Discrepancy between impedance and immunofluorescence platelet counting has implications for clinical decision making in patients with idiopathic thrombocytopenia purpura


Bowles et al (2006) reminded us of the limitations of impedance platelet counting, a well known and frequently reported phenomenon (Ault, 1996; Harrison et al, 2000). Ault (1996) demonstrated a difference between the impedance and the two-angle light scatter optical platelet count on the Abbott Diagnostics CELL-DYN 4000 (Abbott Diagnostics). The optical count correlated better with the immunofluorescence count due to the presence of non-platelet particles in the samples, which were included in the impedance count on the basis of size. Harrison et al (2000) found that impedance counts on three different analysers underestimated the platelet count in idiopathic thrombocytopenia purpura (ITP) but the optical count on the ADVIA 120 gave significantly better results when compared with an immunofluorescence reference method. The immunofluorescence method used in these studies was the red blood cell/platelet (RBC/PLT) ratio flow cytometric method. This has since become the accepted international reference method (Harrison et al, 2001). The method uses two antiplatelet monoclonal antibodies to label the platelets and the count is derived from the ratio of fluorescent platelet events to the number of red cells detected, negating the need for the addition of calibration beads to the sample to obtain the absolute platelet count. The immunofluorescence TruCount method employed by Bowles et al (2006) used calibration beads but the nature of the antiplatelet monoclonal antibodies is not described. The RBC/PLT ratio method demonstrates superior precision to the bead method (Harrison et al, 2000) and is independent of pipetting and dilution artefacts, which influence the bead ratio method.

It is important to note that impedance platelet counting methods on different analysers may give different results on the same sample (Segal et al, 2005). This is due to differences in method analysis (fixed or moving size thresholds), linearity over the entire range and the number of cells counted.

In this study, mean platelet volume (MPV) and platelet distribution width did not predict impedance/immunofluorescence discrepancies (data not shown), although a high MPV represents the presence of large platelets, which may be excluded from the impedance count. An earlier study by the same group found that MPV, using the same haematology analyser, could guide the clinician as to the likely presence or absence of bone marrow disease (Bowles et al, 2005). MPV has also been reported as a sensitive and specific marker to discriminate thrombocytopenia caused by ITP and that caused by aplastic anaemia (Kaito et al, 2005). The MPV is increased in ITP and therefore should predict impedance/immunofluorescence discrepancies.

For many laboratories, the introduction of optical platelet counting has improved the accuracy in samples with large platelets, e.g. ITP. However, optical counts are not always more accurate in patients with thrombocytopenia due to chemotherapy (Segal et al, 2005), where white cell fragments may incorrectly elevate the platelet count.

The recent introduction of the Immature Platelet Fraction from the optical fluorescent platelet count on the Sysmex XE-2100 has improved the diagnosis and management of patients with thrombocytopenia. Unlike MPV, it is a direct measurement of the rate of thrombopoiesis and can indicate the cause of thrombocytopenia (Briggs et al, 2004).

This article correctly demonstrates that impedance counts will give erroneous platelet results on some abnormal samples and users need to be aware of this. However, most routine haematology laboratories will not have the time, skill or facilities to perform an immunofluorescence count on thrombocytopenic samples. More specifically, the immunofluorescence count is designed to be a reference method for evaluation purposes, assigning accurate platelet counts to controls and for the calibration of haematology analysers.