Synovial Cell Count Before Reimplantation Can Predict the Outcome of Patients with Periprosthetic Knee Infections Undergoing Two-stage Exchange

BACKGROUND: Although synovial fluid can be used to diagnose periprosthetic joint infections (PJI) effectively, only the cutoff values adopted at the time of PJI diagnosis have been standardized, and few data are currently available about effectiveness of synovial fluid examination before definitive reimplantation. QUESTIONS/PURPOSES: We asked: (1) What are the most appropriate thresholds for synovial fluid leukocyte counts (WBC) and neutrophil percentage (PMN percentage) in a patient group undergoing definitive reimplantation after an uninterrupted course of antibiotic therapy for chronic PJI? (2) What is the predictive value of our synovial WBC and PMN percentage threshold compared with previously proposed thresholds? METHODS: In all, 101 patients with PJI were evaluated for inclusion from January 2016 to December 2018. Nineteen percent (19 of 101) of patients were excluded because of the presence of a chronic inflammatory disease, acute/late hematogenous infection, low amount of synovial fluid for laboratory investigations or infection persistence after spacer placement, and adequate antibiotic therapy. Finally, 81% (82 of 101) of patients with a median (range) age of 74 years (48 to 92) undergoing two-stage revision for chronic TKA infection, who were followed up at our institution for a period 96 weeks or more, were included in this study. The patients did not discontinue antibiotic treatment before reimplantation and were treated for 15 days after reimplantation if intraoperative cultures were negative. No patient remained on suppressive treatment after reimplantation. Synovial fluid was aspirated aseptically with a knee spacer in place to evaluate the cell counts before reimplantation. Thirteen percent (11 of 82) of patients had persistent or recurrent infection, defined as continually elevated erythrocyte sedimentation rate or C-reactive protein levels coupled with local signs and symptoms or positive cultures. The synovial fluid WBC counts and PMN percentage from the 11 patients with persistent or recurrent PJI were compared with the 71 patients who were believed to be free of PJI. Receiver operating characteristic (ROC) curve analyses assessed the predictive value of the parameters, and the areas under the curves (AUCs) were evaluated. The sensitivities, specificities, and positive and negative predictive values were determined for the WBC count and PMN percentage. Patients with persistent or recurrent infection had higher median WBC counts (471 cells/µL versus 1344 cells/µL; p < 0.001) and PMN percentage (36% versus 61%; p < 0.001) than did patients believed to be free of PJI. RESULTS: ROC curve analysis identified the best threshold values to be a WBC count of 934 cells/µL or more (sensitivity 0.82 [95% CI 0.71 to 0.89], specificity 0.82 [95% CI 0.71 to 0.89]) as well as a PMN percentage of at least 52% (sensitivity 0.82 [95% CI 0.71 to 0.89] and specificity 0.78 [95% CI 0.67 to 0.86]. We found no difference between the AUCs for the WBC count and the PMN percentage (0.87 [95% CI 0.79 to 0.96] versus 0.84 [95% CI 0.73 to 0.95]. Comparing the sensitivities and specificities of the synovial fluid WBC count and PMN percentage proposed by other authors, we find that a PMN percentage more than 52% showed better predictive value than previously reported. CONCLUSION: Based on our findings, we believe that patients with WBC counts of at least 934 and PMN percentage of 52% or more should not undergo reimplantation but rather a repeat debridement, as their risk of persistent or recurrent PJI appears prohibitively high. The accuracy of the proposed cutoffs is better than previously reported. LEVEL OF EVIDENCE: Level III, diagnostic study