- Email: [email protected]
AC133 expression on acute myeloid leukemia blasts: correlation to FAB and to CD34 expression and possible implications for peripheral blood progenitor cell purging in AML
Leukemia Research 25 (2001) 191– 196 www.elsevier.com/locate/leukres
AC133 expression on acute myeloid leukemia blasts: correlation to FAB and to CD34 expression and possible implications for peripheral blood progenitor cell purging in AML F. Fauth *, E. Weidmann, H. Martin, B. Schneider, S. Sonnhoff, D. Hoelzer Department of Hematology, Klinikum der Johann Wolfgang Goethe Uni6ersita¨t, Medizinische Klinik III-Ha¨matologie/Onkologie, Knochenmarktransplantation, Theodor-Stern-Kai 7, 60590 Frankfurt, Germany Received 2 February 2000; accepted 12 August 2000
Abstract AC133 is an antigen expressed on CD34 + hematopoietic progenitor cells. In acute myeloid leukemia (AML) it is expressed on leukemic blasts of most FAB subtypes. However, few data are available regarding coexpression of other surface antigens. We measured AC133 expression on AML blasts from 28 consecutive patients at initial diagnosis (n= 26) or at diagnosis of first relapse (n=2) and on 26 leukapheresis products from 14 patients. In AML AC133 correlated with CD34 expression (Spearman r= 0.4711, P= 0.0114) and even stronger with combined CD34/CD33 expression (Spearman r= 0.5083, P= 0.0068). In leukapheresis products AC133 expression correlated with CD34 expression (Spearman r= 0.7495, P = 0.002) and the yield of the obtained amount of CD34+ cells (Spearman r= 0.6484, P= 0.0121). In conclusion AC133 expression is closely related to CD34 expression in AML. In leukapheresis products AC133 provides an additional marker for selection of PBPC autografts in AC133AML. © 2001 Elsevier Science Ltd. All rights reserved. Keywords: AC133; Surface antigens; Acute myeloblastic leukemia; Autologous peripheral blood progenitor cell transplantation
1. Introduction AC133 has been first described as an antigen expressed on certain subsets of hematopoietc progenitor cells such as most of CFU-GM and a small proportion of BFU-E [1] in normal bone marrow, early myeloid progenitors such as CD34+/CD38 − cells, CD34+ / CD33 + cells and precursors of dendritic cells [2], but only a small proportion of CD10+ and CD19 + B-cell precursors [3]. It has also been described to be expressed on the AML cell line MUTZ-2 [4]. Abbre6iations: AML, acute myeloid leukemia; aPBPCT, autologous peripheral blood progenitor cell transplantation; BFU-E, burst forming unit — erythroid; CD, cluster differentiation; CFU-GM, colony forming unit — granulocytic monocytic; FAB, French American British classification for acute myeloid leukemias; G-CSF, granulocyte — colony stimulating factor; IL, interleukin; SCF, stem cell factor; PBPCT, peripheral blood progenitor cell transplantation. * Corresponding author. Tel.: +49-69-63017410; fax: + 49-6963016089. E-mail address: [email protected] (F. Fauth).
CD34 + /AC133+ progenitor cells isolated from leukapheresis products of chemotherapy and G-CSF mobilized high-risk breast cancer patients were shown to be expandable ex vivo using SCF, IL-3 and IL-11 to erythroid precursors and CFU-GM by Vavrova et al. [6]. In leukemic blasts AC133 is present not only on myeloid but also on lymphoid blasts and on a subset of B-lymphoid precursor cells and CML cells [5]. Buhring et al. [7] found that 15/18 tested CD34+ and 4/5 CD34 − AML samples expressed AC133. Horn et al. [8] demonstrated coexpression of AC133 and CD34 in 24/30 (80%) AML cases. In a small (n= 6) number of patients Miraglia et al. [9] found three of five CD34+ samples to express AC133. The function of AC133 on a molecular level is not clear. A structure different from other surface antigens on hematopoietic cells and homology with prominin, a protein located at the apical surface of murine epithelium have been demonstrated [9,10].
0145-2126/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 5 - 2 1 2 6 ( 0 0 ) 0 0 1 1 2 - 0
192
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
To investigate the potential role of AC133 as a diagnostic marker in AML and as a marker for selection of peripheral blood stem cells in autologous peripheral blood progenitor cell transplantation (aPBPCT) in AML we measured the expression of AC133, CD34 and other surface antigens on human leukemic blast cells in consecutive 28 patients with de novo AML at initial diagnosis and on leukapheresis products (n =26) from 14 patients (13 AML, one secondary AML).
