The goals of this observational study are:

1. to evaluate the concordance between traditional method (culture) and genomic method in diagnosing prosthetic joint infections;
2. to define an algorithm for laboratory diagnosis of prosthetic joint infections, which could be easily applied in our hospital setting;
3. to test different DNA extraction protocols to obtain high yields of microbial nucleic acids from infected human tissue samples.

Subjects A total of 70 patients from the Department of Reconstructive Surgery and Osteoarticular Infections (CRIO) and the Hip Units of IRCCS Ospedale Galeazzi -Sant'Ambrogio for suspected infection of their prosthesis will be enrolled in the study. Each sample will be analized with both traditional culture and Next Generation Sequencing (NGS).

On the basis of culture results patients will be divided into 3 groups.

Sample collection for Molecular analysis Sampling from peri-prosthetic tissue with macroscopic signs of infection will be taken from the synovial membrane, bone-prosthetic interface and intramedullary canals. Synovial fluid will be sterilely collected in DNAse and RNAse free containers for NGS analysis before performing the arthrotomy. Sterile swabs and pads will be used to collect samples from interface between tissue and bone implant and from intramedullary canal for NGS analysis. Bone tissue will be collected with new sterile instrumentation. Samples will be frozen within 2 hours from collection at -80°C until .

Sample collection for culture Sample collection of periprosthetic tissues and implant will be performed during the surgery after the proper exposure of the affected site. Synovial fluid will be collected in sterile screw cap tubes. Periprosthetic tissues and implant will be collected in sterile containers by using new sterile instrumentation. All samples will be sent to the Laboratory within 2 hours from collection. Samples after treatment with 0.1% (w:v) Dithiothreitol to dislodge bacteria from biofilm will be cultured on agar plates and broth for aerobes and anaerobes for 15 days. Microbial identification will be performed by biochemical assays.

A further tissue sample, collected in a separate sterile screw cap tube, will be used for host DNA removal to isolate microbial DNA, for shotgun NGS. Samples will be treated with saponin and/or different commercial kits for microbial DNA extraction. The dsDNA will be quantified with a Qubit fluorometer, and its purity will be determined with a nanodrop spectophotometer.

Synovial fluid analysis Synovial fluid for leukocyte count and differential will be collected in EthylenDiaminoTetraAcetic Acid (EDTA) containing tubes and analyzed by means of a Sysmex XN haematological analyser.

Determination of synovial protein, glucose, C-Reactive Protein (CRP) and leukocyte esterase will be performed on synovial fluid collected in clot activator containing tubes. Measurement of proteins, glucose and CRP will be carried out on Atellica CH and IM analyser (Siemens) while leukocyte esterase was determined by used of strips for urinalysis.

Molecular analysis 70 samples will be sent to a specialized laboratory which will perform nucleic acid extraction from biological specimen types using tailored proprietary techniques that include concentrating cellular material, mechanical lysis, and removing organic inhibitors and human host DNA. Every specimen will be processed with a positive and negative extraction control to assure quality.

At the end of the study any residual biological material will be destroyed

Statistical methodology Sample size The sample size has been calculated considering as the primary outcome the agreement of the innovative molecular method and the traditional culture of periprostethic tissue and synovial fluid in defining a patients as positive or negative for prosthetic joint infections . Based on investigators experience, the proportion of patients, undergoing revision surgery for knee and hip implant failure in IRCCS Ospedale Galeazzi Sant'Ambrogio with diagnosis of prosthetic joint infection based on cultural exam is of 45%. Considering a margin of error of 15%, an alpha = 0.05 and power = 80%, the number of patients to be tested is 70.

Analysis methodology Categorical data will be described as absolute and relative frequencies, while continuous data will be described as mean and 95%-confidence interval, if normally distributed, or as median and interquartile range, in case of non-normal distribution. Normal distribution will be assessed by the Shapiro-Wilk test.

Differences in proportion will be evaluated by Fisher's exact test or Chi-squared test; continuous variable comparison among the three groups will be performed using One-Way Analysis Of Variance (ANOVA) test for normally distributed data or with Kruskal-Wallis test, in case of non-normal distribution.

Agreement between molecular methods with traditional culture of periprostethic tissue and synovial fluid will be tested using Choen's kappa.

Sensitivity and specificity of each assay will be also assessed.

Inclusion Criteria:

• Age >18 years
• Males and females
• Patients with a clinical indication for knee and hip revision surgery admitted at IRCCS Ospedale Galeazzi - Sant'Ambrogio

Exclusion Criteria:

• Insufficient amount of periprosthetic tissue and synovial fluid collected
• Patients unable to sign the Informed Consent
• Pregnancy and breast feeding with declaration of patient
• Age <18 years

A total of 70 patients from the Department of Reconstructive Surgery and Osteoarticular Infections (CRIO) and the Hip Units of IRCCS Ospedale Galeazzi -Sant'Ambrogio for suspected infection of their prosthesis will be enrolled in the study.

Only data useful for publication will be made avalilable

IPD will be avalable form the end of the study