Cell Population Data NE-SFL and MO-WX From Sysmex XN-3000 Can Provide Additional Information for Exclusion of Acute Promyelocytic Leukemia From Other Acute Myeloid Leukemias: A Preliminary Study.

Dear Editor,

Acute promyelocytic leukemia (APL) is diagnosed by detection of PML-RARA using reverse transcriptase (RT)-PCR, but reporting of results takes at least two days, and this situation is an obstacle for prompt management of patients. Treatment and disease course in APL are different from those in other AMLs because APL patients require all-trans-retinoic acid (ATRA) for remission and for reducing the risk of disseminated intravascular coagulation (DIC), which is a dangerous complication of APL. Hence, treatment of APL with ATRA can be more important than starting chemotherapy in other AMLs, and rapid exclusion of APL from other AMLs is important, allowing for rapid implementation of appropriate treatment. Nonetheless, discrimination of APL from other AMLs by morphological evaluation can cause misdiagnosis in some cases with APL-like morphology. The recently launched Sysmex XN-3000 analyzer (Sysmex, Kobe, Japan) can show various cell population data (CPD), and some of these characteristics, e.g., NE-SFL [fluorescent light intensity of the neutrophil area on the WDF (white blood cell differential) scattergram] and NE-WY (fluorescent light distribution width of the neutrophil area on the WDF scattergram) provide useful information for detection of sepsis [1]. We evaluated the performance of CPD items in peripheral blood (PB) for differential diagnosis of APL and other AMLs.

Ten APL patients and 21 patients with other AMLs diagnosed from July 2014 to December 2015 at Pusan National University Hospital, Korea were prospectively enrolled. Thirty-one PB samples obtained at diagnosis were analyzed on an XN-3000 analyzer. Demographic characteristics, complete blood cell counts (CBC) results, and 46 CPD items were compared between APL patients and other AML patients. For CPD items with significant differences, ROC curve analysis was performed to evaluate the performance in differential diagnosis of APL and other AMLs and to determine the best cutoff values for APL. Using these cutoff values, the sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy for each item were calculated. This study was approved by the institutional review board of Pusan National University Hospital.

Results of APL patients showed significantly lower values on four CPD items [NE-SFL, LY-WY (fluorescent light distribution width of the lymphocyte area on the WDF scattergram), MO-WX (lateral scattered light distribution width of the monocyte area on the WDF scattergram), and MO-WY (fluorescent light distribution width of the monocyte area on the WDF scattergram)] compared with those of other AML patients (Table 1). NE-SFL showed the greatest AUC (0.829) and higher sensitivity (90.0%) and NPV (93.3%) than other characteristics did in the discrimination of APL from other AMLs with the cutoff ≤49.8. In contrast, MO-WX showed higher specificity (90.5%), PPV (77.8%), and accuracy (83.9%) with the cutoff ≤260.0. Additional analysis revealed that the combination "NE-SFL ≤49.8 or MO-WX ≤260.0" shows the highest sensitivity (100.0%) and NPV (100.0%), whereas the combination "NE-SFL ≤ 49.8 and MO-WX ≤260.0" yields the highest specificity (100.0%) and PPV (100.0%) (Table 2).

Table 1

Comparison of the routine and cell population data (CPD) items obtained on a XN-3000 automatic blood cell analyzer between patients with AML and those with acute promyelocytic leukemia (APL)

