Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Objectives: The management of bloodstream infection, a life-threatening disease, largely relies on early detection of infecting microorganisms and accurate determination of their antibiotic susceptibility to reduce both mortality and morbidity. Recently we developed a new technique based on atomic force microscopy capable of detecting movements of biologic samples at the nanoscale. Such sensor is able to monitor the response of bacteria to antibiotic's pressure, allowing a fast and versatile susceptibility test. Furthermore, rapid preparation of a bacterial pellet from a positive blood culture can improve downstream characterization of the recovered pathogen as a result of the increased bacterial concentration obtained. Methods: Using artificially inoculated blood cultures, we combined these two innovative procedures and validated them in double-blind experiments to determine the susceptibility and resistance of Escherichia coli strains (ATCC 25933 as susceptible and a characterized clinical isolate as resistant strain) towards a selection of antibiotics commonly used in clinical settings. Results: On the basis of the variance of the sensor movements, we were able to positively discriminate the resistant from the susceptible E. coli strains in 16 of 17 blindly investigated cases. Furthermore, we defined a variance change threshold of 60% that discriminates susceptible from resistant strains. Conclusions: By combining the nanomotion sensor with the rapid preparation method of blood culture pellets, we obtained an innovative, rapid and relatively accurate method for antibiotic susceptibility test directly from positive blood culture bottles, without the need for bacterial subculture. (C) 2017 The Authors. Published by Elsevier Ltd.
Sandor Kasas, Anthony Vocat, Gino Cathomen, Eric Delarze, Florian Tagini, Laura Munch
, , ,
Camille Véronique Bernadette Goemans, Florian Huber