Performance Evaluation of the Sysmex XE-2100 Hematology Analyzer

ISLHLaboratory Hematology 6:83-92© 2000 Carden Jennings Publishing Co., Ltd.Official PublicationPerformance Evaluation of the Sysmex XE-2100Hematology AnalyzerJERELYN WALTERS, 1 PATRICIA GARRITY 21 ACL Laboratories, West Allis; 2 Milwaukee Area Technical College, Milwaukee, WisconsinABSTRACTA new hematology analyzer, the XE-2100 (Sysmex, Kobe,Japan), was evaluated for performance in the core laboratory of amultiple-hospital system. Complete blood count performance wasevaluated in comparison to the SE-9000 (Sysmex). Differentialperformance was evaluated according to the National Committeefor Clinical Laboratory Standards (NCCLS) document H20-A 1.Precision, carryover, correlation, and linearity studies all showedexcellent results. Stability studies showed no significant change indifferential parameters for a period of 68 hours at both refrigeratedand room temperatures. It appears that the stability of the automateddifferential on the XE-2100 now exceeds the stability of redblood cell size parameters that were stable for 36 hours whenrefrigerated. Results comparing the XE-2100 automated differentialwith the 400-cell manual differential were very good. The newnucleated red blood cell (NRBC) measurement showed excellentprecision and linearity. Correlation with manually enumeratedNRBCs exceeded acceptable statistical standards. The evaluationled to the conclusion that the XE-2100 is an accurate and precisehigh-speed analyzer. Lab Hematol. 2000;6:83-92.KEY WORDS: Automation · Hematology ·Sysmex XE-2100 analyzer ·White blood cell differentialINTRODUCTIONThe Sysmex XE-2100 (Sysmex, Kobe, Japan) is a new discretehematology analyzer designed for high-volume testing in clinicallaboratories. The analyzer provides a 14-parameter hemogram aswell as a 5-part differential, integral reticulocyte analysis includingimmature reticulocyte fraction, and a nucleated red blood cellCorrespondence and reprint requests: Jerelyn Walters, 11026 W. Godsell,Milwaukee, WI 53130 (e-mail: garrityp@matc.edu)(Received March 9, 2000; accepted May 10, 2000)(NRBC) count. Differential parameters, reticulocyte analysis, andNRBC counts are determined using the principles of flow cytometrywith a semiconductor laser and appropriate fluorescent dyes.Instrument throughput is 150 samples per hour for the full hemogram,5-part differential, and NRBC count. The bidirectionalinterface capability with the laboratory information system allowsfor sample-to-sample discrete testing.The aim of this study was to evaluate the performance of theSysmex XE-2100 and assess its capability as an integrated hematologyanalyzer. The parameters evaluated for accuracy, precision, carryover,linearity, and stability were as follows: hemoglobin (HGB),hematocrit (HCT), red blood cells (RBCs), white blood cells(WBCs), 5-part automated differential, and impedance plateletcount (PLT-I). The routine blood cell count parameters were comparedwith the SE-9000 (Sysmex). Differential performance wasevaluated according to the National Committee for Clinical LaboratoryStandards (NCCLS) document H20-A 1 [1]. The NRBCparameter was evaluated to assess accuracy, precision, and linearity.This study was performed in the core laboratory of a large healthcare system. Examination of the Sysmex XE 2100 reticulocyteparameters and optical platelet count (PLT-O) was performed in aseparate study, and the results are not reported here.MATERIALS AND METHODSInstrumentsThe Sysmex XE-2100 is an integrated multiparameter hematologyanalyzer with the capacity to store analysis data plus histogramsfor 10,000 samples using a Windows NT–based platform.An example of the Sysmex XE-2100 histogram/scattergram anddata analysis printout can be seen in Figure 1. There are 11 qualitycontrol files available for use. This analyzer also incorporatesa system of flags to alert the operator to the presence of the following:atypical lymphocytes; abnormal lymphocyte/L-blasts;left shift; immature granulocytes; blasts; NRBCs; RBC fragments;RBC lyse resistance; RBC agglutination; dimorphic population;turbidity/hemoglobin interference; iron deficiency;HGB defects; platelet clumps; WBC, NRBC, reticulocyte, andplatelet (PLT) abnormal scattergrams; and RBC and PLT abnormal83

