Assessing the Performance of Shotgun Metagenomics in the Diagnosis of Complex Prosthetic Joint Infections

Prosthetic joint infection (PJI) is one of the most serious and devastating complications of orthopaedic surgery, leading to a high risk of recurrence and disability, as well as increased mortality and management costs. Despite improvements in antibiotic prophylaxis procedures and surgical asepsis measures, the significant increase in the number of prostheses fitted worldwide has been accompanied by an increase in the number of infections. The infection rate has been estimated at between 1% and 2% after hip and knee arthroplasty.

Appropriate diagnosis and medical and surgical management of PJI are therefore essential to preserve and/or restore adequate motor function, minimise the risk of complications and prevent excessive morbidity. The microbiological diagnosis of PJI must be as early and exhaustive as possible in order to introduce rapid and effective antibiotic therapy and avoid the development of a biofilm (gangue around the material) or chronic infection (quiescent bacteria).

However, the diagnosis of PJI can be difficult to make in certain situations. Learned societies have established a definition of PJI and defined diagnostic scores combining clinical, biological, anatomopathological and cytological criteria. An initial definition was approved in 2011 by the Musculoskeletal Infection Society (MSIS). This definition was modified and subject to an international consensus review in 2013 (MSIS diagnostic score). In 2018, an international consensus meeting reviewed and adapted the MSIS score. This adapted score is more appropriate to current Medical Biology practices and to the non-accessibility of all diagnostic tests in laboratories (leucocyte esterase, alpha-defensin, ...).

In this definition of PJI, the positivity of 2 intra-operative samples to the same bacterial species is considered to be a major criterion. A wide range of bacteria can cause PJI: aerobic/anaerobic/intracellular/mycobacterial; somePJI can be polymicrobial. It is therefore essential to accurately identify these pathogens in order to administer appropriate antibiotic therapy and avoid chronicity of infection. Despite the optimisation of practices, culture of samples is negative in 5 to 30% of cases, despite the presence of diagnostic criteria for PJI. The most common causes are a lack of culture sensitivity, prior antibiotic administration and/or the presence of difficult or slow-growing pathogens. In these cases, intravenous broad-spectrum antibiotic therapy is administered, resulting in additional management costs, the occurrence of adverse treatment effects and the risk of acquiring resistance or intestinal dysbiosis.

In this context, "classic" molecular techniques are routinely used to overcome the limitations of culture for microbiological detection: bacterial-specific (including PCR targeting Staphylococcus aureus) or non-specific (bacterial universal PCR targeting the gene encoding 16S rDNA) (Figure 1). The latter approach was previously evaluated by the CRIOGO group (3Centre de Référence en Infections Ostéo-articulaires du Grand Ouest") with detection performance deemed disappointing in the context of PJI (sensitivity of 73.3%, specificity of 95.5%). Innovative molecular techniques for Next Generation Sequencing (NGS) are being developed, including shotgun metagenomics (sequencing of all the genetic material in a sample). Recent studies have evaluated the sensitivity of shotgun metagenomics in PJI, estimated at between 90.2% and 93.0% compared with bacterial culture and at around 95% compared with the MSIS diagnostic score.

However, these few recent studies evaluating shotgun metagenomics have only been carried out on a single sample per patient, which is insufficient according to the recommendations of the international and national consensuses on the management of PJI. In fact, four or even five intraoperative samples must be taken and analysed in microbiology to make the diagnosis of PJI. This high number of samples improves the sensitivity and completeness of bacterial detection and facilitates the interpretation of positive cultures for potentially contaminating skin bacteria (coagulase-negative Staphylococci, Cutibacterium acnes, etc.). To date, only one study has assessed the performance of shotgun metagenomics applied to several intraoperative samples per patient. Further studies are therefore needed to refine the performance of shotgun metagenomics in the context of PJI and to better assess the contribution of this costly technique, which requires considerable expertise to perform and interpret.

The setting up of a prospective, multicentre study in centres associated with the CRIOGO will make it possible to assess the performance of shotgun metagenomics in the management of chronic PJI. The performance of shotgun metagenomics will be assessed on the basis of four different samples per patient, in six centers specialising in the diagnosis of PJI, which makes the METAGENOS study unique compared with other studies. At the end of the project, the aim is to define the indications for using this innovative technique and to harmonise future regional practices.