Counting cells in cerebrospinal fluid (CSF) using automated analyzers is generally problematic due to low precision at low cell numbers. To overcome this limitation, Sysmex (Kobe, Japan) developed the high-sensitive analysis (hsA) research mode specifically for counting cells in fluids that contain low cell counts. We evaluated this mode by counting RBCs, WBCs, and differentiated WBCs in CSF samples.
We analyzed 248 CSF samples using the hsA mode and compared these results with those obtained using the manual counting method. We also evaluated the linearity, detection limits, carryover, and precision of the hsA mode.
Using the hsA mode, the lower limit of quantification for RBCs and WBCs was 10 and 2 cells/μL, respectively. Comparing the two methods revealed good agreement with respect to WBCs (y = 1.08x + 0.52), RBCs (y = 1.07x + 0.00), lymphocytes (y = 1.00x + 0.00), neutrophils (y = 1.05x + 0.00), and monocytes (y = 0.88x + 0.07). Regression analysis for samples containing low WBCs (<10 cells/μL) and low RBCs (<50 cells/μL) also had good agreement, although a slight positive bias was found for RBCs. Linearity was good (r(2) ≥ 0.99) for all parameters evaluated. Carryover was negligible and never exceeded 0.04%.
The XN hsA research mode provides reliable cell counts in CSF samples, even in samples containing low numbers of WBCs and RBCs.