Full Terms & Conditions of access and use can be found athttps://www.tandfonline.com/action/journalInformation?journalCode=iclb20

Scandinavian Journal of Clinical and LaboratoryInvestigation

ISSN: 0036-5513 (Print) 1502-7686 (Online) Journal homepage: https://www.tandfonline.com/loi/iclb20

Report of a new six-panel flow cytometry markerfor early differential diagnosis of APL from HLA-DRnegative Non-APL leukemia

Mohammad Mosleh, Mahdieh Mehrpouri, Sasan Ghaffari, Zeinab Saei,Mahnaz Agaeipoor, Farzaneh Jadali, Esmaeil Shahabi Satlsar & RoohollahGholampour

To cite this article: Mohammad Mosleh, Mahdieh Mehrpouri, Sasan Ghaffari, Zeinab Saei,Mahnaz Agaeipoor, Farzaneh Jadali, Esmaeil Shahabi Satlsar & Roohollah Gholampour (2019):Report of a new six-panel flow cytometry marker for early differential diagnosis of APL from HLA-DR negative Non-APL leukemia, Scandinavian Journal of Clinical and Laboratory Investigation,DOI: 10.1080/00365513.2019.1700427

To link to this article: https://doi.org/10.1080/00365513.2019.1700427

Published online: 12 Dec 2019.

Submit your article to this journal

View related articles

View Crossmark data

ORIGINAL ARTICLE

Report of a new six-panel flow cytometry marker for early differential diagnosisof APL from HLA-DR negative Non-APL leukemia

Mohammad Mosleha,b, Mahdieh Mehrpourib, Sasan Ghaffaric,d, Zeinab Saeie, Mahnaz Agaeipoore,Farzaneh Jadalie, Esmaeil Shahabi Satlsara,e and Roohollah Gholampourf

aFaculty of Paramedical sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran; bRasad Pathobiology and GeneticsLaboratory, Department of Flow Cytometry, Tehran, Iran; cStudent Scientific Research Center, Tehran University of Medical Sciences, Tehran,Iran; dCell-based Therapies Research Center, Digestive Disease Research Institute, Tehran University of Medical Sciences, Tehran, Iran;eDepartment of Flow Cytometry, Takhte Tavous Pathobiology Laboratory, Tehran, Iran; fTakhte Tavous Pathobiology Lab

ABSTRACTAlthough AML-M3 (APL) and HLA-DR negative non-APL are characterized by negative HLA-DR antigen,they are different entities with similar morphology in some cases. The aim of this study is the precise,differential diagnosis of APL from HLA-DR negative non-APL by flow cytometry to narrow the diagno-sis window. Bone marrow or blood samples of 580 AML patients were analyzed, and flow cytometryand molecular analysis were performed for the diagnosis of blood disorders. In 105 HLA-DR negativeAML patients, expression of HLA-DR, CD33, CD117, CD11b, CD64, CD34, CD9 and myeloperoxidasestaining pattern were evaluated. Fifty-six patients were diagnosed with APL, and 49 patients werediagnosed with HLA-DR negative non-APL. The APL blasts expressed CD33, CD117, CD64, and CD9 in100%, 80.3%, 94.6%, and 100% of the cases, respectively. HLA-DR negative non-APL blasts expressedCD33, CD117, CD64 and CD9 in 75.5%, 59.1%, 32.6%, and 73.4% of the cases, respectively. APL cellswere negative for HLA-DR, CD11b, and CD34 in 96.4%, 94.6%, and 91.0%, respectively. Blasts in HLA-DR negative non M3-AML were negative for CD11b, CD117, and CD34 in 77.5%, 40.9%, and 22.4%,respectively. We also investigated myeloperoxidase (MPO) staining pattern and found strong diffusereaction in APL cells while HLA-DR negative non-APL cells showed focal positive reaction. In all of theAPL patients, except for one, PML/RARA translocation was positive, and in another case with HLA-DRnegative non-APL, PML/RARA and other translocations were not detected. The six-panel combinationprofile rapidly and specifically identifies APL from other HLA-DR negative AML.

