Applying molecular algorithms to predict decreased susceptibility to ceftriaxone from a report of strains of Neisseria gonorrhoeae in Amsterdam, the Netherlands

Machine learning to predict ceftriaxone resistance using single nucleotide polymorphisms within a global database of Neisseria gonorrhoeae genomes

IMPORTANCE

Antimicrobial resistance in Neisseria gonorrhoeae is an urgent global health issue. The objectives of the study were to use a global collection of 12,936 N. gonorrhoeae genomes from the PathogenWatch database to evaluate different machine learning models to predict ceftriaxone susceptibility/decreased susceptibility using 97 mutations known to be associated with ceftriaxone resistance. We found the random forest classifier model had the highest performance. The analysis also reported the relative contributions of different mutations within the ML model predictions, allowing for the identification of the mutations with the highest importance for ceftriaxone resistance. A machine learning model retrained with the top five mutations performed similarly to the model using all 97 mutations. These results could aid in the development of molecular tests to detect resistance to ceftriaxone in N. gonorrhoeae. Moreover, this approach could be applied to building and evaluating machine learning models for predicting antimicrobial resistance in other pathogens.

Resistance-Guided Therapy for Neisseria gonorrhoeae

Antimicrobial-resistant Neisseria gonorrhoeae infections are a threat to public health. Novel strategies for combating such resistance include the development of molecular assays to facilitate real-time prediction of antimicrobial susceptibility. Resistance to ciprofloxacin is determined by the presence of a single mutation at codon 91 of the gyrase A gene; molecular assays to guide therapy are commercially available. Resistance to cefixime is conferred via 1 of 6 critical mutations in either the mosaic penA gene or specific loci in the nonmosaic region. Resistance to ceftriaxone is conferred through mutations in 1 of 4 genes: penA, ponA, penB, and mtr; however, the ability to predict reduced susceptibility based on those genes varies by geographic region. Here, we highlight the work done toward the development of 3 such assays for ciprofloxacin, cefixime, and ceftriaxone, discuss the status of our current understanding and ongoing challenges, and suggest future directions.

Reliability of Genetic Alterations in Predicting Ceftriaxone Resistance in Neisseria gonorrhoeae Globally

Antimicrobial resistance in N. gonorrhoeae is increasing globally, and ceftriaxone is the recommended treatment for empirical therapy in most settings. Developing molecular assays to detect decreased ceftriaxone susceptibility is critical. Using PathogenWatch, a public database of N. gonorrhoeae genomes, antibiotic susceptibility data and DNA sequences of different genes associated with ceftriaxone resistance were extracted. That information was used to determine the sensitivity and specificity of different molecular markers and algorithms to predict decreased susceptibility to ceftriaxone. A total of 12,943 N. gonorrhoeae genomes were extracted from the PathogenWatch database, of which 9,540 genomes were used in the analysis. The sensitivity and specificity of specific molecular markers and algorithms were largely consistent with prior reports. Small variation (<10%) in either sensitivity or specificity occurred. Certain algorithms using different molecular markers at various prevalence of decreased ceftriaxone susceptibility identified a potentially clinically useful range of positive and negative predictive values. We validated previously described mutations and algorithms in a large public database containing a global collection of N. gonorrhoeae genomes. Certain mutations and algorithms resulted in sensitivity and specificity values consistent with those of prior studies. Further research is needed to integrate these markers and algorithms into the development of molecular assays to predict decreased ceftriaxone susceptibility. IMPORTANCE Antimicrobial resistance in Neisseria gonorrhoeae (N. gonorrhoeae), the causative agent of gonorrhea, is rising globally. Ceftriaxone is the last remaining antibiotic for empirical treatment of gonorrhea. Developing molecular tests to predict ceftriaxone resistance can help to improve detection and surveillance of ceftriaxone resistance. Here, we utilized PathogenWatch, a public global online database of N. gonorrhoeae genomes, to evaluate different genetic markers in predicting decreased susceptibility to ceftriaxone. We compiled MICs for ceftriaxone from the PathogenWatch database and used a computational approach to extract all the genetic markers from the genomic data. We determined the sensitivity and specificity for predicting decreased ceftriaxone susceptibility among several combinations of genetic markers. We identified several combinations of genetic markers with high predictive values for decreased susceptibility to ceftriaxone. These combinations of genetic markers might be promising candidates for future molecular tests to predict ceftriaxone resistance.