2. Patients and methods
2.1. Patients Twenty-eight consecutive patients with AML were evaluated. All 28 patients were diagnosed to have AML de novo, two of them in first relapse. Median age was 55 (18 –76) years, gender distribution was 14 females and 14 males. At the time of initial diagnosis leukemic cells were obtained from bone marrow aspiration, or peripheral blood when marrow was not aspirable. All patients were studied prior to treatment. Informed consent was obtained for use of aliquots of leukapheresis products. Other analyses were performed from blood or bone marrow samples taken for routine procedures.
2.2. Measurement of surface antigen expression Mononuclear cells were stained by double color direct immunofluorescence phycoerythrin (PE) and fluorescein-isothiocyanate (FITC)-labeled with a series of monoclonal antibodies (MoAb) and studied immediately. The following combinations were used (FITC/ PE): CD3/HLA DR; CD10/CD19; CD41/CD13; CD14/CD11c; CD34/CD33; CD15/CD11b; CD34/ AC133; POX. A gate was set primarily on the subpopulation of malignant cells by CD34 positivity and side scatter. In a second step the percentage of positive cells in this gate was measured for the tested antigens. Positivity was defined by using isotype controls, any intensity equal or lower than in isotype control was regarded as negative. Analysis of phenotypical characteristics was performed using a FACScalibur flow cytometer and the Cellquest software (Becton Dickinson, Heidelberg, Germany). The AC133 (PE) MoAb was purchased from Miltenyi Biotec, Bergisch Gladbach, Germany, the POX (FITC) antibody was purchased from DAKO Inc. Denmark, all others were purchased from Becton Dickinson, Heidelberg, Germany. These were CD3, CD10, CD14, CD15, CD34 and CD41a (FITC) and HLA-DR, CD11c, CD13, CD19, and CD33 (PE).
2.3. Classification FAB type was determined by morphological criteria. Additionally cytogenetics and immunological analysis were evaluated. However, in case of a cytogenetic aberration clearly associated with a certain FAB type (e.g. inv (16)(p13q22) with FAB M4Eo or t(8;21) with FAB M2, the patient was scored according to karyotype. If there was a difference between morphology and immunology, the patient was scored after morphological criteria. Immunological phenotype was not respected for FAB classification.
2.4. Leukaphereses Twenty-six leukaphereses from 14 patients with AML (one secondary AML, 13 de novo AML) were performed at the local blood bank. Patients were mobilized with G-CSF which was started with rising WBC and continued until the day of the last leukapheresis in a dose of 5–10 mg/kg BW. All patients were at complete morphological remission at this time after having received chemotherapy according to ongoing protocols. No blast cells were detected in the leukapheresis samples. We measured 26 leukaphereses from 14 patients. Median mononuclear cell yield was 1.76× 1010. Median CD34 + cell yield was 1.24× 108 per leukapheresis and 2.80×108 per patient. Immunophenotyping was done by double staining cells with the following monoclonal antibodies (FITC/ PE): CD34/CD19, CD45/CD34 and CD34/AC133. The first two panels are used on a routine basis in leukapheresis products in our department. Positivity was defined by using isotype controls as described above. Early progenitor cells were gated by forward and sideward scatter and the antigens were measured by staining with the moAbs and studied immediately. Correlation of AC133 to CD34 expression was calculated on the basis of medium values of all leukaphereses of each patient.
2.5. Statistical analysis All shown data are median values except those stated differently. For statistical analysis, the Mann –Whitney test for non-paired observations was used. Correlations were calculated with the Spearman non-parametric correlation test.
3. Results The FAB distribution in the investigated patients was M0 =1, M1 = 4, M2 = 5, M3 = 0, M4 = 9, M4Eo =4, M5 =3, M6 = 2, M7 =0. Most patients had a normal
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
193
Table 1 AML blasts: patient’s gender, age and expression of CD34 and AC133 No.
