CD15 focus score for diagnostics of periprosthetic joint infections. Neutrophilic granulocytes quantification mode and the development of morphometric software (CD15 quantifier)

The aim of this project was to devise a quantification method for neutrophils within a single focal point through the development of a CD15 focus score which enables bacterial infections in synovial-like interface membranes (SLIM) to be diagnosed. In this study a histopathological classification of 91 SLIM removed during revision surgery from the hips (n = 59) and knees (n = 32) was performed. Neutrophils were identified immunohistochemically by means of a CD15-specific monoclonal antibody. The quantitative evaluation of CD15-positive neutrophils (CD15Ne) used the principle of maximum focal infiltration (focus) together with an assessment of a single focal point (0.3 mm(2)). This immunohistochemical approach made it possible to develop the CD15 quantifier software, which automatically quantifies CD15Ne. The SLIM cases with positive microbiological findings (n = 47) had significantly (p < 0.001, Mann-Whitney U-test) more CD15Ne/focal point than cases with negative microbiological findings (n = 44). A count of 50 CD15Ne/focal point was identified as the optimum threshold when diagnosing periprosthetic joint infections (PJI) using the CD15 focus score. If the microbiological findings are used as a gold standard, the diagnostic sensitivity is 0.83, and the specificity is 0.864 with a positive predictive value (PPV) of 0.87, a negative predictive value (NPV) of 0.83, an accuracy of 0.846 and an area under the curve (AUC) of 0.878. The evaluation of findings for the preparations using the CD15 quantifier software (n = 31) deviated by an average of 12 cells from the histopathological evaluation findings (CD15 focus score). Above a cell count of 62, the CD15-quantifier needs on average 32 s less than the pathologist. The immunohistochemical CD15 focus score has a high diagnostic value and allowed the development of the CD15 quantifier software. This provides an automated procedure, which shortens the mentally tiring and time-consuming process of microscopic cell counting and thus makes a contribution towards the standardization of tools for diagnosing PJI.