84 J.Walters and P. GarrityNegativeFIGURE 1. Sysmex XE-2100 histogram/scattergram and data analysis printout. SFL indicates side fluorescent; SSC, side scatter;FSC, forward scatter; RF, radio frequency; DC, direct current; WBC, white blood cell; BASO, basophil; IMI, immature myeloidinformation; NRBC, nucleated red blood cell; RET, reticulocyte; PLT-O, optical platelet count; RBC, red blood cell; PLT, platelet;HGB, hemoglobin; HCT, hematocrit; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, meancorpuscular hemoglobin concentration; RDW-SD, red blood cell distribution width by standard deviation; RDW-CV, red bloodcell distribution width by coefficient of variation; MPV, mean platelet volume; NEUT, neutrophil; LYMPH, lymphocyte;MONO, monocyte; EO, eosinophil; IRF, immature reticulocyte fraction; IP, interpretive.distribution. User-defined flags can be set to identify abnormalnumerical results. An example of the Sysmex XE-2100 histogram/scattergramand data analysis printout for an abnormalsample can be seen in Figure 2. HGB determination is achievedusing the sodium lauryl sulfate (SLS)-hemoglobin method; SLSis a cyanide-free reagent.RBCs and PLT-I are enumerated in the RBC/PLT channelusing the sheath flow direct current (DC) detection method [2].HCT is simultaneously determined using the RBC pulse heightdetection method [3].The total and differential WBC counts are determined by flowcytometry, and each cell is analyzed by 3 different dispersion

Sysmex XE-2100 Performance Evaluation 85WBC IP Message(s) RBC/RET P Message(s) PLT IP Message(s)Neutrophilia RET Abn Scattergram PLT Abnormal DistributionMonocytosisReticulocytosisLeukocytosisAnisocytosisNRBC PresentMacrocytosisImmature Gran?Left Shift?FIGURE 2. Sysmex XE-2100 histogram/scattergram and data analysis printout: abnormal sample. Diff. Morph. indicates differentialmorphology. For other definitions, see legend to Figure 1.angles: forward-scattered light, lateral-scattered light, and lateralfluorescent light. The immature myeloid information (IMI) is aseparate channel that provides additional WBC information onthe presence of immature granulocytes, blasts, and hematopoieticprogenitor cells. This channel uses a hemolytic reagent that selectivelyprotects immature cells that are analyzed by the radio frequency(RF)/DC detection method [4]. Nucleated RBCs aredetected in a separate flow channel using both a specific lysingreagent and a fluorescent dye.Reticulocytes are enumerated by flow cytometry using a semiconductorlaser and a nucleic acid fluorescent dye. The dye reactionis based on the same principle as that of the Sysmex R series reticulocyteanalyzers [5]. By analyzing both forward-scattered light andside fluorescence information, reticulocyte counts and reticulocytematurity information were obtained. A PLT-O count was also performedin this channel.The ability to perform platelet counts by 2 different technologiesprovides more accurate platelet information. The sheath flow

86 J.Walters and P. GarrityTABLE 1. Within-Run Precision (n = 10)*Sample 1 Sample 2 Sample 3 Sample 4 Sample 5Mean SD CV, %* Mean SD CV, % Mean SD CV, % Mean SD CV, % Mean SD CV, %Red blood cells, 10 6 /µL 4.19 0.03 0.7 4.66 0.03 0.6 3.81 0.02 0.5 4.13 0.03 0.7 4.58 0.01 0.2Hemoglobin, g/dL 12.1 0.06 0.5 14 0.05 0.4 12.5 0.03 0.2 12.6 0.00 0.0 13.3

