Developing a Simple Algorithm Based on Multiparameter Flow Cytometry for Fast Screening of Acute Promyelocytic Leukemia

Introduction: Acute promyelocytic leukemia (APL) is a genetically and molecularly well-defined type of acute leukemia that is curable but has a frequent early mortality due to bleeding. So, there is a need for a fast diagnostic screening in order to start appropriate therapy. Multiparameter flow cytometry (MFC) is usually performed in all types of acute myeloid leukemias (AMLs) but only few features have been described as characteristic of APL.

Aim: to develop a diagnostic algorithm based on the intensity of expression of several antigens examined by MFC in AML that could reliably discriminate between APL and the other types of AML.

Material and Methods: Consecutive newly diagnosed AMLs treated in our Institution during the last 2 years entered the study. Immunophenotyping was included in the diagnostic workup. An 8-color platform based on the Euroflow recommendations was used. The mean fluorescence intensity (MFI) of each antigen tested was assessed and those best discriminating between APL and all other types of AML were obtained by a discriminant analysis. Phenotypic characteristics of normal myeloblasts taken from examinations of bone marrow (BM) MFC performed for the diagnosis of cytopenias were used as controls.

Results: 24 cases of APL and 56 cases of other primary AML entered the study. Median age: 39 (23-56) and 62(26-81) years respectively. Concerning ELN risk groups of non-APL cases, 13 were favorable risk, 26 were intermediate and 09 were adverse risk. In 8 cases risk assessment was not possible due to the absence of cytogenetics. Moreover, among APL patients, 7 cases had a FLT3-ITD mutation. Among non-APL AMLs, 4 had FLT3-ITD mutation, 4 had NPM1 and 10 had FLT3-ITD and NPM1mutation.

Concerning antigen expression, CD34 was expressed in only 1/24 APL samples, and in 18/56 samples from non-APL AMLs. The following flow features were differentially expressed in both groups: SSC (p <0.0001), CD45 (p=0.02), CD13 (p=0.001), CD64 (p=0.004), HLA-DR (p<0.0001) and CD33 (p<0.0001) (Table 1). In the discriminant analysis, MFI CD34 and MFI HLA-DR were able to accurately classify APL and non-APL AML in only 62.5%. However, after the addition of the ratio of SSC between blasts and lymphocytes, these 3 parameters were able to differentiate APL from non-APL AML in 91.2% of the cases.

Conclusion: MFC was adequate for a fast screening of APL in most cases. Expression of CD34 was not very useful, as many AMLs do not express this antigen, similar to APL, but SSC, together with HLA-DR could discriminate both types of leukemia in most cases.