Simultaneous detection of genotype and phenotype enables rapid and accurate antibiotic susceptibility determination

Abstract

Multidrug resistant organisms (MDROs) are a serious threat to human health 1,2. Fast, accurate antibiotic susceptibility testing (AST) is a critical need in addressing escalating antibiotic resistance, since delays in identifying MDROs increase mortality 3,4 and use of broad-spectrum antibiotics, further selecting for resistant organisms. Yet current growth-based AST assays, such as broth microdilution 5 , require several days before informing key clinical decisions. Rapid AST would transform the care of infected patients while ensuring that our antibiotic arsenal is deployed as efficiently as possible. Growth-based assays are fundamentally constrained in speed by doubling time of the pathogen, and genotypic assays are limited by the ever-growing diversity and complexity of bacterial antibiotic resistance mechanisms. Here, we describe a rapid assay for combined Genotypic and Phenotypic AST through RNA detection, GoPhAST-R, that classifies strains with 94-99% accuracy by coupling machine learning analysis of early antibiotic-induced transcriptional changes with simultaneous detection of key genetic resistance determinants to increase accuracy of resistance detection, facilitate molecular epidemiology, and enable early detection of emerging resistance mechanisms. This two-pronged approach provides phenotypic AST 24-36 hours faster than standard workflows, with <4 hour assay time on a pilot instrument for hybridization-based multiplexed RNA detection implemented directly from positive blood cultures.

Machine Learning and Multidrug-Resistant Gram-Negative Bacteria: An Interesting Combination for Current and Future Research

“…It is also important to note that we limited our review to ML algorithms trained on clinical (patients' characteristics and recorded clinical features) and laboratory data (phenotypical identification and susceptibility test) usually available in routine clinical practice in most hospitals. In this regard, the possible dissemination and appropriate use of innovative tests able to provide information on either the precise mechanisms of resistance or other phenotypic/genotypic features of MDR-GNB may further and considerably improve the ability of ML algorithms to help clinicians predicting MDR-GNB risks in routine care [32,33]. Very importantly, this should occur against a background of wide adherence to the FAIR principles (findable, accessible, interoperable, and reusable) connected to the availability of standardized systems [34][35][36][37][38].…”

Targeted surveillance strategies for efficient detection of novel antibiotic resistance variants

“…For example, under the general assumption that novel resistance variants are more likely to appear in underrepresented lineages, phylogeny-aware surveillance could be paired with a diagnostic approach such as genomic neighbor typing 56 , where any isolates with either susceptible or low confidence calls that appear to be divergent from the genomes in the reference database would be prioritized for confirmatory phenotyping. Similarly, a diagnostic that predicts AMR phenotypes through a combination of transcriptomic and genomic typing 63 may facilitate targeted surveillance by identifying isolates with ambiguous predictions ( e.g. , isolates with transcriptional signatures of resistance that lack known genomic markers of resistance) that could be prioritized for confirmatory phenotyping.…”

Differential DNA accessibility to polymerase enables 30-minute phenotypic β-lactam antibiotic susceptibility testing of carbapenem-resistant Enterobacteriaceae

“…We note that many stateof-the-art phenotypic AST methods are initially published without validation of performance directly on clinical samples, e.g. a recent breakthrough demonstrating phenotypic AST on isolates and on blood cultures [58]. Urine is a relevant matrix for a CRE diagnostic because UTIs are the most common source of CRE isolates [76], and because of the large number of hospital-acquired infections that involve catheters or other long-term indwelling medical devices [11], where CRE infections cause major problems.…”

“…Rapid phenotypic methods based on quantification of nucleic acids (NAs) have shown great promise for a rapid POC AST due to the speed, specificity, and robustness of NA detection [53][54][55][56][57][58]. There is an additional advantage to using NA quantification as a readout of the bacterial response to antibiotic: because rapid pathogen ID from clinical samples is commonly performed via NA analysis, it would likely be easier to integrate an NA-based phenotypic AST into a combined ID/AST workflow performed from the same clinical sample.…”