picture_as_pdf Download as PDF
Session: Session 1

High throughput identification of the most prevalent sepsis causative micro-organisms by a Scout-triggered proteomic assay directly from positive blood culture

Maud GREGSON1, Romain CARRIÈRE1, Iulia MACAVEI1, Tiphaine CECCHINI2, Chloé DESBIOLLES2, Roxane PRAT2, François VANDENESCH2, Jérôme LEMOINE1

1Universite Claude Bernard Lyon1, ISA, UMR5280, CNRS, Villeurbanne, France
2Hospices Civils de Lyon, Lyon, France

Introduction

Ultra-fast and accurate identification (ID) of sepsis causative micro-organisms directly from positive blood culture (PBC) is a key step in patient treatment. Even if MALDI-TOF MS is the standard of care, it has some limitations regarding ID performance for polymicrobial samples and differentiation inside some species complexes. A multiplexed proteomic assay using liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) covering 113 most prevalent species involved in bacteraemia has been developed and tested against artificial PBC and clinical routine samples.

 

Material and method

Upon positivity, after several sample preparation steps including the cells lysis and the proteins digestion by the trypsin, peptides mixtures were analysed by LC-MS/MS within a Scout-triggered MRM acquisition method. The total workflow lasts less than 45 min.

To validate this workflow, 3282 bacterial strains were spiked in BC bottles with human sterile blood and incubated in a BACT/ALERT VIRTUO. In a second time, 169 patient PBC from Lyon Hospital were blindly selected. Species identification given by MALDI-TOF MS was compared with the one found with the developed method.

 

Results

From spiked PBC, rates of 100% (3282/3282) of specificity and 93.8% (3080/3282) for the sensitivity were reached. Some indistinguishable species inside the Enterobacter and the Klebsiella complexes have been differentiated with the developed LC-MS/MS method.

From patient PBC, successful ID was obtained for 93% (153/169) of the samples on Day 0 and in total agreement with MALDI-TOF MS identification (obtained on Day 1). Among this cohort, 10 samples were polymicrobial and for 6 of them, the two pathogens were correctly identified.



Conclusion

We aimed to develop a new diagnostic tool, simple to use, rapid and ready to be implemented in routine hospital services. The Scout MRM-MS method enables the ID of sepsis-causative pathogens with performances equalling or exceeding those of MALDI-TOF MS.