Can the Current Thresholds for Synovial Cell Count and Neutrophil Percentage to Diagnose Prosthetic Joint Infection be Applied to Metal-on-Metal Rotating Hinge Total Knee Arthroplasty?

Background: Synovial leukocyte count analysis is an important tool in the diagnosis of PJI. However, results can be poor if metallosis is present. The issue of metallosis was established for some rotating hinge knee arthroplasty designs. Methods: This was a retrospective single-center analysis that included 108 patients who had a singledesign metal-on-metal rotating hinge knee arthroplasty who underwent revision surgery and had prior synovial fluid analysis performed. The diagnostic accuracy of leukocyte count and the percentage of polymorphonuclear neutrophiles (% PMN) were investigated. Results: Patients who had a PJI had a higher median leukocyte count and % PMN compared to aseptic revisions (7,905/mL (interquartile range (IQR) 2,419 to 37,400) and 85% (IQR 70.3 to 93.8) versus 450 (IQR 167 to 1,215) and 46% (IQR 28.5 to 67.4%), P < .001). The respective areas under the curves were 0.916 (95% confidence interval 0.862 to 0.970) for leukocyte count and 0.821 (95% confidence interval 0.739 to 0.902) for % PMN. We calculated an optimal cut-off value of 1,200 leukocytes/mL (Sensitivity 94.5%/ specificity 75.5%) and 63% PMN (Sensitivity 85.5%/specificity 73.6%) to define PJI. Established thresholds were less sensitive, but more specific. The "infection likely" scenario of the European Bone and Joint Infection Society (EBJIS) definition was closest to the calculated thresholds. Conclusion: Currently used thresholds for leukocyte cell count and %PMN to define PJI were less sensitive and specific in the diagnosis of PJI in metal-on-metal RHK knees. Surgeons must consider lower cut-off values when evaluating such implants for PJI.