Evaluation of the XE-5000 for the automated analysis of blood cells in cerebrospinal fluid

Abstract

Objectives Counting of cells in cerebrospinal fluid is an important clinical laboratory test and elevated white blood cell counts in cerebrospinal fluid are frequently seen in CNS disorders. Quantification of red blood cell concentrations in CSF may help to interpret certain diagnostic constellations and may result from subarachnoid haemorrhage, surgical procedures or contamination due to traumatic puncture. Table top analyser XE-5000 (Sysmex, Norderstedt, Germany) offers, beside its use as a haematology analyser, a protocol for the quantification of red and white blood cells in body fluids such as CSF including the differentiation between polymorphonuclear and mononuclear cells. A detection limit of 1 cell/mm 3 would render this device suitable for automated CSF analysis. Design and methods White blood cell counting was compared between Fuchs–Rosenthal counting chamber and XE-5000 in 273 routinely collected lumbar and ventricular CSF samples. Red blood cell counting was compared between UF-100 and XE-5000. Differentiation was performed on a slide stained after Pappenheim and compared to the differential count of the XE-5000. Results Linearity was established between 1 and 10,000 cells/mm 3 for white blood cells and between 1000 and 1 ⁎ 10 3 particles/mm 3 for red blood cells. Functional sensitivity was established at 20 cells/mm 3 for white blood cell counting and at 1000 particles/mm 3 (lowest reported concentration) for red blood cell counting. When comparing between microscopic and automatic white blood cell counts no statistically significant slope and offset were detected in lumbar CSF samples while a significant slope and offset were detected when comparing ventricular CSF samples. Most patients were classified correctly according to their WBC count (non-pathologic, mildly, moderately, and highly elevated) by both methods although more patients had pathologic white blood cell counts on XE-5000. A significant slope and offset were detected when comparing red blood cell counts between UF-100 and XE-5000. Conclusions In summary despite its high imprecision at low white blood cell counts (< 20 particles/mm 3) most patients were classified correctly and therefore XE-5000 is suitable for automated quantification of white blood cells in cerebrospinal fluid in a defined diagnostic setting. This could significantly improve automation in the relatively time- and manual work-intensive field of cerebrospinal fluid diagnostics. However, careful review of plausibility of the results continues to be compulsory.