Minimizing treatment-induced emergence of antibiotic resistance in bacterial infections.

Abstract

Currently, treatment of bacterial infections focuses on choosing an antibiotic which matches a pathogen’s susceptibility, with less attention to the risk that even susceptibility-matched treatments can fail due to resistance emerging in response to treatment. Here, combining whole-genome sequencing of 1,113 pre- and post- treatment bacterial isolates with machine-learning analysis of 140,349 urinary tract infections (UTIs) and 7,365 wound infections, we find that treatment-induced emergence of resistance could be predicted and minimized at the individual-patient level. Emergence of resistance was common, yet driven not by de novo resistance evolution, but rather by rapid reinfection with a different strain resistant to the prescribed antibiotic. As most infections are seeded from the patient’s own microbiota, these resistance-gaining recurrences can be predicted based on the patient’s past infection history, and their expected risk minimized by machine learning personalized antibiotic recommendations, offering a means to reduce the emergence and spread of resistant pathogens.