Synovial C-reactive protein features high negative predictive value but is not useful as a single diagnostic parameter in suspected periprosthetic joint infection (PJI).

OBJECTIVES: Synovial fluid C-reactive protein (syCRP) has been recently described as a new biomarker in preoperative diagnostics to identify periprosthetic joint infections (PJI). The aim of this study was to evaluate syCRP in a large cohort of patients with suspected PJI and to calculate the optimal cut-off to diagnose PJI. METHODS: Between September 2015 and June 2017, we prospectively included patients with suspected PJI, in which syCRP was additionally measured along with routine preoperative diagnostic serum and synovial biomarkers. We analysed the sensitivity and specificity of syCRP using receiver operating characteristic curves. RESULTS: We included 192 cases (hip n = 80, knee n = 91, shoulder n = 21) with a final diagnosis of PJI in 26 cases (14.0%). Combined for all joints, the syCRP values were significantly higher in the PJI group than in the no PJI group (median: 13.8 vs. 0 mg/l; p < 0.001). The optimal cut-off (Youden Index: 0.71) for the PJI diagnosis combined for all joints was at a syCRP value of 2.9 mg/l with a sensitivity of 88%, a specificity of 82%, and a negative predictive value of 98%. CONCLUSIONS: SyCRP features high negative predictive value but is not useful as a single diagnostic parameter in suspected periprosthetic joint infection (PJI).

[Quantitative detection and elimination of neuroblastoma cells in vitro. Comparative study of the efficacy of an immunomagnetic method and lymphokine-activated killer cells]. / Détection quantitative et élimination des cellules de neuroblastome in vitro.Etude comparative ente l'efficacité d'une méthode immunomagnetique et celle des cellules LAK.

Much effort is devoted to eliminating residual tumor cells - which escape detection by conventional methods - from remission bone marrow harvested for autologous transplantation. The quantitative determination of the efficacy of these purging methods is generally difficult. This report describes an in vitro reproducible tumor model. Normal mononuclear blood cells were deliberately contaminated with cells from the human neuroblastoma cell line SK-N-AS. The frequency of clonogenic SK-N-AS cells was determined in limiting dilution culture before and after treatment. Two methods of purging these tumor cells from the mixed cell suspension were compared, 1) an immunomagnetic method taking advantage of the existence of a monoclonal antibody (mAb) specific for 90-95% of the tumor cells, and 2) a method based on cell mediated cytotoxicity testing the activity of previously lymphokine activated killer (LAK) cells. Immunomagnetic purging eliminated 90% (1 log) of all SK-N-AS cells and 99% (2 log) of the mAb-binding clonogenic tumor cells. In contrast, LAK cell treatment removed only between 0.1 to 0.3 log (23-50%) of the clonogenic SK-N-AS cells.