Normal and Chronic Lymphocytic Leukemia (CLL) Lymphocytes Can Be Distinguished in Flow-Cytometry-Based Automated Differential Cell Counts by Their Positional Parameters.

Introduction and objectives. Flow-cytometry based automated blood cell counters measure the volume (V), conductivity (C) and light scatter (S) of cells and make available these values for each subpopulation of white blood cells. Although differences in V, C and S among several lymphoproliferative disorders, reactive lymphocytes and normal lymphocytes have been described, the ability to use this this information to detect lymphoproliferative disorders has not been defined. Methods. We registered VCS data of all consecutive cases of lymphocytosis (lymphocytes > 3.5x10e9/l) detected in routine laboratory testing during 50 non-consecutive days. Lymphocytes were evaluated morphologically, immunophenotypically and by conventional cytogenetics if necessary and the patients were evaluated clinically to document diagnosis adequately. Discriminant analisys modelling was used to assess the ability to predict diagnosis of VCS data. Results. 202 cases of lymphocytosis were detected in 120 males and 82 females, mean age 53 years (SD 16). Patients were considered to have normal lymphocytes (N=102, 50%), to present a reactive lymphocytosis (N=43, 21%, 20 of them HIV infected patients) or achieved a final diagnosis of lymphoproliferative disorder (N=57, 28%, 43 of them CLL). The mean volume of CLL lymphocytes was significantly lower than that of normal and reactive lymphocytes (p=0.009) and so was light scatter (p=0.018). When mean VCS values and their standard deviations were included in a discriminant analysis model, a unique function adequately classified 166/189 cases (87%) as CLL versus reactive or normal. Sensitivity and specificity were 73% and 92% respectively and false positive and false negative cases were 26% and 8% respectively. Inclusion of a second discriminant function did not improve significantly the predictive power of the model. Conclusion, VCS data from flow-cytometry based automated blood cell counters can be used to differentiate CLL from normal or reactive lymphocytes by simple discriminant analysis classification.