Items	Results, median (range)		P value		Items	Results, median (range)		P value
	AML (N=21)	APL (N=10)				AML (N=21)	APL (N=10)
Sex (M:F)	12:9	5:5	0.709		LY-WX	501.0 (289.0–709.0)	502.5 (421.0–782.0)	0.833
Age (yr)	65.0 (19.0–83.0)	43.0 (26.0–68.0)	0.021		LY-WY	915.0 (713.0–1,750.0)	767.5 (615.0–1,017.0)	0.008
WBC (×109/L)	8.94 (0.45–204.0)	3.57 (0.44–59.86)	0.043		LY-WZ	604.0 (503.0–845.0)	590.0 (517.0–840.0)	0.800
HGB (×10 g/L)	8.8 (6.6–11.7)	9.9 (7.8–11.6)	0.019		MO-WX	351.0 (67.0-514.0)	237.0 (56.0-375.0)	0.009
PLT (×109/L)	33.0 (8.0–202.0)	41.0 (13.0–377.0)	0.447		MO-WY	862.0 (33.0–1,313.0)	692.5 (32.0–904.0)	0.049
PB Blasts (%)	30.0 (10.0–80.0)	3.0 (1.0–20.0)	<0.001		MO-WZ	675.0 (271.0–932.0)	713.5 (42.0–825.0)	0.899
PB Promyelocytes (%)	0.0 (0.0–2.0)	72.5 (50.0–90.0)	<0.001		WBC-N ( × 109/L)	8.94 (0.45–204.0)	3.57 (0.44–59.86)	0.053
FAB classification (M1: M2: M3: M4: M5: M7)	3:13:0:1:3:1	0:0:10:0:0:0	<0.001		WBC-D (× 109/L)	8.78 (0.46–202.98)	3.40 (0.41–59.09)	0.053
NRBC (×1012/L)	0.02 (0.00–2.65)	0.02 (0.00–0.23)	<0.001		TNC-N ( × 109/L)	8.97 (0.46–204.63)	3.59 (0.45–60.09)	0.053
NRBC (%)	0.20 (0.00–11.40)	0.45 (0.00–2.30)	0.468		TNC-D ( × 109/L )	8.81 (0.47–203.61)	3.42 (0.42–59.32)	0.053
NEUT (×109/L)	2.43 (0.08–111.49)	1.43 (0.20–7.96)	0.473		Delta NEUT (× 109/L)	1.26 (0.07–76.97)	1.41 (0.08–7.72)	0.512
LYMPH (×109/L)	3.05 (0.26–45.67)	0.91 (0.16–13.52)	0.056		Delta NEUT (%)	16.90 (1.30–74.30)	39.0 (5.40–82.60)	0.118
MONO (×109/L)	3.16 (0.06–108.62)	0.40 (0.01–38.33)	0.051		Delta LYMPH ( × 109/L)	3.00 (0.26–45.16)	0.91 (0.15–13.52)	0.051
EOSINO (×109/L)	0.01 (0.00–16.09)	0.01 (0.00–0.32)	0.628		Delta LYMPH (%)	26.90 (3.40–91.90)	29.80 (12.60–58.50)	0.866
BASO (×109/L)	0.03 (0.00–0.59)	0.01 (0.00–0.04)	0.082		HFLC (× 109/L)	0.03 (0.00–0.76)	0.00 (0.00–0.01)	0.060
NEUT (%)	21.30 (1.30–74.80)	51.75 (13.30–83.20)	0.076		HFLC (%)	0.20 (0.00–7.60)	0.00 (0.00–2.30)	0.267
LYMPH (%)	30.20 (3.40–91.90)	30.40 (12.60–58.50)	0.966		BASO-D (× 109/L)	0.04 (0.00–1.69)	0.01 (0.00–0.06)	0.054
MONO (%)	27.30 (3.90–70.60)	15.20 (1.00–64.00)	0.099		BASO-D (%)	0.40 (0.00–2.20)	0.05 (0.00–2.30)	0.165
EOSINO (%)	0.10 (0.00–7.90)	0.00 (0.00–4.30)	0.856		RBC-He (pg)	31.50 (16.90–38.00)	30.05 (28.70–38.00)	0.767
BASO (%)	0.30 (0.00–1.40)	0.15 (0.00–0.70)	0.354		Delta-He (pg)	2.30 (−6.90 to 5.80)	4.20 (−0.20 to 6.40)	0.113
IG (×109/L)	0.05 (0.00–34.52)	0.04 (0.00–0.24)	0.719		RET-Y (ch)	179.2 (110.7–195.2)	184.6 (168.0–195.8)	0.272
IG (%)	0.70 (0.00–16.90)	0.55 (0.00–25.00)	0.672		RET-RBC-Y (ch)	174.5 (112.6–184.8)	169.5 (157.7–183.1)	0.499
NE-SSC (ch)	148.8 (124.8–181.6)	141.0 (123.7–165.6)	0.375		IRF-Y (ch)	171.3 (145.6–196.6)	181.5 (160.4–199.3)	0.183
NE-SFL (ch)	52.2 (42.9–89.7)	45.5 (40.7–52.4)	0.004		RPI	0.40 (0.10–2.40)	0.45 (0.00–2.40)	0.881
NE-FSC (ch)	74.8 (44.0–91.9)	74.0 (58.6–90.6)	0.673		RET-UPP	2.00 (0.00–81.00)	4.00 (0.00–101.00)	0.913
LY-X (ch)	84.8 (74.3–97.7)	83.3 (78.5–117.9)	0.342		RET_TNC	105.0 (4.0–1,883.0)	25.5 (9.0–470.0)	0.070
LY-Y (ch)	73.6 (62.9–99.6)	76.6 (64.3–108.9)	0.099		HYPO-He (%)	0.70 (0.10–47.90)	0.55 (0.20–3.70)	0.983
LY-Z (ch)	61.3 (54.6–65.6)	63.5 (59.8–70.7)	0.056		HYPER-He (%)	0.40 (0.10–3.10)	0.40 (0.20–0.90)	0.651
MO-X (ch)	120.7 (101.2–141.0)	123.2 (107.0–130.6)	0.460		FRC (× 1012/L)	0.014 (0.000–0.180)	0.011 (0.000–0.090)	0.526
MO-Y (ch)	119.3 (82.8–223.8)	113.5 (56.5–142.0)	0.291		FRC (%)	0.64 (0.00–5.07)	0.41 (0.01–2.67)	0.597
MO-Z (ch)	65.4 (57.5–76.8)	68.6 (48.0–77.6)	0.486		IPF (× 109/L)	2.90 (0.10–15.50)	1.20 (0.50–13.90)	0.190
NE-WX	439.0 (33.0–649.0)	362.0 (278.0–574.0)	0.499		H-IPF	1.60 (0.30–34.10)	0.90 (0.20–9.10)	0.197
NE-WY	998.0 (97.0–3,086.0)	780.0 (537.0–2,295.0)	0.139		MicroR (%)	2.50 (0.40–48.10)	2.90 (0.70–5.90)	0.597
NE-WZ	900.0 (91.0–1,285.0)	852.5 (685.0–1,382.0)	0.999		MacroR (%)	5.40 (2.50–21.10)	3.45 (2.60–10.40)	0.066

