Influence of white blood cells on the filterability of human erythrocytes: modelling of the filtration process and experimental results.

In medical research and clinical practice, the increasing application of filtration measurements, to characterize the erythrocyte deformability has aroused an intensive discussion about possible erroneous interpretations of measured values (filterability indices). In the present paper the influence of white blood cells (WBC) on these indices is described and discussed in relation to a mathematical model of the filtration process. We found for two gravity filtration systems, using cellulose filters and concentrated erythrocytes suspensions sensitivities of their filterability indices of (1 +/- 0.1)% per 100 WBC/microliter (initial filtration rate measurement) and (3 +/- 0.3)% per 100 WBC/microliter (hemofiltrometer). In comparison to Nuclepore filtration systems with sensitivities of 5-12% per 100 WBC/microliters, cellulose filters were much less sensitive to WBC. The necessity, advantages and disadvantages of additional procedures for WBC removal are discussed. A standardization of residual WBC-count in limits of 500 +/- 300 WBC/microliters at a hematocrit of 60% is proposed. Fitting of the filtration curves of the hemofiltrometer to a model function may either improve the accuracy of estimation of filterability indices or divide them into two model parameters with different WBC-sensitivities: the initial filtration rate parameter having a low WBC-sensitivity ((0.9 +/- 0.2)% per 100 WBC/microliter) and the occlusion related parameter having a 12 times higher sensitivity ((11 +/- 1)% per 100 WBC/microliter). Interpretation, advantages and disadvantages of the model parameters are discussed also in comparison to the filterability index.