Original Article · Originalarbeit Infus Ther Transfus Med 2001;28:277–279 Received: June 1, 2001 Accepted: August 6, 2001 Automated Analysis of Bone Marrow Fluid with the Sysmex XE-2100 Blood Cell Counter G. Zini G. Mistretta G. Giordano L. Laurenti G. d’Onofrio Key Words Nucleated red blood cells · Bone marrow aspiration · Immature granulocytes · Erythroblasts · Automated blood cell counters Schlüsselwörter Kernhaltige rote Blutzellen · Knochenmarkpunktat · Unreife Granulozyten· Erythroblasten · Automatische Blutzellzählgeräte Summary Background: Until recently, automated analysis of the cellular composition of bone marrow has not been possible, owing to problems in erythroblast identification, contamination by fat particles, and the heterogeneity of cell types. The XE-2100 automated blood cell counter can effectively count erythroblasts and granulocytes using laser light flow cytometry and DNA fluorochromes. In this study we have evaluated the performance of the XE-2100 for analyzing bone marrow cells. Material and Methods: The XE-2100 hematology analyzer performs white cell differential counts by measuring radiofrequency, direct current resistance, forward and side scatter of laser light, and fluorescence. Erythroblasts are identified by means of their specific fluorescent labelling. We have assessed i) the imprecision of the XE-2100 nucleated red blood cell (NRBC) counting using linear regression and the National Committee for Clinical Laboratory Standards (NCCLS) document H20A method of differences between duplicates, and ii) the comparability of NRBC count with microscopy counts using linear regression and analysis of differences. Results: Reproducibility of the XE-2100 measurements was excellent with coefficients of variation below 15%. We compared bone marrow granulocyte and erythroblast proportions by the XE-2100 with microscopy counts using linear regression and analysis of differences on 207 samples: Correlation between XE-2100 and the microscope was excellent for granulocyte percentage but slightly lower for erythroblasts. There was a tendency for the XE-2100 to slightly underestimate erythroblast percentage in most samples with a mean difference (XE-2100 – microscope) of –13.4%. Zusammenfassung Hintergrund: Noch bis vor kurzem war die automatisierte Analyse der Zellbestandteile von Knochenmarkpunktaten nicht möglich. Gründe waren die fehlende Erkennung und Separation von Erythroblasten, die Kontamination durch Fettpartikel und die Vielfalt der Zellbestandteile selbst. Mit dem automatisierten Blutzellanalysator XE-2100 ist es jetzt möglich, Erythroblasten und Granulozyten durch die Anwendung der Durchflusszytometrie mit Laserlicht und DNAFluoreszenzfarbstoffen sicher zu trennen. In dieser Arbeit haben wir die Verwendbarkeit des XE-2100 zur Knochenmarkanalyse getestet. Material und Methoden: Der Hämatologieanalysator XE-2100 benutzt für die Leukozytendifferenzierung fünf verschiedene Messprinzipien: Hochfrequenzund Gleichstrom-Widerstandsmessprinzip, Vorwärts- und Seitwärtsstreulicht eines Lasers und Fluoreszenzlicht. Die Erythroblasten werden durch ihr spezielles Fluoreszenzverhalten identifiziert. Wir haben zwei Versuchsreihen durchgeführt: 1) Die Impräzision der NRBC(nucleated red blood cell)Messung mittels linearer Regression und des NCCLS-H20AVerfahrens der Differenzbildung aus Doppelmessungen und 2) die Vergleichbarkeit der NRBC-Messung mit der mikroskopischen Bestimmung ebenfalls mittels linearer Regression und Differenzanalyse. Ergebnisse: Die Reproduzierbarkeit der XE-2100-Messungen war bei einem Variationskoeffizient < 15% ausgezeichnet. Wir haben dann die Granulozyten- und Erythroblastenergebnisse des XE-2100 mit den Ergebnissen der mikroskopischen Auswertung an 207 Proben mit Hilfe der linearen Regression und Differenzanalyse verglichen: Die Übereinstimmung zwischen dem XE-2100 und dem Mikroskop war bei den Granulozyten ausgezeichnet und bei den Erythroblasten etwas weniger gut. Die NRBC-Messergebnisse des XE-2100 lagen bei der Mehrzahl der Proben geringfügig unter denen der mikroskopischen Bestimmung. Die Abweichung betrug im Mittel –13,4%. © 2001 S. Karger GmbH, Freiburg Fax +49 761 4 52 07 14 E-mail Information@Karger.de www.karger.com Accessible online at: www.karger.com/journals/iut G Zini Servizio di Ematologia Università Cattolica del Sacro Cuore Largo F-Vito 1, I-00168 Roma, Italy E-mail firstname.lastname@example.org Downloaded by: Univ.of Adelaide 22.214.171.124 - 10/26/2017 5:43:53 AM Research Center for the Development and Clinical Evaluation of Automated Methods in Hematology (ReCAMH) – Hematology Service – Università Cattolica del Sacro Cuore, Roma Introduction 100 XE-2100 