Low Synovial Lymphocyte Percentage as a Novel Diagnostic Marker of Periprosthetic Joint Infection.

Objectives: Correctly diagnosing periprosthetic joint infection (PJI) remains a major clinical challenge. While acute PJI is straightforward to identify because of its pronounced clinical presentation, chronic infections remain challenging to detect since clinical signs are subtle and standard criteria lack sensitivity. By analyzing a wide range of serum and synovial parameters in patients undergoing revision arthroplasty, we sought to identify the most accurate diagnostic PJI markers. Methods: A retrospective analysis of 400 patients undergoing knee or hip revision arthroplasty, including 145 patients who had PJI and 255 patients who had aseptic failure, was conducted. Diagnosis of PJI was established using the 2018 International Consensus Meeting criteria. A comprehensive evaluation of medical records, serum, and synovial biomarkers was performed. For each marker, receiver operating characteristic curves, calculating the area under the curve and optimal cutoff values, were determined. Results: Synovial biomarkers such as alpha-defensin and C-reactive protein did not demonstrate superior diagnostic performance compared with polymorphonuclear neutrophil (PMN) count and percentage. Importantly, low synovial lymphocyte percentage (<34.7%) emerged as the most accurate marker for PJI diagnosis (area under the receiver operating characteristic curve [AUC] = 0.96, sensitivity = 0.85, specificity = 0.96), independent of location or infection subtype (acute and chronic). This was further supported by a conditional inference tree model for diagnosing chronic PJI, which identified synovial lymphocyte percentage and PMN count as key decision nodes and demonstrated excellent diagnostic performance (AUC = 0.95; sensitivity = 0.93; specificity = 0.91). Conclusions: Our study provides evidence that low synovial lymphocyte percentage is a reliable diagnostic marker of PJI. Integrating the assessment of synovial lymphocytes into clinical practice could enable more timely diagnosis and, therefore, effective treatment, ultimately improving patient outcomes. However, as this was a single-center study limited to hip and knee revision, external validation is warranted to confirm the generalizability of our findings. Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.