ARTICLE HISTORYReceived 29 June 2019Accepted 28 November 2019

KEYWORDSAPL; HLA-DR negative AML;flow cytometry; PML-RARA;CD64; CD9

Introduction

Acute myeloid leukemia (AML) is a type of cancer in whichbone marrow (BM), peripheral blood (PB), and other softtissues are infiltrated by abnormally-differentiated cells ofthe hematopoietic system [1]. Although AML was deemedincurable for the past 50 years, 30–40% of the patients whoare 60 years of age or younger and 5–15% of the patientsolder than 60 years currently achieve remission [2].According to the French-American-British (FAB) classifica-tion and the World Health Organization’s new classificationin 2017, acute promyelocytic leukemia (APL) is a subtype ofacute leukemia, first described in 1957 [3]. A balanced,reciprocal translocation between chromosome 15 (PMLgene) and 17 (RARa gene) can be detected in 70–90% casesof APL [4,5]. Rare, variant translocations with classic mor-phologic findings of APL are seen in 1–2% of APL cases.Translocation between chromosome 11 (ZBTB16 gene) and17 (RARa gene) has a distinct morphology in APL with amore regular nucleus compared to the bilobed nucleus typ-ically found in APL. APL blasts with t(11:17) are character-ized by coarse granules and, less often, by fine or no

granules [6]. APL is historically characterized by a rapid,fatal course with a high incidence of hemorrhagic complica-tions [7]. After the introduction of arsenic tetra retinoicacid (ATRA), however, the mortality rate associated withAPL has decreased, resulting in complete remission in mostpatients [8].

Due to a high risk of fatal bleeding, current protocolsrecommend APL to be considered as a medical urgencyupon diagnosis [9]. Immunophenotyping of arrested blastcells by flow cytometry has become a rapid and reliablemethod of lineage determination, since it is faster and lesscostly than cytogenetic tests. In flow cytometry evaluation,APL blasts are CD11b negative, CD34 negative, andCD117± and positive for CD13, CD33, CD64, and cytoplas-mic myeloperoxidase (MPO). MPO staining of APL blastsalso shows a diffuse and strong positive patterns. The pres-ence of Auer rods is also witnessed in the cytoplasm of saidblasts. Negative expression of CD10, CD11a, CD11c,CD45RO, and CD133 is another common feature [10].Most AML cases express HLA Class II antigens, such asHLA-DR, on the surface of their blast cells upon diagnosis,and HLA-DR expression in non-APL cases is common.

CONTACT Esmaeil Shahabi Satlsar esmaeilshahabi@yahoo.com Department of Flow Cytometry, Takhte Tavous Pathobiology Laboratory, Tehran, Iran� 2019 Medisinsk Fysiologisk Forenings Forlag (MFFF)

SCANDINAVIAN JOURNAL OF CLINICAL AND LABORATORY INVESTIGATIONhttps://doi.org/10.1080/00365513.2019.1700427

APL, known to lack the HLA-DR antigen, is an exception.Lack of HLA-DR antigen expression is hence anothermarker used for distinguishing APL from other AML sub-types. However, data suggest that HLA-DR negativity canalso be witnessed in non-M3 AML. Due to bleeding ten-dency in some cases of HLA-DR negative non M3-AML[11], early detection and accurate differentiation of APL iscrucial to the patient’s health.

In this study, we evaluated HLA-DR, CD117, CD11b,CD34, CD64, and CD9 in a six-panel combination for a def-inite distinction between APL blasts and HLA-DR negativenon-APL blasts.

Materials and methods

Patients’ selection

In this study, 580 AML patients who referred to TakhteTavous Pathobiology Laboratory, Tehran, Iran, were studied.BM and PB samples were taken from the patients betweenJune 2014 and September 2018 after obtaining informedconsent. The diagnosis of AML subtypes was based on crite-ria established by the WHO classification of tumors of hem-atopoietic and lymphoid tissues [12].The samples wereevaluated using flow cytometry with our specific panel.Then, APL cases were confirmed by detecting the presenceof PML-RARa or variant translocation with RT-qPCR. Weonly included de novo APL and HLA-DR negative non-APLpatients and excluded other AML subtypes, such as AML-M5 and -M6. This study was approved by the EthicsCommittee of Shahid Beheshti Medical University.

