Characterizing low platelet count sampleswith discrepant results using a reference method and a fully automated comprehensive hematology analyzer, Alinity hq

Mika Araki*¹, Yoko Yatabe*¹, Kaori Yahagi*¹, Tomoko Arai*¹, Yutaka Nagai*³, Hiromitsu Yokota*¹,  Yoichi Nakayama*², Takayuki Mitsuhashi*³, Masatoshi Wakui*³,  Hiromichi Matsushita*³

Correspondence
Lab Med Int 2025; 4(3): 100-102

†Correspondence: Hiromichi Matsushita, Department of Laboratory Medicine, Keio University School of Medicine,
Tokyo 160-8582, Japan
Tel: +81-3-3353-1211 (Ext. 68623); Fax: +81-3-3359-6963
E-mail: hirommat”@”keio.jp
Received November 8, 2024; accepted May 19, 2025
*1 Department of Clinical Laboratory, Keio University Hospital, Tokyo 160-8582, Japan
*2 Abbott Japan LLC, Tokyo 108-6305, Japan
*3 Department of Laboratory Medicine, Keio University School of Medicine, Tokyo 160-8582, Japan

  〔Lab Med Int 2025; 4(3): 100-102〕 

Key Words

Hematology analyzer, Platelet count

Text

Dear Editor,
Automated cell counters currently used in many laboratories can rapidly deliver accurate blood count parameters based on multiple principles of cell counting. Among these blood cell parameters, platelet count is an important decision-making factor for platelet transfusion therapy; thus, an accurate platelet count (especially at a low count) is imperative for appropriate and safe transfusion therapy.
The fully-automated comprehensive hematology analyzer Alinity hq (Abbott, Abbott Park, IL; hq) is based on the flow cytometry-based technique. It measures complete blood count and white blood cell differential by analyzing optical signals obtained by the Multi Angle Polarized Scatter Separation (MAPSS^TM) technology¹⁾²⁾. Additionally, it can enumerate platelets without switching measurement channels and does not require a dedicated staining reagent. We previously reported that measurements of low platelet counts (<100 × 10⁹/L) by hq were consistent with those obtained by a flow cytometry-based reference method and the PLT-F and PLT-O methods applied in the XN-series (Sysmex Corporation)³⁾. However, results of a small number of peripheral blood samples with a platelet count <100 × 10⁹/L deviated from those of the reference method. Therefore, we further analyzed these samples. This study was approved by the Keio University School of Medicine Ethics Committee (approval number: 20170393).
A total of 68 EDTA-2K-treated peripheral blood samples was obtained from outpatients. Of these, 63 samples were analyzed in a previous study³⁾, and the remaining 5, once excluded due to suboptimal quality, were also included in the present analysis. The reference method was developed by the Blood Cell Counting Standardization Subcommittee of the Japanese Society for Laboratory Hematology (JSLH), wherein the absolute platelet count was calculated using a flow cytometry-based method compliant with ICSH/ISLH2001 (indirect method)⁴⁾. The samples were initially analyzed using the JSLH method, followed by the hq method and blood smear preparation. The correlation coefficient between the JSLH and hq methods indicated a strong correlation, as observed previously³⁾. However, samples from three cases showed a deviation of more than 2 SD in the Bland-Altman analysis between the JSLH and hq methods (Figure 1). These samples were further investigated using blood smears and (IAS × ALL [WBC]) scattergrams were generated using hq (Table 1).
In Case 1, the platelet count was 42.3 × 10⁹/L and 50.6 × 10⁹/L by the JSLH method and hq, respectively, with the morphological flag “rst RBC”, indicating that RBCs were resistant to hemolytic process in hq, which resulted in a higher value in hq measurement. An abnormal cluster was observed in the area of lower-than-normal fluorescence intensity on the scattergram, and the blood smear showed poikilocytes and Howell-Jolly bodies. In Case 2, the platelet count was 74.6× 10⁹/L and 62.1 × 10⁹/L by the JSLH method and hq, respectively, with the morphological flag “Plt Clump?” indicating platelet clumps, which resulted in a lower value in hq measurement. An abnormal cluster was also found on the hq scattergram, although it was smaller than that observed in case 1, and the blood smear showed no platelet aggregation; instead, a small number of giant platelets were recognized. In Case 3, the platelet count was 93.1 × 10⁹/L and 70.4 × 10⁹/L by the JSLH method and hq, respectively, with the morphological flag “Plt Clump?”, again resulting in a lower value in hq measurement. The hq scattergram showed a tiny abnormal cluster; however, no apparent platelet aggregation was observed in the blood smear (Figure 2).
The hq results for these three cases showed minor deviations compared with those obtained using the JSLH method (9-23 × 10⁹/L), These deviations were associated with morphological flags such as “Plt Clump?” and “rst RBC”, as well as abnormal clusters on the hq scattergram. These clusters may indicate the presence of platelet aggregation or hemolysis-resistant red blood cells, which may be underrepresented  by microscopic observation. Therefore, special attention should be paid to these abnormal clusters and associated morphological flags, as they may affect the accuracy of platelet counts determined by hq. 
In conclusion, although hq delivers highly reliable platelet counts even for samples with very low platelet count, samples with morphological flags in real-world applications should be carefully investigated by observing the scattergrams and blood smears and comprehensively determining whether they are consistent with measured results.

