Cryoglobulins Interfere with Platelet Counts by Optical and Impedance Methods but not with the CD61 Immunoplatelet Count

Automated complete blood counts (CBCs) use supplementary warnings (flags) to provide alerts for the existence of hematologic abnormalities (1). We observed on several occasions platelet-count interference by cryoglobulins, where the impedance and optical results were in error to a similar extent and did not trigger a delta alert. Platelets usually are counted by impedance analysis or by optical procedures (refractive and laser light-scattering properties). The CELL-DYN CD4000 system (Abbott Laboratories) simultaneously provides both and offers a further optional approach using immunoplatelet analysis (2)(3). In the latter method, a fluorescein isothiocyanate-labeled monoclonal antibody (CD61; Clone RUU-PL 7F12, IgG1 subclass) binds the glycoprotein (Gp)IIIa part of the GpIIb/IIIa receptor subunit (3). This glycoprotein is expressed by all human platelets, both resting and activated, but not by erythrocytes or leukocytes. The immunochemical platelet count is largely unaffected by sample interferences and allows determination of platelet counts to <1 × 109/L (4). The analytical performance is virtually identical to that reported for alternative flow cytometric immunomethods (5). We recently encountered four patients with cryoglobulinemia in whom erroneously high platelet counts were obtained by both optical and impedance platelet counts. Patient 1 had lymphoplasmacytoid lymphoma. Because the CD4000 analyzer flagged the presence of variant lymphocytes, a peripheral blood smear was examined. This confirmed the presence of lymphoma cells and also revealed thrombocytopenia, whereas the automated impedance and optical analyses suggested platelet counts of ∼355 × 109/L (Table 1 ). The result from a chamber (hemacytometry) count (38 × 109/L) also indicated thrombocytopenia, and a CD4000 CD61 immunoplatelet count (42 × 109/L) further substantiated the morphologic conclusions. Therefore, both the impedance and optical platelet counts were overestimated and, because the errors were of similar numeric magnitude, no CD4000 delta alert was given. Consequently, this patient’s thrombocytopenia would have been missed had the manual slide review not been performed for other reasons. We processed these same samples with other hematology analyzers (Sysmex SE8000, Sysmex K4500, Sysmex XE2100, and Bayer Technicon H3) and found platelet count overestimates similar to that of the CD4000. Clinical features and results from platelet counts (× 109/L) obtained by different methods. Clinical features and results from platelet counts (× 109/L) obtained by different methods. Patient 2 (Table 1 ) had Waldenström macroglobulinemia, but at the time of CBC analysis he was not thrombocytopenic. Both the CD4000 (impedance and optical) and Technicon H3 (optical) instruments significantly overestimated the platelet count. Two other patients seen during the past 18 months demonstrated similar patterns of interference. Both had cryoglobulinemia, one associated with centroblastic lymphoma (patient 3) and the other with Waldenström macroglobulinemia (patient 4; see Table 1 ). These patients also had several episodes of thrombocytopenia during their hospital stays for cytoreductive chemotherapy. Because of these observations, we wished to develop a review procedure for recognizing such situations in the absence of a significant difference between CD4000 impedance and optical platelet counts. Frequency distributions of platelet size may show increased steepness or asymmetry in the lower ends of platelet size histograms because of smaller-sized non-platelet particles such as cryoglobulin. The Technicon H3 analyzer displayed a typical peak indicative of cryoglobulinemia (6) in its optical platelet size plot (Fig. 1A ; patient 2), but this was not seen in the CD4000 impedance histogram (Fig. 1B ). In addition, the CD4000 optical scatter graph (Fig. 1C ) did not appear unduly abnormal, although there was some suggestion of a significant accumulation of events at the lower end of the 7° axis scatter. Thus, atypical CD4000 platelet plots are not always seen in cryoglobulinemia, although other interferences are usually detected with high efficiency. For comparison, whereas the Technicon H3 appeared to provide a more immediate visual indication of sample cryoglobulins, large platelet count overestimates were still obtained in most cases. Platelet distribution for patient 2. (A), histogram representing platelet size distribution of patient 2 (Table 1 ), who had cryoglobulinemia as determined with the Technicon H*3 analyzer. The reported platelet count was 819 × 109/L, although the observed histogram with a high event frequency in the very low size range suggests that it was erroneous. Analysis of the same sample with the CD4000 impedance (B; 578 × 109/L) and optical (C; 619 × 109/L) platelet counting methods does not reveal any obvious abnormal distributions, although there appears to be an unusually sharp cutoff at the lower threshold on the 7° axis of the optical plot. Subsequent analysis with the CD4000 immunoplatelet method revealed a platelet count that was significantly lower (226 × 109/L) than the counts obtained by the other methods. In the immunoplatelet graphic plot (D), a large number of gray events (non-platelet) are seen to overlay the cluster of platelets (black events). Examination of graphic plots is not used in all laboratories, and its efficiency may be limited when workload is high. Without it, operators depend on the presence of other hematologic abnormalities, instrument flags, or clinical information. Because patients with cryoglobulinemia, both at clinical presentation and during treatment, almost always show abnormal erythrocyte or leukocyte results, morphologic review is likely. When spurious platelet counts are suggested by graphic or morphologic review, a systematic approach to subsequent investigation can be initiated. Initially, warming samples (>30 min at 37 °C) before instrument analysis generally resolves cryoglobulin interference and enables platelet counting by standard optical or impedance methods. When interference persists after rewarming, other causes should be considered. When doubt remains regarding the true platelet count, the conventional alternative is manual hemacytometry, but its imprecision is unacceptably high when the platelet count is <20 × 109/L. The CD4000 CD61 immunoplatelet count, however, has high precision in thrombocytopenia (4) and, moreover, is not influenced by the presence of cellular or plasma interferences (Fig. 1D ). Despite the cost of the test (approximately US $16 in Germany), immunoplatelet counting should be considered in situations such as platelet transfusions and therapeutic interventions with bleeding risk. Few reports have described platelet count interference by cryoglobulins, although awareness of the potential analytical problems is high. Because the presence of thrombocytopenia was masked in two of the four cases described here, we believe that cryoglobulin interference is an important concern for the clinical laboratory. We thank Dr. Hans Seeger for helpful discussions and Prof. Victor W. Armstrong for valuable editorial support.