Morphological evaluation

Myeloperoxidase staining was applied using Sigma-AldrichMyeloperoxidase Leukocyte Kit. All steps were performedper the manufacturer’s instructions. Briefly, freshly preparedBM or whole blood films were fixed with a formaldehyde-ethanol solution and then washed and dried in the dark.They were then placed in peroxidase indicator and washedwith running tap water. Finally, slides were counterstainedin acid hematoxylin solution and rinsed in running deion-ized water. Air-dried slides were examined by anoptic microscope.

Real-time quantitative RT-PCR assay

PML-RARa-L, PML-RARa-S, and PLZF-RARa fusion tran-script levels were quantified by RT-qPCR. First, RNA wasextracted from BM samples using Ribo Exreagent (GeneAll,Portugal) according to the manufacturer’s protocol. Wethen determined the integrity of extracted RNA by gel elec-trophoresis, and samples with intact 28S and 18S were con-sidered appropriate. Reverse transcription was performed ontotal RNA (1 mg) using SuperScript II reverse transcriptase(Invitrogen, USA). PCR analyses of the different types offusion transcripts were performed using a Taqman-basedreal-time quantitative PCR in an ABI PRISM 7000 sequence

detection system (Applied Biosystems [ABI], Foster City,CA) and ABL was selected as the internal control. We eval-uated our results based on the Europe Against Cancer(EAC) program which defines a framework for assessmentand standardization of RT-qPCR assays using common pri-mer-probe sets for each respective fusion transcript [13].

Flow cytometry

Five ml of fresh BM or blood sample was collected intovacutainer tubes containing K2-EDTA and gently homogen-ized several times for sufficient mixing. Samples wereanalyzed using a large panel of the following fluorochrome-conjugated monoclonal antibodies. These antibodiesincluded CD64 (PE, Dako), CD117 (PerCP, Immunostep),CD34 (PC7, Beckman coulter) HLA-Dr (FITC, Dako),CD14 (FITC,Dako, Denmark), CD45 (ECD, Coulter),CD11b (ECD, Coulter), CD13 (PE,Dako), CD33 (FITC,Dako), CD10 (FITC,Dako), CD3 (PE,Dako),CD5(FITC,Dako), CD19(PE,Dako), CD20 (FITC,Dako),andCD9 (FITC Coulter). BM and blood cells were incubatedwith antibodies for 30min at 4 �C, and then erythrocyteswere lysed using a standard lyse/wash technique. Antigenicexpression in blast cells was systematically analyzed bymulti-parametric flow cytometry (Beckman-Coulter FC500,USA) via MXP software. Cell surface antigen expression wasconsidered positive if there was >20% expression of theanalyzed events, while a cutoff of >10% was set for CD117and CD34 antigens.

Statistical analysis

All statistical analyses were performed by SPSS software(version 21). T-test was used to analyze the differencesbetween APL and HLA-DR negative non-APL groups basedon their distribution. We also used one-way ANOVA toevaluate three or more than three groups, and a p-value of<.05 was considered statistically significant.

Results

Patients’ baseline characteristics

Of the 580 AML patients admitted to our laboratory, 105patients were negative for HLA-DR antigen. Of these, 56patients were diagnosed with APL and 49 with non-APLimmunophenotype. The median age of APL patients (25male and 31 female) was 26.5 years (range, 1� 52 years),and the median age of non-APL patients (27 male and 22female) was 47.5 years (range, 17� 78 years). Following theuse of molecular assay to confirm the results of flow cytom-etry, it was found that immunophenotyping performancefor an exact diagnosis of APL had 98.2% sensitivity, 97.9%specificity, and 98.0% accuracy. The positive predictive valuewas 98.1%.

2 M. MOSLEH ET AL.

Myeloperoxidase staining patterns were differentbetween APL and HLA-DR negative non-APL specimen

In this study, all BM or blood samples were stained withWright’s and myeloperoxidase staining. The Wright’s stain-ing in the APL group showed that 27 patients (48.2%) hadhypergranular variation with abundant cytoplasmic granules,Auer rods, and faggot cells, while 29 patients) (51.8%) weremicrogranular variant with convoluted and dumbbell-shapednuclei (Figure 1). In the HLA-DR negative non-APL group,we observed cytoplasmic granules and convoluted nuclei insome cases.

MPO staining showed a diffuse and strong positive reac-tion on promyelocytes in APL patients, but there weregranular or focal patterns in patients with HLA-DR negativenon-APL (Figure 2).