P values were obtained by the Mann-Whitney U test (for all items except sex and FAB classification) and χ2 test (for sex and FAB classification).

Abbreviations and definitions: WBC, white blood cells; HGB, hemoglobin; PLT, platelets; PB peripheral blood; FAB, French-America-British; NRBC, nucleated red blood cells; NEUT, neutrophils; LYMPH, lymphocytes; MONO, monocytes; EOSINO, eosinophils; BASO, basophils; IG, immature granulocytes; NE-SSC, the lateral scattered light intensity of the NEUT area on the WDF scattergram; NE-SFL, the fluorescent light intensity of the NEUT area on the WDF scattergram; NE-FSC, the forward-scattered light intensity of the NEUT area on the WDF scattergram; LY-X, the lateral scattered light intensity of the LYMPH area on the WDF scattergram; LY-Y, the fluorescent light intensity of the LYMPH area on the WDF scattergram; LY-Z, the forward-scattered light intensity of the LYMPH area on the WDF scattergram; MO-X, the lateral scattered light intensity of the MONO area on the WDF scattergram; MO-Y, the fluorescent light intensity of the MONO area on the WDF scattergram; MO-Z, the forward-scattered light intensity of the MONO area on the WDF scattergram; NE-WX, the lateral scattered light distribution width of the NEUT area on the WDF scattergram; NE-WY, the fluorescent light distribution width of the NEUT area on the WDF scattergram; NE-WZ, the forward-scattered light distribution width of the NEUT area on the WDF scattergram; LY-WX, the lateral scattered light distribution width of the LYMPH area on the WDF scattergram; LY-WY, the fluorescent light distribution width of the LYMPH area on the WDF scattergram; LY-WZ, the forward-scattered light distribution width of the LYMPH area on the WDF scattergram; MO-WX, the lateral scattered light distribution width of the MONO area on the WDF scattergram; MO-WY, the fluorescent light distribution width of the MONO area on the WDF scattergram; MO-WZ, the forward-scattered light distribution width of the MONO area on the WDF scattergram; WBC-N, the WBC count calculated from the WNR channel; WBC-D, the WBC count calculated from the WDF channel; TNC-N, the total nucleated cell count (WBC+NRBC) calculated from the WNR channel; TNC-D, the total nuclear cell count (WBC+NRBC) calculated from the WDF channel; Delta NEUT, the number of particles obtained by subtracting the IG count from the NEUT count; Delta LYMPH, the number of particles obtained by subtracting the HFLC count from the LYMPH count; HFLC, the count of the upper LYMPH area of the WDF scattergram; BASO-D, the basophil counts calculated from the WDF channel; RBC-He, the correlation between RBC-Y and MCH to convert RBC-Y into [pg] units; Delta-He, subtraction of RBC-He from RET-Hb; RET-Y, the forward scattered light intensity of the RET area on the RET scattergram; RET-RBC-Y, the forward scattered light intensity of RBC area on the RET scattergram; IRF-Y, the intensity of forward scattered light from the IRF area on the RET scattergram; RPI, reticulocyte production index; RET-UPP, the count in the upper area of the RET scattergram; RET-TNC, the count in the TNC area of the RET scattergram; HYPO-He, the ratio of the count in the low level area of the forward scattered light signal in the RBC (mature red blood cell) area of the RET scattergram, to mature red blood cells; HYPER-He, the ratio of the count in the high-level area of the forward scattered light signal in the RBC (mature red blood cell) area of the RET scattergram to mature red blood cells; FRC, the count in a specific area below the RBC area in the RET scattergram; IPF, immature platelet fraction; H-IPF, the ratio to the total platelet count of the count of platelets that appear in the area of stronger fluorescent light intensity within the IPF area of the PLT-F scattergram; MicroR, Micro RBC ratio; MacroR, Macro RBC ratio; ch, channel.

