10-Parameter Flow Cytometry as a New Tool to Improve Diagnosis and MRD Follow-Up of Acute Leukemias

During the last 20 years flow cytometric immunophenotyping has become the preferred method for lineage assignment and maturation analyses to specify the subtypes of acute leukemias. Together with moleculargenetics and morphology the comprehensive informations obtained by multiparametric flow cytometry result in a more accurate diagnosis. In addition, visualization of minimal residual disease (MRD) during treatment has increased knowledge about predictive impact of leukemic cell reduction on remission status (Ratei R et al., Leukemia 2008; 1: 1–7). To date, detection of clonality and aberrant antigen expression is based mostly on 2–5 colour flow cytometric analyses and diagnosis is done according to the immunological standardized classification system of acute leukemias (Bene MC et al., Leukemia 1995; 9: 1783–1786 (EGIL); Wood BL et al., Cytometry part B (Clinical cytometry) 2007; 72B: S14–S22). Recently, advances in flow cytometry (Wood B et al., Arch Pathol Lab Med 2006; 130: 680–690; McLaughlin et al., Cytometry A 2008; 73 (5): 400–410 and 411–420) on both, new 8/10 colour, 10/12 parameter flow cytometers (i. e. Navios^™, BeckmanCoulter, Krefeld, Germany; FACS-Canto-II^™, BD, Heidelberg, Germany) and newly patented fluorescent dyes including tandem dye technology may provide a new tool for detection of leukemic cells, useful for accreditation in leukemia diagnosis in the near future.

Here we present our first data with this enhanced 10 parameter flow cytometric analyses based on our previous experience in immunophenotyping as well as aberrant marker expression over more than one decade (Ebener U et al., Klin Pädiatr 2000; 212: 90–98; Huenecke S et al., Eur J Haematol 2008; 80: 532–539) using BeckmanCoulter and BD flow cytometers. Bone marrow and peripheral blood samples (n=991) were obtained from n=187 pediatric patients suffering from Acute Lymphoblastic Leukemia (ALL) or from Acute Myelogenous Leukemia (AML) at initial diagnosis, in case of relapse and at defined time-points as given in BFM-ALL/-AML therapeutic regimen. Flow cytometric analyses were done on both, the dual laser FACSCalibur^™ and/or the three laser FACS-Canto-II^™ equipped with CELLQuest^™ or FACS-DIVA software (BD) as indicated in [Table 1] . A mixture of fluorochrome labelled monoclonal antibodies (MoAbs) reflecting the various hematological cell lines and differentiation stages (Swart et al., JIM 2005; 305: 75–83) were used considering membrane bound, intracytoplasmatic and nuclear antigens. In our overall patient cohort 95% and 93% of our patients show leukemia associated antigens, which allow MRD monitoring of ALL and AML cells, respectively during therapy in accordance to large studies (Campana D., BJH 2008; 142: 481–489). We could demonstrate the following aberrant antigens and co-expressions: CD7, CD10^high, CD13, CD15, CD20, CD33, CD56, CD66c, CD133 for cALL; CD13, CD33, CD133, CD135 for pro B ALL; CD2, cytoplasmatic CD3, CD34^neg for cortical T ALL; CD7, CD19; CD56 for AML, respectively. CD45^neg of leukemic cells provides an additional information in our gating strategy (25% of our patients). Similar results were obtained on the samples of our smaller patient cohort measured on both flow cytometers simultaneously. Thereby, our diagnosis did not differ regarding the percentage of leukemic cells and the expression of asynchronous antigen profile in any case up to now. [Table 2] summarizes our first experience using eight colour combination panels in leukemia (ALL, AML) or MRD diagnosis combining lineage mismatched antigens predominantly and non-lineage committed antigens likewise. In addition we present the simultaneous labelling of leukemic cells and those for antigen expression that might be targets for possible immunotherapies, i. e. Rituximab (anti CD20), Campath-1 (anti CD52); Myelotarg (anti CD33) or tyrosin kinase inhibitors (c-Kit, CD117; e. g. Imatinib/Dasatinib). [Fig. 1] provides an example of such an eight colour flow cytometric analysis with a CD45^dim/side scatter-gating strategy for a standardized flow cytometric immunophenotyping of an AML sample. In addition, [Fig. 2] demonstrates the follow-up of residual leukemic cells from 3 pediatric patients with cALL during chemotherapy (BFM-ALL regimen) and post stem cell transplantation for both, typical co-expressions of aberrant antigens and an example to discriminate leukemic cells from normal B-hematopoiesis.

Fig. 1 10 parameters/8 colour FACS analysis representing dot plots from a patient suffering from AML M5a. These 37 plots emphasize the high level of information provided by this approach using coloured subpopulations. AML (red) demonstrating CD33+/CD64+/partially CD14+(black) and NG2+(MoAb 7.1) as well as co-expression of CD56^high, further lymphocytes (yellow orange), T lymphocytes (blue), CD56+CD3- NK cells (light blue), SSC=side scatter, FSC=forward scatter, PMN=polymorph nucleated cells, MNC=mononucleated cells. Lysing of erythrocytes was done using BD-Lysing solution (1:10). Intracytoplasmatic or nuclear antigen detection was done using BD-Lysing solution in front of incubating MoAbs or using Fix&Perm cell permeabilization procedure. Quality controls were done as recommended by BD company.