Sysmex XE-2100 Performance Evaluation 87TABLE 3. CarryoverParameter %White blood cells 0.0Red blood cells 0.4Hemoglobin 0.0Hematocrit 0.5Nucleated red blood cells 0.0Impedance platelet count 0.0PrecisionWithin-run precision was evaluated by performing 10 consecutivemeasurements on 5 fresh blood samples. Between-day precisionwas evaluated by processing commercial control over aperiod of 10 days.CarryoverCarryover was assessed using the International Committee forStandardization in Hematology (ICSH) procedure [9]. Carryoverwas performed by analyzing a specimen with a high concentrationconsecutively in triplicate (H1, H2, and H3) followed immediatelyby a low concentration consecutively in triplicate (L1, L2, and L3).The percentage of carryover for each parameter was calculatedusing the following formula:Carryover % = (L1 – L3) 100(H3 – L3)TABLE 4. LinearityParameter Range r 2White blood cellsHigh level 0.17 to 173.00 10 3 /µL 1.000Low level 0 to 1.12 10 3 /µL 1.000Impedance platelet countHigh level 1 to 2054 10 6 /µL 0.998Low level 0 to 382 10 6 /µL 1.000HemoglobinHigh level 0.0 to 26.6 g/dL 1.000Low level 0.0 to 10.3 g/dL 1.000HematocritHigh level 0.2% to 81.4% 1.000Low level 0.1% to 30.1% 1.000Red blood cellsHigh level 0.92 to 9.23 10 6 /µL 0.997Low level 0.01 to 3.48 10 6 /µL 1.000making a 400-cell manual differential impractical. The hemogramand autodifferential results were evaluated using regression analysis.Two blood films were made and stained with a polychromestain. Manual differentials were performed on each sampleaccording to the NCCLS document H20-A 1 [1]. Manual differ-ALinearityLow linearity was evaluated by serial dilution technique offresh patient samples. The analyzer’s primary diluting fluidwas used as the diluent. High linearity was evaluated usingboth fresh patient samples and concentrated samples. Serialdilutions were performed as stated above. Extended PLT-I linearitywas evaluated with a platelet concentrate sample dilutedwith WBCs and platelet-depleted RBCs.StabilityShort-term stability. Blood was collected from 6 normaldonors and stored at room temperature. The samples wereanalyzed in duplicate on the Sysmex XE-2100 at 0, 15, 30, 45,and 60 minutes.Long-term stability. Blood was collected from 10 normal donorsand divided into 19 aliquots. Nine aliquots were stored at 4°C, and9 aliquots were stored at room temperature. The remaining aliquotwas tested immediately. At 4, 8, 12, 16, 24, 36, 48, 56, and68 hours after collection, 1 aliquot each from room temperatureand 4°C was analyzed. All aliquots stored at 4°C were allowed towarm to room temperature before analysis.ComparabilityThere were 200 random blood samples selected from the dailyworkload and analyzed on the Sysmex XE-2100 and Sysmex SE-9000. An additional 12 samples were excluded from the studybecause of incomplete or missing data. Two additional sampleswere excluded because the WBC counts were

88 J.Walters and P. GarrityXE-2100 NEUTROPHIL #XE-2100 LYMPHOCYTE #FIGURE 4. Autodifferential linearity. A. Neutrophils. B. Lymphocytes. C. Monocytes.ential results were compared with the autodifferential resultsusing linear regression analysis.NRBC MeasurementLinearity and precision were evaluated using a whole bloodpatient sample with >40,000 NRBCs/µL. Accuracy was assessedusing the following comparisons:• Sysmex XE-2100 absolute NRBC number with the absoluteNRBC number calculated from a 400-cell manual differential.• Sysmex XE-2100 WBC count with the total nucleated cell countmanually corrected for presence of NRBCs.• Sysmex XE-2100 WBC count with a total nucleated cell countobtained on the Sysmex F-300. This total nucleated cell countincludes NRBCs.RESULTSPrecisionResults from within-run replicate analyses of blood samplesare shown in Table 1. The precision surpassed manufacturer specificationsfor all parameters. All measured RBC parametersshowed excellent precision, with coefficients of variation (CVs)

Sysmex XE-2100 Performance Evaluation 89FIGURE 5. Long-term stability: room temperature. WBC indicates white blood cell.Long-term stability results for refrigerated storage are shown inFigure 6.ComparabilityThe comparison of results between the Sysmex XE-2100 andthe Sysmex SE-9000 showed excellent correlation for all parametersexcept absolute number of basophils (Table 5). Comparison of theSysmex XE-2100 automated differential with the 400-cell manualdifferential showed good correlation (Table 6).NRBCsThe sample with elevated NRBCs analyzed 10 times gave CVsof 1.05% for absolute number of NRBCs and 1.37% for adjustedWBCs. The NRBC parameter showed acceptable linearity up to45,960/µL (Figure 7). NRBC number per µL determined by theSysmex XE-2100 showed excellent correlation with the NRBCnumber per µL calculated from the 400-cell manual differentialand the total nucleated cell count (Figure 8).The corrected WBC count reported by the Sysmex XE-2100compared well with the manually corrected WBC count based onthe 400-cell differential (Figure 9). The corrected WBC countreported by the Sysmex XE-2100 showed poor correlation with theuncorrected total nucleated cell count (Figure 10).DISCUSSIONThe Sysmex XE-2100 performed exceptionally well for theparameters evaluated. Results for precision and carryover studiesexceeded manufacturer specifications. There was excellent correlationbetween instruments for the measured parameters (HGB,HCT, RBCs, PLT-I, WBCs, and automated differential). The poorbasophil correlation was not unexpected because of very lowabsolute numbers of basophils.Results for linearity were very good. WBCs and PLT-I showedexcellent linearity over a wide range (WBCs, 0 to 400,000/µL;PLT-I, 0 to 9,000,000/µL). Extended low linearity of the automateddifferential parameters (neutrophils, lymphocytes, and monocytes)provides a significant advantage when analyzing leukopenic samples.Short-term stability studies showed there is no immediate timelimitation on accuracy of results. Long-term stability studies showed