Table 2

ROC analysis results on the four cell population data (CPD) items with significant differences for discrimination of patients with acute promyelocytic leukemia from those with other AMLs

Items	AUC (95% CI)	Best cutoff	Sensitivity (95% CI)	Specificity (95% CI)	NPV	PPV	Accuracy (95% CI)	P value*
								NE-SFL	LY-WY	MO-WX	MO-WY
NE-SFL	0.829 (0.651–0.939),	≤49.8	90.0 (69.6–98.2)	66.7 (45.4–82.8)	93.3	56.3	74.2 (56.8–86.3)	NA	0.772	0.715	0.335
LY-WY	0.800 (0.618–0.921)	≤808.0	70.0 (39.7–89.2)	81.0 (60.0–92.3)	85.0	63.6	77.4 (60.2–88.6)	0.772	NA	0.963	0.405
MO-WX	0.795 (0.612–0.918)	≤260.0	70.0 (39.7–89.2)	90.5 (71.1–97.4)	86.4	77.8	83.9 (67.4–92.9)	0.715	0.963	NA	0.270
MO-WY	0.721 (0.532–0.866)	≤774.0	80.0 (49.0–94.3)	61.9 (40.9–79.3)	86.7	50.0	67.7 (50.1–81.4)	0.335	0.405	0.270	NA

*P values characterize the results of comparison of AUC values between two items.

Abbreviations: AUC, area under the curve; CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; NE-SFL, the fluorescent light intensity of the NEUT area on the WDF scattergram; LY-WY, the fluorescent light distribution width of the LYMPH area on the WDF scattergram; MO-WX, the lateral scattered light distribution width of the MONO area on the WDF scattergram; MO-WY, the fluorescent light distribution width of the MONO area on the WDF scattergram; NA, not applicable.

Low NE-SFL values may reflect increased infiltration of more mature granulocytes (such as promyelocytes) than myeloblasts because low fluorescence intensity indicates a low RNA/DNA ratio in more mature cells [12]. APL involves abundant promyelocytes, and this situation may partly explain our results. Significantly lower MO-WX in APL patients may be the result of lower monocyte counts in APL (versus AML), as found in our study, possibly yielding small distribution width.

Clinically, rapid exclusion of APL from other AMLs may be more important than accurate confirmation of APL by more time-consuming diagnostic procedures. The present study showed that CPD items, NE-SFL and MO-WX, would be useful for the discrimination of APL from other AMLs, especially for rapid exclusion of APL with high sensitivity and NPV. The discriminating power of the PB promyelocyte count may be better, but the time benefit of the two CPD items may be greater. Thus, we hypothesize the use of NE-SFL and MO-WX as criteria for rapid exclusion of APL, not as diagnostic parameters, because we cannot diagnose APL without PML-RARA, even if NE-SFL and MO-WX are lower than cutoffs. The sample size here is small and therefore, our results should be regarded as a preliminary data, and more comprehensive research such as a multicenter study should be performed to confirm our preliminary results.