OBJECTIVE:The synovial tissue pathotype may determine the treatment response in rheumatoid arthritis (RA); however, biopsies are not widely available. Synovial fluid is a promising tissue surrogate. Our purpose was to compare RA synovial fluid cell counts with histopathology and use synovial fluid to predict tissue inflammation. METHODS:Synovial fluid and tissue were collected during knee arthroplasty. Patients were stratified based on their medication treatment history. Synovial lymphocytic inflammation (SLI) was graded from low to high. Synovial fluid white blood cell (WBC) count and differentials were performed in the clinical laboratory. Descriptive statistics, correlations, receiver operating characteristic curve analysis, and multivariable regression were performed to determine the associations with tissue SLI. RESULTS:Sixty-four patients with RA had paired synovial tissue and synovial fluid data available. The mean Clinical Disease Activity Index (CDAI) score was 17.9. High tissue SLI was observed in 29 patients, and low SLI was observed in 35 patients, with roughly equal distribution among treatment groups. The mean synovial fluid WBC count was 5,661 cells/μL and was not correlated with CDAI but correlated positively with SLI and percentage polymorphonuclear cells (PMN%). Synovial fluid WBC count ≥1,400 cells/μL was sensitive (0.86) and specific (0.91) for high SLI (area under the curve 0.91). In a multivariable regression, PMN% was associated with high SLI (odds ratio [OR] 1.46 [95% confidence interval (CI) 1.14-1.85]). Synovial fluid monocyte percentage was negatively associated with high SLI (OR 0.44 [95% CI 0.27-0.73]). CONCLUSION:Synovial fluid WBC count is sensitive and specific for differentiating high and low lymphocytic synovial inflammation. Further analysis of the synovial fluid as it relates to the adjacent tissue in different cohorts is needed.