Gender (m/f)
Age (years)
FAB
CD34 expression (%)
AC133 expression (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
f f f m f m m f m f m m m f m f m m m f f f m f m f m f
73 19 70 75 58 68 36 68 60 54 40 51 38 58 56 24 44 43 25 69 70 50 65 49 51 73 77 40
M4 M4Eo M5 M2 M1 M4 M4Eo M4 M4 M0 M4 M4Eo M4 M1 M6 M2 M2 M2 M2 M4 M4 M4Eo M4 M1 M6 M5 M5 M1
10.91 29.76 0.15 65.66 95.71 0.20 48.85 0.41 0.15 95.53 65.50 12.70 15.40 54.60 74.28 85.08 95.27 67.86 38.66 3.14 0.55 91.98 0.18 91.82 19.24 67.78 21.97 0.18
11.18 25.42 0.68 50.46 95.92 0.83 50.88 0.55 0.12 0.57 78.50 3.76 0.28 54.92 67.92 26.16 25.53 54.64 53.90 0.89 0.42 0.06 89.88 55.69 0.47 30.23 5.15 0.12
karyotype on cytogenetic evaluation. Four patients had an inv (16); of these, two presented with an additional marker chromosome. Five patients presented with deletions or complex abnormalities. Five patients had less than 5000 WBC/ml at initial diagnosis. For more details see Table 1.
3.2. Correlation of CD34 expression to FAB morphology Expression of CD34 expression in AML M4/M5 patients (n= 12) was significantly lower than in AML M1/M2 patients (n=9, P= 0.0056). Comparing AML
3.1. Correlation of AC133 expression to other surface antigens in AML patients There was a significant correlation between the expression of AC133 and CD34 in the AML group (n= 28; Spearman r = 0.4515; P = 0.0159). This correlation was stronger when AC133 expression was correlated to combined expression of CD34 and CD33 (n=27; Spearman r =0.4940; P =0.0088). The correlation is shown in Fig. 1. Regarding expression of AC133 and expression of POX (n = 27, Spearman r = 0.3383, P=0.0844) a trend towards significance could be observed. In the CD34+ subgroup no correlation between AC133 and a single antigen was found. AC133 expression did not correlate with any other of the evaluated surface antigens (n =27).
Fig. 1. This graph shows the correlation of AC133 and CD34/CD33 expression on AML blast cells.
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
194
Table 2 This table shows the gender, age and FAB subtype of the patients that underwent leukapheresis, the number of leukaphereses, the CD34 yield per leukapheresis, the total amount of CD34+ cells collected and the percentages of CD34 and AC133 expression of mononuclear cells in the leukapheresis products Patients
Leukapheresis products
No.
Gender
Age
FAB
No. of LA
CD34+ MNC (/LA)
CD34 yield (/LA)
CD34 (%)
AC133 (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
m f m f m m m m m m m f m f
52 54 36 58 44 38 47 56 56 60 40 54 63 44
M2 M0 M4Eo M1 M2 M4 M3 M5 M6 M4 M4 M4 M2 M4
2 2 2 2 1 2 1 1 2 2 2 3 2 2
8.20×107 1.22×108 1.16×108 6.98×108 3.14×108 5.36×108 6.77×108 4.42×108 1.06×108 9.81×107 1.29×108 1.01×108 2.10×108 7.51×107
1.64×108 2.44×108 2.31×108 1.40×109 3.14×108 1.07×109 6.77×108 4.42×108 2.11×108 1.96×108 2.57×108 3.04×108 4.21×108 1.50×108
1.04 3.29 2.00 8.42 4.07 7.77 15.81 3.14 3.05 1.24 1.85 1.62 5.38 0.96
0.73 3.00 2.20 6.76 0.58 6.99 8.69 2.57 3.55 0.80 1.09 1.54 3.71 1.12
M4/M5 patients with AML M0 + M1 + M2 + M6 patients (n= 12), this difference was even more significant (P = 0.0018). In AML M4Eo median CD34 expression was 39.31 (12.7 –91.82)% (n = 4) differing significantly from AML M4 patients (n =9, P =0.0336).
3.3. Correlation of AC133 expression to FAB morphology A lower AC133 expression in the AML FAB M4/M5 patients (n=12) than in the patients with AML FAB M1/M2 could not be demonstrated in this study (n= 9, P= 0.255). Median AC133 expression in AML M4Eo patients was 38.15 (3.76 – 55.69)% (n =4). This was not significantly different from AML M4 patients (n= 9, median 0.83 (0.12 –89.88)%, P =0.1986)
3.4. AC133 expression in leukapheresis products In the leukapheresis products, there was a positive correlation of AC133 expression to CD34 expression (Spearman r= 0.7495, P =0.002), to the amount of CD34 + cells in the leukapheresis (Spearman r= 0.6352, P = 0.0147) and to the CD34 yield as a whole (Spearman r =0.6484, P =0.0121) (26 aphereses from 14 patients). CD19 did not correlate with AC133 expression. The general data of the leukapheresis products are summarized in Table 2.