Immunophenotyping is helpful with distinguishing APLfrom HLA-DR negative non-APL

We first compared the CD45/SSC pattern of leukemia cellsbetween APL and HLA-DR negative non-APL groups. Allleukemia cells in the APL group were located in the mono-cytic region with dim to moderate CD45 expression andmoderate SSC pattern. In most of the HLA-DR negativenon-APL samples, blasts were located in the lymphoidregion (blastic region) which is indicated by the dim expres-sion of CD45 and low SSC pattern. Some HLA-DR negativenon-APL cases with monocytic differentiation had a dim tomoderate CD45 expression with higher SSC pattern similarto APL blasts.

Non-APL cases with high expression of CD34 and HLA-DR are easily distinguished from APL. However, the

Figure 1. Peripheral blood smear stained with Wright’s stain. APL cases were both hypergranular with Auer rods (A) and microgranular (B). HLA-DR negative non-APL cases showed cytoplasmic granules (C).

Figure 2. Myeloperoxidase staining patterns in classic APL (A), variant APL (B), and HLA-DR negative non-APL (C) cases.

Figure 3. Flowchart describing the differential diagnosis of APL from HLA-DR negative non-APL using flow cytometry.

SCANDINAVIAN JOURNAL OF CLINICAL AND LABORATORY INVESTIGATION 3

diagnosis of CD34 and HLA-DR negative non-APL cases isa challenge. In most APL cases, blast cells are negative forCD34, HLA-DR, and CD11b while almost consistentlyexpressing CD117, CD64, and CD9 (Figure 3). Evaluatingthe expression pattern of the studied markers revealed thatdifferences in expression of these markers, despite beingpositive in both APL and non-APL groups, can help distin-guish between the two leukemia (Table 1). In APL patients,expression of CD117 is associated with dim and partialexpression of CD64 (p< .05). In HLA-DR negative non-APL patients, there was no correlation between CD117 andCD64 expression. On the other hand, CD9 expression alsoshows a significant difference between the two groups(p< .05). In APL cases, blast cells a showed moderate tobright, and homogenous expression of CD9, but in HLA-DR negative non-APL, blast cells showed dim andHeterogeneous or negative CD9 expression (Figure 4).

A small percentage of APL blasts showed unusual expres-sion patterns. Five cases (8.9%) had a dim and partialexpression of CD34, positive expression of CD117 and CD9,

but a negative expression of HLA-DR, CD11b, and CD64.Three cases (5.4%) had a partial expression of CD11b andnegative for HLA-DR, CD34 and partial expression ofCD117 and positive CD64 expression. Two cases (3.6%) hada positive expression of HLA-DR, CD117, and CD64, but anegative expression for CD11b and CD34 (Table 2).

In 4 cases with AML-M2 (FAB classification), CD64expression was seen in a similar pattern with APL blasts. Inthese 4 cases blasts were negative for CD11b and HLA-DRand expressed CD117 but had a dim to moderate expressionof CD34. Blast cells were also positive for CD64 but a pat-tern of CD9 expression in these 4 cases showed a dim andheterogeneous expression. By using the 6-panel combin-ation, these exceptional cases can be distinguished fromAPL. However, considering the low SSC and the pattern ofMPO staining, they can be easily differentiated from APLby focal MPO reaction. Molecular analysis was carried outfor the presence of the PML-RARa and was negative inthese cases.

In HLA-DR negative AML patients, there is was signifi-cant overlap in the expression of the studied markers com-pared to APL. In 32.6% of the HLA-DR negative non-APLpatients, blast cells were positive for CD64, which included12 cases with M4/M5 subgroups (co-expressed with CD14)and 4 cases with M2 subgroup of AML. In 59.1% of theHLA-DR negative non-APL patients (29/49), CD117 waspositive with different fluorescent intensities. In 17 cases,the expression of CD117 was seen with co-expression ofCD34, and in the remaining patients (12 cases), CD34 wasnegative. In HLA-DR negative patients with CD34 expres-sion, blast cells showed negative expression of CD11b and

Table 1. Differential expression of CD117, CD64, CD11b, CD9 and CD34 inAPL and HLA-DR negative non-M3 AML.

