Automated hematology analyzers (HA) are widely used to obtain complete blood cell counts (CBC). CBC are critical for the accurate assessment of patients’ clinical status; thus, the results must be accurate and precise. However, in some cases, spurious CBC can occur, which may require additional action on the part of the clinician to obtain correct results [1, 2]. We report a case of an incorrect white blood cell (WBC) count obtained on a Sysmex XN hematology analyzer (Sysmex Corporation, Kobe, Japan).

A 56-year-old man visited our hospital for follow-up of hepatocellular carcinoma with lung metastases. He had started sorafenib (a molecular targeted multi-kinase inhibitor) 3 months before and nivolumab (a humanized monoclonal antibody to the immune checkpoint receptor programmed death 1 [PD-1]) 2 days before his peripheral blood was drawn for laboratory tests, including CBC. The initial WBC count measured on a Sysmex XN hematology analyzer (Sysmex Corporation) was 0.02×10³/μL; the differential count was not available due to the low WBC count. His hemoglobin concentration was 8.5 g/dL and his platelet count was 230×10³/μL. After a thorough washout, the repeated analysis showed the same results as initially obtained. The sample was also re-examined in the “low-WBC” mode, which gave a WBC count of 7.16×10³/μL. A peripheral blood smear (PBS) was automatically prepared and reviewed according to the laboratory protocol. The WBC count in the PBS was closer to the 7.16×10³/μL count obtained in the “low-WBC” mode. WBC aggregation was not observed. A few hypersegmented neutrophils were seen and the monocytes were slightly activated.

Under suspicion of falsely decreased WBC count, the raw data from the hematology analyzer were reviewed. The analyzer displayed the following error message: “Difference between WNR and WDF. Check the results.” The reported WBC count (0.02×10³/μL) was obtained from the WBC and nucleated red blood cells (WNR) channel, while the WBC result from the white cell differential (WDF) channel was 7.16×10³/μL, the same as the result obtained in ‘low-WBC’ mode, and for which the differential count was available (Table 1). The scattergrams of the two channels were also reviewed (Fig. 1). The scattergram from the WDF channel was more compatible with the PBS findings than the scattergram from the WNR channel. Additional tests were performed to determine the cause of this falsely decreased WBC count. The blood viscosity and neutrophil respiratory burst activity were both within normal limits.

Table 1 . CBC data from the WDF and WNR channels of the Sysmex XN hematology analyzer

	WDF channel	WNR channel
WBCs	7.16×103/μL	0.02×103/μL
Neutrophils	70.1% (5.02×103/μL)	Not available
Lymphocytes	18.4% (1.32×103/μL)
Monocytes	10.5% (0.75×103/μL)
Eosinophils	1.0% (0.07×103/μL)
Basophils	0.0% (0.00×103/μL)
Nucleated RBCs	None	None

Abbreviations: CBC, complete blood cell count; WDF, white cell differential; WNR, white blood cell and nucleated red blood cells; WBC, white blood cell; RBC, red blood cell.

Figure 1. (A) Peripheral blood smear (100×, Wright stain). (B) Scattergrams from the white cell differential (WDF) (left) and white blood cell and nucleated red blood cells (WNR) (right) channels of the Sysmex XN hematology analyzer.

The XN, a fully automated hematology analyzer introduced in 2011, has several new channels compared to the previous model, the Sysmex XE (Sysmex Corporation). Among these, the WNR channel is designed to automatically measure nucleated red blood cells (RBC) at the same time as the CBC is measured. The result from this channel is reported as the default WBC count. The differential WBC count is measured separately through the white cell differential (WDF) channel; however, the WBC count from this channel is not automatically reported. According to the manufacturer, in the WNR channel, samples are treated with a more acidic reagent (pH 2.95–3.05; Lysercell WNR) than that used in the WDF channel (Lysercell WDF). Theoretically, the WBC count from both channels should be the same [3]. However, in the present case, the two channels did not give the same results. Two similar cases have been reported [4, 5].

Zandecki et al. [1] reviewed factors that can result in inaccurate WBC counts. First, aggregation of polymorphonuclear neutrophils or lymphocytes in blood collected in ethylenediaminetetraacetic acid (EDTA) may result in low WBC counts. Certain anticoagulants, particularly tripotassium (K3)-EDTA, can also cause spurious WBC counts. In our case, no aggregation of leukocytes was seen in the PBS and the sample was collected in dipotassium (K2)-EDTA. It has also been hypothesized that WBCs that are fragile due to cytotoxic agents such as chemotherapy are more prone to lysing, potentially resulting in low WBC counts. However, another report found that chemotherapy did not cause a difference in WBC counts between the WNR and WDF channels [3]. There was also no evidence of WBC lysis in our case. Finally, hyaluronic acid has been reported as a possible cause of WBC aggregation in the presence of the acidic WNR reagent, leading to inaccurate counts [6]. This could be an issue in paraneoplastic syndromes with hypersecretion of hyaluronic acid or hyaluronidase. Unfortunately, we did not measure hyaluronate or hyaluronidase levels to evaluate this possibility.

All clinical laboratories operating automated hematology analyzers should be aware of this phenomenon. When the instrument displays the error message “Difference between WNR and WDF. Check the results”, the raw data from the hematology analyzer and PBS should be carefully reviewed.

None declared.