3.5. Correlation of AC133 expression to initial WBC, cytogenetics and age No correlation was found between AC133 expression and initial WBC, cytogenetics or age (data not shown).
4. Discussion AC133 is an antigen expressed on leukemic blasts as well as on normal hematotpoietic progenitor cells. However, its value as a marker and its function in leukemic cells is still unclear. By analyzing data from AML samples and leukapheresis products we investigated, wether it was possible to define by AC133 expression a cell population whose presence correlates with other important risk factors in AML such as FAB morphology, age, karyotype or initial WBC. Our second aim was to evaluate its role as a possible selection alternative for collection of peripheral blood rogenitor cells in autologous peripheral blood progenitor cell transplantation (aPBPCT) in AML. Expression of AC133 and CD34 correlated well in our study, stronger correlation was achieved comparing AC133 with combined CD34/CD33 expression, whereas CD33 expression alone did not correlate with AC133 expression. These data are in line with findings of Snell et al. [11] who found a correlation between AC133 and CD34 expression in normal progenitor cells indicating that this feature is not distinctive between normal and leukemic progenitor cells. In contrast to the data of Buhring et al. [7] we found that in 6/7 CD34 − samples (expression B 1% of cells) AC133 expression was also below 1%. Of these six samples five showed M4/M5 morphology, one showed M1 morphology. These results show that AC133 is expressed on CD34 + cells when they are already CD33+ but disappears when these cells become CD34−. In this study no AC133− subpopulation of CD34+ cells in AML blasts could be defined.
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
We conclude that AC133 expression defines a short time period in differentiation of AML blasts and that AC133 is rather a surrogate marker for CD34+ / CD33 + cells but a marker to define different populations of blast cells in AML. Regarding correlation of AC133 to FAB morphology our results could not show a trend towards different expression of AC133 in AML M4/M5 and AML M1/ M2. These data contrast with data from Miraglia et al. [9] who found bright expression of AC133 only in AML FAB M4/M5 but are in line with the findings of Horn et al. [8] and Buhring et al. [7], who observed AC133 expression in all FAB subtypes in 30 patients. A larger series of patients might be required to settle this topic. However, since we observed a significantly lower expression of CD34 in AML M4/M5 compared to other morphologies, which is well in line with previous publications [12 –14], any difference in AC133 expression could be explained by the correlation of AC133 and CD34 expression. This means that AC133 is but a surrogate marker for CD34 expression in terms of immunological characterisation of FAB subtypes in AML. Our second aim was to evaluate whether AC133 may be a potential tool for purging AML autografts for autologous peripheral blood progenitor cell transplantation (aPBPCT). In several studies the effect of purging autografts with cytotoxic agents, iron-bound monoclonal antibodies (beads) or physical methods have been investigated in vivo and in vitro [15 – 20]. For bone marrow purging with mafosfamide a beneficial effect on survival has been shown in several studies [15,21] as could be demostrated for bone marrow purging with monoclonal antibodies [22]. Because CD34+/AC133 + cells have repopulating capacity of in cell culture [6] and in animal models [1,2] whereas CD34+/AC133 − cells have not [2], these cells should be a target cell population in leukapheresis for aPBPCT. The correlation between AC133 and CD34 expression found in the evaluated leukaphereses confirms the findings of the literature. AC133 could therefore be used for positive selection of early hematopoietic progenitor cells as an alternative of CD34 or for elimination of leukemic progenitors by negative selection in AC133− AML. From the results of this study we conclude that AC133 expression on CD34+ myeloid blasts in AML is well correlated to CD33 expression of the respective cells but is rather a surrogate marker for the combined expression of CD34 and CD33 than it is able to define a specific subpopulation of AML blasts. AC133 expression on G-CSF-mobilized peripheral blood progenitors in AML patients in CR correlates with CD34 expression, indicating its usefulness as an alternative to CD34 for selecting peripheral blood progenitors for autografts or purge them from leukemic progenitors by negative selection in AC133− AML.
