gms | German Medical Science

Objectives: The differentiation between aseptic loosening and periprosthetic joint infection is still one of the most challenging tasks of revision arthroplasty. Research on infection-associated cell communication has given rise to today's routine laboratory investigations. Within the field of arthroplasty synovial extracellular vesicles (EVs), com-prise a yet uninvestigated intercellular communication pathway. The aim of this study was to test if prosthetic joint synovial fluid samples contain EVs, to develop a protocol to isolate EVs and then to characterize them with respect to their amount, size, surface markers and their differences between aseptic prosthetic loosening (APL) and periprosthetic joint infection (PJI).

Methods: The non-cellular fraction of prosthetic joint aspirates from either knee or hip were stored at -80°C. Patients were classified into two groups according to the 2018 definition of periprosthetic joint infection into PJI and APL, the later was confirmed by postoperative microbiological and pathological analysis. Ultracentrifugation was used as an EV isolation method. Electron microscopy confirmed the presents of EVs and nanoparticle tracking analysis (NTA) was used for EV size measurement and quantification. Analysis of EV surface proteins was achieved by using a bead-based multiplex technology (Exosome Kit, MACSPlex), which allowed the detection of 37 different surface epitopes. Clinical data such as age, gender and laboratory results were retrieved from the hospital patient file system.

Results and conclusion: 20 Patients (30% male, 70% female) were included. The mean age at arthrocentesis was 67.3±10.3 years, 6 hips and 14 knees were sampled, halve of the samples from prosthetic joints with ASL and halve from PJIs. Using our ultracentrifugation-based isolation protocol we were able to identify EVs in prosthetic joint synovial fluid. Between the PJI and ASL groups NTA showed significant higher particle concentration (p=0.0038) and significantly smaller particle size (p<0.0001) in the infected group, which was confirmed by electron microscopy. On the EV surfaces CD44 and HLA-DR/-DP/-DQ were significantly more expressed within the non-infected groups (p<0.0001, p=0.0005), while CD63 was significantly more expressed within the infected group (p<0.0001). Using our EV isolation protocol we confirmed the presents of EVs in prosthetic joint synovial fluid from the knee and hip. We visualized EVs by electron microscopy and characterized their size, amount and specific surface epitopes. Further, our results suggest differences between EVs from PJI and APL. This pilot study lays the foundation for further research and new investigation possibilities in the field of revision arthroplasty.