Fig. 2 Detection of cALL cells during the follow-up of therapy. A: Leukemic cells with excessive expression of CD10 for successful discrimination to normal B-hematopoiesis; Mean fluorescence intensity (MFI) of leukemic blasts: 936±13 (mean±standard deviation); MFI of pre-B hematopoiesis: 638±97. B: Leukemic cells with CD66c co-expression; MFI of leukemic blasts: 593±32; MFI of residual hematopoiesis: 102±22. C: CD33 co-expression of My+cALL; MFI of leukemic blasts: 458±20; MFI of residual hematopoiesis: 96±17. Pathological blast population is visualized via blue circles and percentages are given in blue numbers. Normal likewise residual hematopoiesis is illustrated via a green circle and percentages are given in green numbers. Negative controls for all figures are presented in the first column.

Table 1 Immunosubtyping acute childhood leukemias by dual and dual versus three laser FACS. Diagnosis dual laser FACS (FACSCalibur™) patients (samples) dual laser FACS (FACSCalibur™) and three laser FACS (FACSCanto II™) patients (samples) Bone marrow and peripheral blood aspirates from pediatric patients (n=187) suffering from ALL (n=152) or AML (n=35) were obtained at initial diagnosis, in case of relapse and at defined time-points as given in BFM-ALL/-AML trial. Diagnosed leukemia subtypes as well as percentages of pathological blasts in the analysed samples using dual and three laser FACS (n=41) were the same in all cases cALL 103 (633) 22 (30) pro B-ALL 7 (70) 2 (2) T-ALL 17 (54) 1 (2) AML 30 (193) 5 (7) total 157 (950) 30 (41)

Table 2 8 colour immunophenotyping panels for advanced characterization of ALL and AML cells based upon the BD FACSCanto-II™. Fluorescence Excitation Blue Laser 488 nm Red Laser 633 nm Violet Laser 405 nm Fluorochromes FITC PE R-PE PerCP PerCP-Cy5.5 PE-Cy5 PE-Cy7 APC APC-Cy7 APC-H7 Pacific Blue Horizon V450 AmCyan Fluorescein isothiocyanate (FITC), Phycoerythrin (PE), Peridinin Chlorophyll Protein (PerCP), PerCP-Cyanin5.5 (PerCP-Cy5.5), PE-Cy7, Allophycocyanin (APC), APC-Cy7, APC-H7, Pacific Blue, Horizon Violet 450, AmCyan served as fluorescent dyes. All fluorochrome labelled MoAbs were obtained from BD, with the exception of CD1a (Beckman Coulter). Each panel also includes cross-lineage markers and represents our first experience in 8 colour FACS, exemplarily. Isotype-matched nonreactive immunoglobulins (BD) served as control. Instrument performance was done automatically in recommendation to BD. Cy=cytoplasmic. Panels are composed to simultaneously visualize at least two to three cell lineages (T, B, and myelo-monocytic) and to realize possible co-expressions of asynchronous antigens in one and the same tube. Thereby for the establishment of new panels it is of major importance to prove CD clusters in different colours panel #1 CD2 CD1a CD7 CD5 CD10 CD19 CD3 CD45 panel #2 (cy) MPO CD10 CD34 CD7 IgM CD33 CD3 CD45 panel #3 (cy) TdT CD79a CD235a CD19 CD22 CD10 CD20 CD45 panel #4 CD19 CD52 CD7 CD33 CD10 CD13 HLA-DR CD45 panel #5 CD19 CD66c CD34 CD10 CD33 CD13 CD15 CD45 panel #6 Kappa Lambda CD19 CD34 IgM CD10 CD20 CD45 panel #7 CD34 CD133 CD117 CD33 CD7 CD10 CD15 CD45 panel #8 CD65 CD61/CD41a CD2 CD34 CD19 CD13 HLA-DR CD45 panel #9 CD8 CD117 CD33 CD4 CD94 CD13 CD56 CD45 panel #10 CD64 7.1 [NG-2] CD19 CD33 CD56 CD4 CD14 CD45

The availability of certified flow cytometers capable for detecting 8, 10 or more colours may spur the discovery of advanced leukemia and MRD diagnosis. While the non disposability of monoclonal antibodies in various colours has limited multiparametric flow cytometry in the past, the availability of essential diagnostic reagents, especially the fluorescent dyes, provide the pole position for the advantage in flow cytometry. For the first time this enables a more informative characterization of a leukemic cell population together with an enhanced detection limit for rare events. Furthermore the new insights, especially the expression of antigens like CD52, CD33, CD117 and others, broaden the possibility to apply success-oriented experimental therapeutic agents to high risk patients. In addition, instrument performances and setups ensure more accurate and precise flow cytometric measurements because of the availability of features like automated set-ups and automated compensation of newly developed fluorochromes. Nevertheless, this high technology requires a high score qualified scientific stuff for time-consuming interpretation and innovative panel design. In summary, these proven multi colour platforms are capable to standardize protocols for further accreditation in clinical laboratories, but certainly need solid investigations and inter-laboratory comparison to show that data are reliable and reproducible.