Sysmex XE-2100 Performance Evaluation 91TABLE 6. Comparison of Manual Differential With XE-2100MinimumMaximumManual XE-2100 Manual XE-2100 Correlation Coefficient Slope InterceptNeutrophils, % 5.5 4.9 98.0 97.0 0.970 0.980 –0.87Lymphocytes, % 2.0 1.8 93.0 93.1 0.970 0.960 2.28Monocytes, % 0.0 0.0 23.5 30.7 0.850 0.999 1.01Eosinophils, % 0.0 0.0 10.5 11.8 0.900 0.980 0.04Basophils, % 0.0 0.0 7.0 4.1 0.650 0.500 0.15impressive precision of the automated differential parameters foralmost 3 days. The stability of the automated differential exceedsthe stability of RBC measurements. Thus, the MCV/HCT are nowthe limiting factors in sample stability for hematologic analysis.These parameters were stable at refrigerated temperatures for approximately36 hours. Room temperature storage limits stability to a periodof 12 to 24 hours.Comparability of the 400-cell manual differential with the SysmexXE-2100 automated differential was good, including results oflow-level cell types (monocytes, eosinophils, and basophils).Because of the increased accuracy of the automated differential, thescreening potential of Sysmex XE-2100 has been enhanced.The new technology to measure NRBCs appears to be anextremely accurate and precise parameter. Good correlation hasbeen shown between the corrected WBC on the Sysmex XE-2100and the manually corrected total nucleated cell count. The poorcorrelation seen between the Sysmex XE-2100–corrected WBCcount with the uncorrected total nucleated cell count is further evidencethat the Sysmex XE-2100 provides an accurate WBC countin the presence of NRBCs. The functionality of having both anFIGURE 7. Nucleated red blood cell (NRBC) linearity.FIGURE 9. White blood cell (WBC) correlation: Sysmex XE-2100 corrected WBC count versus manually corrected WBCcount.FIGURE 8. Nucleated red blood cell (NRBC) correlation: SysmexXE-2100 NRBC count versus calculated NRBC countfrom differential and total nucleated cell count.FIGURE 10. White blood cell (WBC) correlation: Sysmex XE-2100 corrected WBC count versus uncorrected total nucleatedcell count.

92 J.Walters and P. Garrityaccurate WBC count and differential and an absolute number ofNRBCs is a significant progression in automated hematology. Thistechnological advance substantially reduces the need for manualintervention on these abnormal samples.In conclusion, the Sysmex XE-2100 has proven to be an accurateand precise high-speed analyzer. It is a suitable analyzer forboth high-volume laboratories and laboratories that test manyabnormal samples. The stability and precision of the automateddifferential enhances its application as a screening analyzer andreduces the need for manual differentials.REFERENCES1. National Committee for Clinical Laboratory Standards (NCCLS). ReferenceLeukocyte Differential Count (Proportional) and Evaluation of InstrumentMethods. Wayne, PA: NCCLS Document H20-A 1; 1992.2. Thom R, Hampe A, Sauerbrey G. Die elektronische volumenbestimmungvon blutkörperchen und ihre fehlerquellen [Electronic determination ofthe volume of blood corpuscles and its sources of errors]. Z Gesamte ExpMed. 1969;151:331-349.3. Jones AR. Counting and sizing of blood cells using aperture–impedancesystems. In: van Assendelft OW, England JM, eds. Advances inHematological Methods: The Blood Count. Boca Raton, FL: CRC Press;1982:59-63.4. Fujimoto K. Principles of measurement in hematology analyzers manufacturedby Sysmex Corporation. Sysmex J Int. 1999;9:31-40.5. Hohenwallner W, Wiesinger K, Wimmer E. The reticulocytes: automaticcounting, indication and interpretation. Sysmex J Int. 1992;2:120-135.6. Cornbleet J. Spurious results from automated hematology cell counters.Lab Med. 1983;14:509-514.7. TOA Medical Electronics. Sysmex SE-9000 System Operation Manual.Kobe, Japan: TOA Medical Electronics; 1994.8. TOA Medical Electronics. Sysmex XE-2100 System Operation Manual.Kobe, Japan: TOA Medical Electronics; 1999.9. International Committee for Standardization in Hematology (ICSH). Protocolfor evaluation of automated hematology analyzer. ICSH. 1984;6:69.