Diabetic foot wound infections are predominantly polymicrobial. However, 'conventional' microbiological culture does not identify all the bacteria potentially involved in these infections and requires time, which can have a negative impact by delaying treatment and/or the prescription of appropriate antibiotic therapy.

Real-time metagenomics analysis using Oxford Nanopore Technologies' MinION technology has demonstrated sufficient power to identify virtually all microbial genomes in a given sample, providing additional information on their antibiotic resistance profile and in silico prediction of genes encoding virulence factors within than 4 hours. Based on these rapid results, a management protocol could be defined specifically for each patient with a view to personalised medicine.

The aim is to study the diversity of bacterial and fungal species identified using the MinION method and compare this diversity with the results obtained using conventional methods (routine culture) from bone biopsies taken from DFOM patients.

Diabetes mellitus is one of the most common diseases in the world. Among its complications, 34% of patients living with diabetes will develop foot ulcers during their lifetime. Once this lesion has developed due to triopathy (arteriopathy, neuropathy, immunopathy), over 50% of these wounds will become infected, leading to serious consequences: bone damage (60 to 80% of cases), amputation (20% of infected wounds), mortality (68% at 5 years) and morbidity. The additional costs associated with this condition exceed $850 billion worldwide, with $1 million spent every 30 seconds due to complications from these wounds in the US. Diabetic foot ulcers (DFUs) are therefore a major public health problem.

For infectious disease specialists, one of the main difficulties in treating infections in these wounds is distinguishing between bacterial colonization - a physiological process, and infection - a pathological process. Studies of the microbiota of foot wounds in diabetic patients show the polymicrobial nature of these lesions, which contain commensal bacteria from the skin microbiota, opportunistic pathogenic bacteria and bacteria from the environment. Among Gram-positive cocci, Staphylococcus aureus is the species most often identified in diabetic foot osteomyelitis (DFO) (23.4%), followed by Pseudomonas spp. (11.1%), Escherichia coli (11.5%), Proteus spp. (8.3%), Klebsiella spp. (6.9%), and Enterococcus spp. (5.4%). Coagulase-negative staphylococci, although involved in less than 4% of infections, often remain difficult to identify by mass spectrometry at species level, even though some of them are known for their pathogenicity. The other major difficulty is the time it takes to obtain results when osteomyelitis is suspected. In such situations, obtaining a bone sample is the standard method and the best way to identify the pathogen(s) responsible and their sensitivity to antibiotics. Bone biopsy can be performed intraoperatively or percutaneously, as recommended by the International Working Group of the Diabetic Foot (IWGDF) in 2023. In addition, to avoid false-negative cultures, experts suggest delaying bone biopsy in patients already on antibiotics, ideally for at least two weeks. In the microbiology laboratory, the standard diagnosis based on conventional microbiological culture can take up to two weeks to identify the causative bacteria and perform antibiotic susceptibility testing on the responsible pathogen(s), bringing the total time to establish a diagnosis to 4 weeks.To reduce this delay, certain culture-free molecular microbiology techniques, in particular metagenomic next-generation sequencing (mNGS), have shown that the microbiota of most wound infections is more diverse and abundant than that revealed by conventional culture methods. However, very few metagenomic data on bone biopsies from DFOMs are available. As molecular biology tools were unable to distinguish between live and dead bacterial cells and cannot identify the bacterial genera contributing to the clinical status of the infection, their use in daily practice is not recommended by the IWGDF and SPILF (Société de Pathologie Infectieuse de Langue Française), as their results could lead to the unnecessary use of broad-spectrum antibiotics. Recent studies based on metagenomic analysis of bone biopsies, soft tissue biopsies and swabs from foot wounds in diabetic patients have identified microbial diversity as a marker of infection. The number of bacteria involved in confirmed cases of infection is estimated at over 70, a figure difficult to achieve with conventional routine methods. The use of PCR based on amplification of the gene coding for 16S rRNA is considered unsuitable due to the frequent polymicrobial nature of DFOM samples. Multiplex PCR methods are not exhaustive in identifying all pathogens. Finally, conventional culture methods are often time-consuming, and species identification using laboratory methods such as MALDI-TOF mass spectrometry is often a source of confusion or failure. Recently, a new rapid sequencing tool has been developed: the MinION. It is small and fast and can sequence a bacterial or viral genome from single-microbe samples in less than four hours, or determine a panel of microorganisms present in a more complex sample. This tool is particularly useful for cerebrospinal fluid.

Inclusion Criteria:

• N/A

Exclusion Criteria:

• N/A

147 bone biopsies from 43 diabetic patients collected in the orthopedic surgery operating room