CD Markers APL (n¼ 56)HLA-DR negativenon-APL (n¼ 49) p-Value

CD117þ 45 (80.3%) 29 (59.1%) <.001CD64þ 53 (94.6%) 20 (32.6%) <.001CD11b- 53 (94.6%) 38 (77.5%) <.001CD34- 51 (91%) 11 (22.4%) <.001HLA-DR- 54 (96.4%) 49 (100%) .11CD9 56 (100%) 36 (73.4%) <.001

A p-Value of <.05 was considered significant.

Figure 4. Immunophenotypic analysis of APL and HLA-DR negative patients. Note that dual negative CD117/CD64 was not seen in almost all APL patients. #: DimCD64 is associated with dim CD117 in APL. A: CD117 (FL1) and CD64 (FL2), B: CD34 (FL1) and HLA-DR (FL2), C: CD34 (FL1) and CD11b (FL2), D: CD117 (FL1) and

CD64 (FL2), E: CD117 (FL1) and CD64 (FL2), F: CD34 (FL1) and HLA-DR (FL2). Dual negative CD117/CD64 in HLA-DR negative AML. Dim CD64/negative CD117

in HLA-DR negative AML.

4 M. MOSLEH ET AL.

CD64 and were often positive for CD117. Patients withpositive expression of CD11b in HLA-DR negative non-APLgroup mostly showed M4/M5 immunophenotypes.

Discussion

Early and accurate diagnosis of APL can significantly reduceits mortality. Therefore, it is essential to improve anddevelop the flow cytometric diagnosis of the disease, whichis faster and less expensive than molecular testings. HLA-DR negative AML is typically divided into APL and HLA-DR negative non-APL cases. Other researchers have similarclassification [14,15] and report APL percentages rangingfrom 10% to 12% of all AML cases [16,17]. In this study, wesought to determine a flow cytometry panel for differentialdiagnosis of APL from HLA-DR negative non-APL. Ourresults of analyzing 580 newly-diagnosed AML patientsshow six markers capable of distinguishing between APLand HLA-DR negative non-APL leukemia. Negative expres-sion of HLA-DR, CD11b, and CD34, bright and homogen-ous expression of CD9, positive CD64 and CD117± stronglydiscriminates APL cases from HLA-DR negative non-APLpatients in our six-panel combination flow cytometry. DimCD64 expression is associated with dim CD117 expressionand moderate CD64 expression is associated with negativeCD117 expression.

Some HLA-DR negative non-APL patients have morpho-logic and clinical characteristics similar to APL patients[15]. Yet, Fenu et al. reported three HLA-DR negative AMLpatients who, based on morphology and immunophenotype,were thought to be variants of APL. After cytogenetic andmolecular analyses, however, they were discovered to be‘atypical M2’ subtypes [18]. Thus, negative HLA-DR expres-sion and morphology alone might not be enough for distin-guishing APL from other AML subtypes. CD64 expressionis common in APL but is highly variable [19]. We firstinvestigated CD64 expression as a marker to differentiateAPL from the non-APL specimen and found significantcontrasting expression of CD64 between the two groups.Interestingly, leukemic cells in 94.6% of all APL patientsexpressed heterogeneous dim or moderate expression ofCD64 compared to CD117 expression. In a study by Liuet al., dim and homogeneous CD64 expression was consid-ered appropriate for the diagnosis of APL [20]. Expressionof CD34, CD64, CD11b, CD117, and CD9 was in agreementwith RT-PCR for PML-RARa in APL patients.

Several studies have discussed the immunophenotypicfeatures of APL, which include high SSC, homogeneousexpression of CD33, CD13, CD64, and CD117, and lack ofCD11b, CD11c, CD14, HLA-DR, CD34 expression.Expression of other markers, such as CD2 and CD56, canalso be seen in some cases. However, there is no consensus

view on the immunophenotypic characteristics ofAPL [11,20,21].

Attempts have been made to detect APL using flowcytometry, and the negative expression of CD34 and HLA-DR were introduced as a valuable combination to predictthe disease [22]. We did not observe a simultaneous expres-sion of these markers in any of the cases. This could prob-ably be due to the difference in sample size. As stated in apaper by Dong et al., the diagnosis of APL cannot be solelybased on negative HLA-DR and CD34 expression [11].

