Hygiene (DOHMH) Sexual Health Clinic (SHC) patients, primarily in 2012–2013. We reconstructed the gonococcal phylogeny, defined transmission clusters using a 10 nonrecombinant single nucleotide polymorphism threshold, tested for clustering of demographic groups, and placed NYC isolates in a global phylogenetic context. Results The NYC gonococcal phylogeny reflected global diversity with isolates from 22/23 of the prevalent global lineages (96%). Isolates clustered on the phylogeny by patient sexual behavior (P <.001) and race/ethnicity (P <.001). Minimum inhibitory concentrations were higher across antibiotics in isolates from men who have sex with men compared to heterosexuals (P <.001) and white heterosexuals compared to black heterosexuals (P <.01). In our dataset, all large transmission clusters (≥10 samples) of N. gonorrhoeae were susceptible to ciprofloxacin, ceftriaxone, and azithromycin, and comprised isolates from patients across demographic groups. Conclusions All large transmission clusters were susceptible to gonorrhea therapies, suggesting that resistance to empiric therapy was not a main driver of spread, even as risk for resistance varied across demographic groups. Further study of local transmission networks is needed to identify drivers of transmission. [ABSTRACT FROM AUTHOR]

Prediction of antibiotic resistance phenotypes from whole genome sequencing data by machine learning methods has been proposed as a promising platform for the development of sequence-based diagnostics. However, there has been no systematic evaluation of factors that may influence performance of such models, how they might apply to and vary across clinical populations, and what the implications might be in the clinical setting. Here, we performed a meta-analysis of seven large Neisseria gonorrhoeae datasets, as well as Klebsiella pneumoniae and Acinetobacter baumannii datasets, with whole genome sequence data and antibiotic susceptibility phenotypes using set covering machine classification, random forest classification, and random forest regression models to predict resistance phenotypes from genotype. We demonstrate how model performance varies by drug, dataset, resistance metric, and species, reflecting the complexities of generating clinically relevant conclusions from machine learning-derived models. Our findings underscore the importance of incorporating relevant biological and epidemiological knowledge into model design and assessment and suggest that doing so can inform tailored modeling for individual drugs, pathogens, and clinical populations. We further suggest that continued comprehensive sampling and incorporation of up-to-date whole genome sequence data, resistance phenotypes, and treatment outcome data into model training will be crucial to the clinical utility and sustainability of machine learning-based molecular diagnostics. [ABSTRACT FROM AUTHOR]

Approximately 20% of Shigella isolates tested in New York City, New York, USA, during 2013-2015 displayed decreased azithromycin susceptibility. Case-patients were older and more frequently male and HIV infected than those with azithromycin-susceptible Shigella infection; 90% identified as men who have sex with men. Clinical interpretation guidelines for azithromycin resistance and outcome studies are needed. [ABSTRACT FROM AUTHOR]