Figure 1   Bland-Altman analysis of 68 peripheral blood samples comparing hq measurements with the reference values obtained by the JSLH method. Three samples indicated by circles with arrowheads exhibited deviations exceeding ±2 standard deviations. 

Table 1  Platelet counts in the three samples whose results were inconsistent when analyzed by the JSLH and hq methods.

* Reference method developed by the Blood Cell Counting Standardization Subcommittee of the Japanese Society for Laboratory Hematology (JSLH).

Figure 2   The hq scattergrams and blood smears for three samples with inconsistent results when analyzed using the JSLH and hq methods. Red arrows in the scattergrams indicate abnormal clusters (grey dots). Flags presented in the hq analysis are shown at the top right corner of each scattergram. ALL, axial light loss; Ba, basophils; Eo, eosinophils; IAS, interme angle scatter; Ly, lymphocytes; Mo, monocytes; Ne, neutrophils; PltClmP, platelet clumps; rstRBC, resistant red blood cells. 

Author contributions
Conceptualization: Arai T and Yokota H
Investigation: Araki M, Yatabe Y, and Yahagi K
Visualization: Arai T, Nagai Y and Yokota H
Supervision: Matsushita H
Writing – original draft: Mitsuhashi T and Wakui M
Writing – review & editing: Nakayama Y, Mitsuhashi T and Matsushita H

Acknowledgments: 
None

Conflicts of interest: 
Yoichi Nakayama is an employee of Abbott Japan LLC, whereas Takayuki Mitsuhashi is a medical advisor of Abbott Japan LLC. There are no other conflicts of interest to declare.

Research funding:
none.

References

1. Van der Beken Y, Van Dalem A, Van Moer G, et al. Performance evaluation of the prototype Abbott Alinity hq hematology analyzer. Int J Lab Hematol. 2019; 41(4): 448-55.PubMed
2. Gambell P, Rowley G, Pham TAT, et al. Accurate white blood cell differential by Alinity hq: A comparison with flow cytometry and manual differential. Int J Lab Hematol. 2022; 44(2): 288-95.PubMed
3. Lifson MA, Wakui M, Arai T, et al. Alinity hq platelet results are equivalent with the international reference method in thrombocytopenic samples. Int J Lab Hematol. 2021; 43(6): 1357-62.PubMed
4. Japanese Society of Laboratory Hematology (JSLH). Standard Operating Procedure (SOP) for the Indirect Method for Platelet Absolute Count Determination by Flow Cytometry [in Japanese]. Available from: http://jslh.kenkyuukai.jp/images/sys/information/20190926111409-32FF8C754AD98B70E4D0049CDE7CD4DCDF9F15EA8CD6DDD4E73F12A34403093B.pdf (Accessed: 2025-10-02)