Based on side scatter, APL is divided into two groups,low SSC and high SSC [11,20], and in some studies, it isstated that each group expresses the specific markers ofblastic or mature granulocytes. For example, it is mentionedthat CD117 expression is evident in blastic type and is notseen in the granulocytic type [11]. We investigated CD117expression on leukemic blasts of both groups and foundCD117 alone has a low specificity and sensitivity in differen-tiating APL from HLA-DR negative non-APL. Oelschlaegeland colleagues also demonstrated that HLA-DR negativenon-APL patients can be negative for CD117, but APL casesare CD117 positive [14]. Ferrari and colleagues found nosignificant association between CD117 expression in APLversus HLA-DR negative non-APL cases with cup-likenucleus morphology [23].

Importantly, CD117 expression on HLA-DR negativenon-APL cases is mostly associated with positive CD34 andweak MPO staining of blasts. On the other hand, in APLpatients, expression of CD117 was associated with dimexpression of CD64, and negative or dim expression ofCD117 is associated with brighter expression of CD64. Thiscombination is helpful to distinguish between APL andHLA-DR negative non-APL. In our study, all cases, exceptfor one, were positive for CD117 expression, regardless ofthe type of morphology or SSC features.

Considering the myeloperoxidase staining pattern resultsin our study, using MPO staining can raise the sensitivity,specificity, PPV, and NPV of the panel along with theexpression pattern of the mentioned markers. Similar toAPL, MPO reactions can be strong in cases of HLA-DRnegative non-APL, but the pattern of MPO positive stainingis different between these two groups and this difference ismostly overt and clear. In this regard, the type of positivereaction in non-APL cases is focal/granular and in APL isdiffuse. Considering these patterns, it is possible to differen-tiate the type of positive reaction of MPO in these twogroups despite their positivity. There was only an exceptionwith a 15-year-old girl with FAB-M1 morphology who hada diffuse pattern of MPO staining, positive CD64, CD11b,CD117, and negative HLA-DR, CD34, and PML-RARafusion gene transcript. Other flow cytometry studies havefound that heterogeneous expression of CD13, CD15, and

Table 2. APL patients with atypical immunophenotypes. A total of 10 APL patients showed unexpected APL marker expression.

HLA-DR CD34 CD11b CD64 CD117 CD9

Patients 8, 12, 24, 32, 47 Neg Dim and partial expression Neg Neg þ þPatients 2, 19, 54 Neg Neg Partial expression þ þ þPatients 5, 22 þ Neg Neg þ þ þ

SCANDINAVIAN JOURNAL OF CLINICAL AND LABORATORY INVESTIGATION 5

negative CD34 are helpful to distinguish between HLA-DRnegative non-M3 AMLs and AML-M3 [24].

Conclusion

In summary, it may be concluded that HLA-DR-, CD11b-,CD34-, CD64þ, CD9þ, and CD117�/þ six-panel flowcytometry could distinguish APL from HLA-DR negativenon-APL, which in combination with MPO staining, resultsin 100% sensitivity and specificity.

Acknowledgements

We thank the technical staff of Takhte Tavous Pathobiology Lab forgathering the data.

Disclosure statement

No potential conflict of interest was reported by the authors.

References

[1] Estey EH. Acute myeloid leukemia: 2019 update on risk-stratifi-cation and management. Am J Hematol. 2018;93(10):1267–1291.

[2] Cerrano M, Itzykson R. New treatment options for acute mye-loid leukemia in 2019. Curr Oncol Rep. 2019;21(2):16.

[3] Welch JS, Ley TJ, Link DC, et al. The origin and evolution ofmutations in acute myeloid leukemia. Cell. 2012;150(2):264–278.

[4] Rowley J, Golomb H, Dougherty C. 15/17 translocation, a con-sistent chromosomal change in acute promyelocytic leukaemia.Lancet. 1977;309(8010):549–550.

[5] de Th�e H, Chomienne C, Lanotte M, et al. The t (15; 17) trans-location of acute promyelocytic leukaemia fuses the retinoicacid receptor a gene to a novel transcribed locus. Nature. 1990;347(6293):558.

[6] Adams J, Nassiri M. Acute promyelocytic leukemia: a reviewand discussion of variant translocations. Arch Pathol Lab Med.2015;139(10):1308–1313.

