Automated cerebrospinal fluid cell count--new reference ranges and evaluation of its clinical use in central nervous system infections.

OBJECTIVES

The purposes of this study were to establish new reference ranges for leukocytes in the CSF and to examine if the separation of mononuclear cells into lymphocytes and monocytes could be used to differentiate between various CNS infections that present with a similar picture in manual CSF cell counts.

DESIGN AND

METHODS

The automated cell counter Siemens ADVIA 2120 i was used. For the reference range section, we analyzed CSF from 80 neurologically healthy volunteers. For the differential diagnosis section we analyzed cell counts and hospital records from 175 patients with CSF mononuclear pleocytosis.

RESULTS

Correlation was good between automated and manual leukocyte counts for samples with erythrocyte counts <250 cells/μL. For the neurologically healthy volunteers studied in the reference range section, the 95th percentile was 3.0 cells/μL for lymphocytes, 1.0 cell/μL for monocytes and 1.0 cell/μL for granulocytes. In the differential diagnosis section, comparisons were done between the groups Lyme neuroborreliosis and viral CNS infection. There were no significant differences between these two groups regarding cell counts; neither for lymphocytes, median 58 cells/μL vs. 72 cells/μL (P = n.s.); nor for monocytes, median 13 cells/μL vs. 16 cells/μL (P = n.s.); nor for granulocytes, median 1 cell/μL vs. 2 cells/μL (P = n.s.)

CONCLUSIONS

We suggest new CSF cell count reference ranges of <4 cells/μL for lymphocytes, <3 cells/μL for monocytes and <3 cells/μL for granulocytes. The separation of mononuclear cells into lymphocytes and monocytes did not facilitate the discrimination between Lyme neuroborreliosis and viral CNS infection.