PURPOSE OF THE STUDY Laboratory methods are central to prosthetic joint infection (PJI) diagnosis. Most research teams focus on detection of specific inflammatory markers, causative pathogens, or on assessment of the tissue response. This study sought to determine the optimal cut-off values and diagnostic performance of selected synovial markers in relation to the diagnosis of hip or knee PJI. The studied markers were synovial level of glucose, lactate, coefficient of energy balance (CEB) and NGAL (neutrophil gelatinase-associated lipocalin). MATERIAL AND METHODS This prospective study includes 89 patients who underwent revision total knee or hip arthroplasty for septic or aseptic reasons in the period from 2014 to 2017. Among these 89 patients, there are 2 cases of prosthetic hip infection, 22 cases of prosthetic knee infection, 31 aseptic revision total hip arthroplasties and 34 aseptic revision total knee arthroplasties. The diagnostic characteristics of the studied methods were set in relation to the reference standard, the 2013 MSIS (Musculoskeletal Infection Society) criteria. The cut-off values were calculated using the ROC (receiver operating characteristic curve) analysis. RESULTS The synovial glucose test is considered positive if the glucose level drops below 2.65 mmol/L. The area under the curve is 0.813, sensitivity 75.0%, specificity 83.1%. The synovial lactate test is considered positive if lactate level rises above 8.87 mmol/L. The area under the curve is 0.882, sensitivity 70.8%, specificity 95.4%. Synovial NGAL is considered positive if its level exceeds 998 μg/L. The area under the curve is 1.000, sensitivity 100.0%, specificity 100.0%. CEB is considered positive if its value is lower than +4.665. The area under the curve is 0.883, sensitivity 91.7% and specificity 69.8%. Combining of these tests with other synovial markers does not improve the diagnostic performance of the studied tests. CONCLUSIONS The glucose and lactate levels and CEB undoubtedly reflect the presence of an inflammatory process in a prosthetic joint. However, the diagnostic characteristics of these tests are not better than those of other modern diagnostic techniques. As opposed to these tests, synovial NGAL shows excellent diagnostic performance. Nonetheless, the potential of this method shall be verified on larger cohorts of patients. Key words: prosthetic joint infection, periprosthetic infection, total knee arthroplasty, total hip arthroplasty, diagnosis, glucose, lactate, CEB, NGAL.

Author