[7] Choudhry A, DeLoughery TG. Bleeding and thrombosis inacute promyelocytic leukemia. Am J Hematol. 2012;87(6):596–603.

[8] Degos L, Wang ZY. All trans retinoic acid in acute promyelo-cytic leukemia. Oncogene. 2001;20(49):7140.

[9] Coombs C, Tavakkoli M, Tallman M. Acute promyelocytic leu-kemia: where did we start, where are we now, and the future.Blood Cancer J. 2015;5(4):e304.

[10] Peters JM, Ansari MQ. Multiparameter flow cytometry in thediagnosis and management of acute leukemia. Arch Pathol LabMed. 2011;135(1):44–54.

[11] Dong HY, Kung JX, Bhardwaj V, et al. Flow cytometry rapidlyidentifies all acute promyelocytic leukemias with high specificityindependent of underlying cytogenetic abnormalities. Am JClin Pathol. 2011;135(1):76–84.

[12] Vardiman JW, Harris NL, Brunning RD. The World HealthOrganization (WHO) classification of the myeloid neoplasms.Blood. 2002;100(7):2292–2302.

[13] Gabert J, Beillard E, Van der Velden V, et al. Standardizationand quality control studies of ‘real-time’quantitative reversetranscriptase polymerase chain reaction of fusion gene tran-scripts for residual disease detection in leukemia–a EuropeAgainst Cancer program. Leukemia. 2003;17(12):2318.

[14] Oelschlaegel U, Mohr B, Schaich M, et al. HLA-DRneg patientswithout acute promyelocytic leukemia show distinct immuno-phenotypic, genetic, molecular, and cytomorphologic character-istics compared to acute promyelocytic leukemia. Cytometry BCytometry. 2009;76:321–327.

[15] Wetzler M, McElwain B, Stewart C, et al. HLA-DR antigen-negative acute myeloid leukemia. Leukemia. 2003;17(4):707.

[16] Harani MS, Adil S, Shaikh MU, et al. Frequency of fab subtypesin acute myeloid leukemia patients at Aga Khan UniversityHospital Karachi. J Ayub Med Coll Abbottabad. 2005;17(1):26.

[17] Callera F, Mulin CC, Rosa ES, et al. High prevalence ofmorphological subtype FAB M1 in adults with de novo acutemyeloid leukemia in S~ao Jos�e dos Campos, S~ao Paulo. SaoPaulo Med J. 2006;124(1):45–47.

[18] Fenu S, Carmini D, Mancini F, et al. Acute myeloid leukemiasM2, potentially misdiagnosed as M3 variant French-American-Britain (FAB) subtype: a transitional form? Leuk Lymphoma.1995;18(Sup1):49–55.

[19] Xu F, Yin C-X, Wang C-L, et al. Immunophenotypes andimmune markers associated with acute promyelocytic leukemiaprognosis. Dis Markers. 2014;2014:1.

[20] Liu M, Weng X, Gong S, et al. Flow cytometric analysis ofCD64 expression pattern and density in the diagnosis of acutepromyelocytic leukemia: a multi-center study in Shanghai,China. Oncotarget. 2017;8:80625.

[21] Gorczyca W, Sun Z-Y, Cronin W, et al. Immunophenotypicpattern of myeloid populations by flow cytometry analysis.Methods Cell Biol. 2011;103:221–266.

[22] Promsuwicha O, Auewarakul CU. Positive and negative pre-dictive values of HLA-DR and CD34 in the diagnosis of acutepromyelocytic leukemia and other types of acute myeloid leuke-mia with recurrent chromosomal translocations. Asian Pac JAllergy Immunol. 2009;27(4):209.

[23] Ferrari A, Bussaglia E, �Ubeda J, et al. Immunophenotype dis-tinction between acute promyelocytic leukaemia and CD15�CD34� HLA-DR� acute myeloid leukaemia with nucleophos-min mutations. Hematol Oncol. 2012;30(3):109–114.

[24] Orfao A, Chill�on MC, Bortoluci AM, et al. The flow cytometricpattern of CD34, CD15 and CD13 expression in acute myelo-blastic leukemia is highly characteristic of the presence ofPML-RARalpha gene rearrangements. Haematologica. 1999;84(5):405–412.

6 M. MOSLEH ET AL.