80 60 40 y = 0.8945x + 7.9065 2 r = 0.6566 20 0 0 a 20 40 60 80 100 Microscope 80 y = 0.6074x - 1.8603 2 r = 0.4441 60 XE-2100 Precise count and classification of bone marrow cells is useful for diagnosis and classification of diseases such as acute leukemia and myelodysplastic syndromes as well as for the monitoring of the effects of chemotherapy and bone marrow transplantation. Automated quantitative analysis of bone marrow cellular composition by hematology analyzers, however, has always been difficult [1–3]. A number of problems, in fact, arise from difficulty in erythroblast identification, contamination by fat particles, heterogeneity of cell types and maturation levels, and the presence of low-frequency and very large cells. In the hematology routine analyzers of the last generation technological advances have solved some of these difficulties, e.g. the identification of erythroblasts . The XE-2100 automated blood cell counter can effectively count, besides the physiological 5 leukocyte subtypes, erythroblasts and immature granulocytes using laser light flow cytometry and DNA fluorochromes. In this study we have evaluated the performance of the XE-2100 for the analysis of bone marrow cells. 40 20 0 Patients and Methods b Study Samples All samples consisted of 1–2 ml of bone marrow fluid collected in K2EDTA and analyzed on the XE-2100 within 4 h of venipuncture. Before aspiration samples were gently mixed by inversion. No filtration nor any other treatment of samples was carried out. Microscope analysis was performed on May-Grünwald-Giemsa-stained bone marrow smears. A total of 500 leukocytes per film was counted. Since the XE-2100 does not include NRBC count in the WBC count, all results were corrected to a total nucleated cell count (TNCC) calculated as TNCC = NRBC + WBC. 278 Infus Ther Transfus Med 2001;28:277–279 20 40 Microscope 60 80 Fig. 1. Correlation of bone marrow cell counts obtained with the XE2100 and the microscope. a granulocyte percentage. b erythroblast percentage. Results Precision and Comparability Reproducibility of the XE-2100 measurements was assessed through duplicate analysis of 165 samples. Coefficients of variation were below 15% for all counts except erythroblasts and lymphocytes. In 57 samples the instrument did not report a complete differential count. We compared the proportions of bone marrow granulocytes and erythroblasts by the XE-2100 with the microscope count using linear regression and analysis of differences on 207 samples out of a total of 264 samples, in which a complete computation was available. Correlation between XE-2100 and the microscope (fig. 1a) was excellent for granulocyte percentage (r2 = 0.657). Analysis of differences indicated a very good agreement in the great majority of samples, with a mean difference (XE-2100 – microscope) of –0.9%. Correlation between XE-2100 and the microscope was slightly lower for erythroblast percentage (r2 = 0.576) (fig. 1b), with a mean difference (XE-2100 – microscope) of –13.4%. There was a tendency of the XE-2100 to slightly underestimate erythroblast percentage in most of the samples. Zini/Mistretta/Giordano/Laurenti/d’Onofrio Downloaded by: Univ.of Adelaide 126.96.36.199 - 10/26/2017 5:43:53 AM The XE-2100 Counting Method The XE-2100 hematology analyzer (Sysmex, Kobe, Japan) performs white cell differential counts measuring i) radiofrequency and direct current resistance (indicators of cell content density and cell size), ii) forward and side scatter light emitted by a semiconductor laser beam (indicators of cell size and internal components), and iii) side fluorescence. Both mature and immature granulocytes are identified and counted in the WBC channel: Moreover, immature granulocytes are identified in the immature myeloid information (IMI) channel, owing to their partial resistance to surfactant used in that channel. Nucleated red blood cell (NRBC) identification and counting is carried out after red blood cell lysis. Intact white blood cells (WBCs) and NRBCs are stained with a fluorescent dye (in Stromatolyser-NR). In the resulting two-dimensional cytogram the intensity of side fluorescence is represented on the x-axis, while the intensity of the forward scattered light is represented on the y-axis. NRBCs are identified as a well-separated cell cluster to the left of the WBC population. The NRBC count is expressed as a proportion per 100 WBCs and as absolute number per unit volume of blood. We have assessed the imprecision of the XE-2100 NRBC counting using linear regression and the NCCLS-H20A method of differences between duplicates. Instrument comparability has been assessed against 500 cell microscope counts using linear regression and analysis of differences. The clinical usefulness of the XE-2100 NRBC method has been assessed from observations of instrument reports in patients with hematological disorders. 