Flow cytometric analysis of CD64 expression pattern and density in the diagnosis of acute promyelocytic leukemia: a multi-center study in Shanghai, China

Abstract

No unified immunophenotypic profiles and corresponding analytic strategies have been established for the rapid diagnosis of acute promyelocytic leukemia (APL) using flow cytometry (FCM). Here we describe a characteristic immunophenotypic panel that can rapidly and accurately distinguish APL from other types of adult acute myeloid leukemia (AML) using only FCM. By comparing APL cells and non-APL AML cells that share APL common immunophenotypes (CD34−CD117+HLA−DR−) we found that CD64 was a significant factor that differentiated APL from other AMLs. Further retrospective analyses of 205 APL and 629 non-APL AML patients from different hematology centers showed that either the CD64dim and homoCD13+homo CD33+homoMPO+ (myeloperoxidase) CD11c− panel or the CD64dim and homoCD13+homo CD33+homoMPO+ CD11c+CD10−CD117+ SSChigh (high side scatter signal) panel could distinguish APL from non-APL AML patients with nearly 100% sensitivity, specificity and accuracy. Moreover, relative quantification of CD64 expression enhanced the applicability of our APL diagnostic immunophenotypic panels (ADI-panels) in different hematology centers. Application of the ADI-panels will decrease diagnosis time and improve personalized treatment for APL, a life-threatening disease with very rapid progression.

The diagnostic power of CD117, CD13, CD56, CD64, and MPO in rapid screening acute promyelocytic leukemia

“…Importantly, these antigens are signi cant in classifying APL, while cut-off points are optimized rather than a xed value of 20% events (Table 1). These are different from previous studies that used HLA-DR-negative AML as the control group [20,21]. Whereas in comparing the diagnostic performance, we noted that the combination of ve markers, including CD117, CD13, CD56, CD64, and MPO, resulted in excellent accuracy (Figure 2), which are comparable with reports of Liu and Mosleh [20,21].…”

“…These results indicate that the immunophenotype of the non-APL and APL cases are highly closed. In the studies of Liu and Mosleh, although HLA-DR is negative in all control subjects, the expression of other antigens as CD117, CD34, CD11b, CD13, CD33, CD64, and MPO are signi cantly different between APL and non-APL patients [20,21]. So, despite the high diagnostic values presented by Liu and Mosleh, a cut-off value of 20% events applied for all cell antigens might not be useful in APL differential diagnosis, at least from those with APL-like immuno-phenotype as in this study.…”

“…Previous studies have shown the high diagnostic values of cell antigens for APL but with a comparison to an AML control group containing a high percentage of HLA-DR-positive subjects [11][12][13][14][15][16]. Only two studies assessed the diagnostic role of cell antigens for APL in comparison to an HLA-DR-negative AML control group and showed high sensitivity, speci city, and accuracy values (98-100%) [20,21].…”

“…In this study, we used a cut-off point of 20% events to de ne expression status as in previous studies [20,21] but, no signi cant differences of antigen pro le between APL and HLA-DR-negative AML was found (except CD56 and CD64, Additional le 5: Table S2). These results indicate that the immunophenotype of the non-APL and APL cases are highly closed.…”

“…All of the above studies investigated the diagnostic values of ow cytometric antigens that used a control group containing >50% HLA-DR-positive AML patients [11][12][13][14][15][16]. A few reports mentioned the role of these markers in comparison with a similar phenotype control group [20,21]. We compared the antigen expression level between APL and HLA-DR-negative non-APL patients and identi ed the associated markers with APL together with an optimal model in clinical diagnostics.…”