PROTEOMIC PROFILING OF SYNOVIAL FLUID IN SEPTIC AND ASEPTIC REVISION ARTHROPLASTY: INSIGHTS INTO MOLECULAR SIGNATURES OF PERIPROSTHETIC JOINT INFECTION

Aim Despite advances in synovial biomarkers, the molecular profile distinguishing septic from aseptic revision arthroplasty in periprosthetic joint infection (PJI) remains poorly defined. This study aimed to characterize synovial fluid proteomic signatures of septic and aseptic revision total knee arthroplasty (TKA) compared with native-joint controls, and to identify candidate diagnostic and pathophysiological biomarkers of PJI. Methods A retrospective proteomic analysis was conducted on synovial fluid samples from 57 patients who underwent revision TKA between 2021 and 2026: septic revision (n=25), aseptic revision (n=20), and native-joint controls (n=12). Revisions were classified according to EBJIS PJI criteria. Synovial fluid proteomes were analyzed using data-independent acquisition mass-spectrometry. Differential protein abundance analyses were performed for control vs septic, control vs aseptic, and septic vs aseptic comparisons. Subgroup analyses compared high-and low-virulence septic revisions, and primary versus re-revision TKA. Proteins with log 2 fold change ≥ 1 and p-value ≤0.01 were considered significant. Results In total, 2514 synovial proteins were quantified, with septic revisions showing the highest proteomic complexity. In the septic vs aseptic comparison, 978 proteins were upregulated in septic and 83 in aseptic revisions. Beyond previously published PJI marker candidates, septic PJI showed several additional significantly upregulated proteins with log 2 fold change ≥4 (including AZU1, CORO1A, SERPINB1, NAMPT, GCA and CAMP; all p<0.001), predominantly associated with neutrophil activation and antibacterial defense. The three most upregulated proteins in aseptic revisions (MGP, DKK3, and PRG4; log 2 fold change ≥ 2; p < 0.001) are linked to bone and cartilage matrix biology. In the control vs aseptic comparison, a similar extracellular matrix and tissue-remodeling profile was observed with increased expression (log2fold change ≥2) of 87 proteins, including SPP1, POSTN, CTSB, CTSD, SERPINE1, SPARC, COL1A1, and MXRA5 (all p<0.001), suggesting an active matrix-remodeling phenotype rather than a non-infected state. Virulence-stratified analyses identified 16 and 14 proteins upregulated in high-and low-virulence infections, respectively (log 2 fold change ≥1), with markedly higher levels of the major acute-phase protein SAA1 in high-virulence PJI (p<0.001). Re-revision TKA appeared to be associated with increased synovial COL12A1 ((log 2 fold change >1) compared to primary revisions. Conclusions Synovial fluid proteomics robustly distinguishes septic from aseptic revision TKA, confirming known PJI candidate proteins and revealing novel infection-specific and aseptic remodeling signatures. Septic PJI involves a neutrophil, antimicrobial, acute-phase response, while aseptic failure involves extracellular matrix and fibrotic remodeling proteins. Data highlight new candidate biomarkers, including those for pathogen virulence, for future validation and development of proteomics-based diagnostic panels.