0 samples in leukemia, hypoplasia or erythroid hyperplasia (fig. 2). Our study showed that analysis of bone marrow samples with this method should be improved by an algorithm for fat exclusion. A problem was the high percentage of samples with incomplete differential count results (57/264): Dilution of the samples with phosphate buffered saline (PBS) to counts less than 20–30 × 109/l reduced the number of rejections and is presently carried out routinely before analysis. Discussion Fig. 2. XE-2100 scattergrams: In the upper left scattergram (DIFF) the y-axis represents the side fluorescence, the x-axis represents the side scatter. In the upper right scattergram (WBC/BASO) the y-axis represents the forward scatter, the x-axis represents the side scatter. In the lower left scattergram (IMI) the y-axis represents the alternating current signal (radio frequency) the x-axis represents the direct current signal. In the lower right scattergram (NRBC) the y-axis represents the forward scatter, the x-axis represents the side fluorescence. This example is obtained from a sample of bone marrow with normal cell composition. The sigmoid clusters in the WBC/BASO and NRBC channel are due to fat particles. Clinical Usefulness The TNCC was a good indicator of marrow cellularity. The observation of instrument cell distribution plots (scattergrams) provided the morphologist with an immediate general perception of the overall quality and characteristics of the Previous attempts to analyze bone marrow samples with automated blood cell counters have obtained very limited results [1–3], owing to difficulties such as the unpredictable behavior of erythroblasts and interference by cellular debris and lipid particles. The possibility to obtain rapid accurate cell counts from bone marrow samples with the XE-2100, similar to that reported with the Abbott CELL-DYN 4000 hematology analyzer , complements the morphological examination. Automated analysis permits a valuable objective preliminary evaluation of marrow samples, prior to conventional microscope examination. It provides the morphologist with an immediate general perception of the overall quality and characteristics of the sample, the relative proportions of cells and fat, total cellularity, myeloid/erythroid ratio, and presence and characteristics of abnormal cell clusters. We could demonstrate a very good correlation between automated counts and microscopic evaluation with respect to neutrophil percentage, an acceptable correlation with respect to erythroblast percentages, and a consistently lower automated erythroblast count compared to former published studies with flow cytometry and CELL-DYN 4000 [5, 6]. Some problems still have to be solved, such as interference by fat particles [4, 7], which often falsely increase the total cell count, the presence of and interference by platelet clumps, the need to recalculate the TNCC from WBC and NRBC counts, and finally the availability of results only on research screens for a significant number of samples. References Bone Marrow Analysis with Blood Cell Counters 4 D’Onofrio G, Zini G, Tommasi M, Laurenti L, Vergine C, Van Hove L: Quantitative bone marrow analysis using the Abbott CELL-DYN 4000 hematology analyzer. Lab Hematol 1997,3:146–154. 5 Terstappen LWMM, Levin J: Bone marrow cell differential counts obtained by multidimensional flow cytometry. Blood Cells 1992;18:311–323. 6 Sakamoto C, Yamane T, Ohta K, Hino M, Tsuda I, Tatsumi N: Automated enumeration of cellular composition in bone marrow aspirate with the CellDyn 4000 automated hematology analyzer. Acta Haematol 1999,101:130–134. 7 Lesesve JF, Goupil JJ, Latger V, Buisine J, Lecompte T: Artefactual elevation of the automated white cell count in the context of a bone marrow aspirate analysis. Clin Lab Haematol 2000,22:56– 57. Infus Ther Transfus Med 2001;28:277–279 279 Downloaded by: Univ.of Adelaide 188.8.131.52 - 10/26/2017 5:43:53 AM 1 Tatsumi J, Tatsumi Y, Tatsumi N: Counting and differential of bone marrow cells by an electronic method. Am J Clin Pathol 1986;86:50–55. 2 Bentley SA, Taylor MA, Killian DE, .Schoultz SB, McLannan L, Bishop CA, Shea TC, Brecher ME: Correction of bone marrow nucleated cell counts for the presence of fat particles. Am J Clin Pathol 1995;104:60–64. 3 Den Ottolander GJ, Baelde HA, Huibregtsen L, Paauwe JLE, Van der Burgh JF: The H1 automated differential counter in determination of bone marrow remission in acute leukemia. Am J Clin Pathol 1995:103:492–497.