195
References [1] Yin AH, Miraglia S, Zanjani ED, Almeida-Porada G, Ogawa M, Leary AGXOJ, Kearney J, Buck DW (1997) AC133, a novel marker for human hematopoietic stem and progenitor cells. Blood 90 (12), 5002-5012. [2] de Wynter EA, Buck D, Hart C, Heywood R, Coutinho LH, Clayton A, Rafferty JA, Burt D, Guenechea G, Bueren JA, Gagen D, Fairbairn LJ, Lord BI, Testa NG (1998) CD34 + AC133 + cells isolated from cord blood are highly enriched in long- term culture-initiating cells, NOD/SCID-repopulating cells and dendritic cell progenitors. Stem Cells 16(6), 387-396. [3] Buhring HJ, Seiffert M, Bock TA, Scheding S, Thiel A, Scheffold A, Kanz L, Brugger W. (1999) Expression of novel surface antigens on early hematopoietic cells. Ann N Y Acad Sci 872, 25-38. [4] Kratz-Albers K, Zuhlsdorp M, Leo R, Berdel WL, Buchner T, Serve H (1998) Expression of a AC133, a novel stem cell marker, on human leukemic blasts lacking CD34-antigen and on a human CD34+ leukemic line:MUTZ-2 [letter]. Blood 92(11), 4485-4487. [5] Waller CF, Martens UM, Lange W. Philadelphia chromosomepositive cells are equally distributed in AC133 + and AC133fractions of CD34 + peripheral blood progenitor cells from patients with CML (1999) [letter] Leukemia 1999; 13(9):14661467. [6] Vavrova J, Filip S, Vokurkova D, Blaha M, Vanasek J, Jebavy L (1999). Ex vivo expansion CD34 +/AC133 +-selected autologous peripheral blood progenitor cells (PBPC) in high-risk breast cancer patients receiving intensive chemotherapy [In Process Citation]. Hematol Cell Ther 41(3), 105-112. [7] Buhring HJ, Seiffert M, Marxer A, Weiss B, Faul C, Kanz L, Brugger W (1999) AC133 antigen expression is not restricted to acute myeloid leukemia blasts but is also found on acute lymphoid leukemia blasts and on a subset of CD34 + B-cell precursors [letter]. Blood 94 (2), 832-833. [8] Horn PA, Tesch H, Staib P, Kube D, Diehl V, Voliotis D (1999) Expression of AC133, a novel hematopoietic precursor antigen, on acute myeloid leukemia cells [letter]. Blood 93 (4), 1435-1437. [9] Miraglia S, Godfrey W, Yin AH, Atkins K, Warnke R, Holden JT, Bray RAX, Waller EK, Buck DW. (1997) A novel fivetransmembrane hematopoietic stem cell antigen: isolation, characterization, and molecular cloning. Blood 90 (12), 5013-5021. [10] Corbeil D, Roper K, Weigmann A, Huttner WB. (1998) AC133 hematopoietic stem cell antigen: human homologue of mouse kidney prominin or distinct member of a novel protein family? [letter]. Blood 91 (7), 2625-2626. [11] Snell V, Jackson E, Buck D, Andreeff M. (1998) Expression of the AC133 antigen in leukemic andnormal progenitors. Blood 92 119a (abstr, suppl 1). [12] Batinic D, Tindle R, Boban D, Tiefenbach A, Rajic L, Labar B, Nemet D, Boranic M (1989) Expression of hematopoietic progenitor cell-associated antigen BI-3C5/CD34 in leukemia Leukemia Research 13 (1), 83-85 [13] Harada N, Okamura S, Kubota A, Shimoda K, Ikematsu W, Kondo S, Harada M, Niho Y (1994) Analysis of acute myeloid leukemia cells by flow cytometry, introducing a new light-scattering classification Cancer Research in Clinical Oncology 120 (9), 553-557 [14] Sperling C, Buchner T, Creutzig U, Ritter J, Harbott J, Fonatsch C, Sauerland C, Mielcarek M, Maschmeyer G, Loffler H, et al. (1995) Clinical, morphologi, cytogenetic and prognostic implications of CD34 expression in childhood and adult de novo AML Leukemia an Lymphoma 17 (5-6), 417-426 [15] Gorin NC, Labopin M, Meloni G, Korbling M, Carella A, Herve P, Burnett, A, Rizzoli V, Alessandrino EP, Bjorkstrand B,
196
[16]
[17]
[18]
[19]
F. Fauth et al. / Leukemia Research 25 (2001) 191–196 et al. (1991) Autologous bone marrow transplantation for acute myeloblastic leukemia in Europe: further evidence of the role of marrow purging by mafosfamide. European Co-operative Group for Bone Marrow Transplantation (EBMT). Leukemia 5 (10), 896-904. Gorin NC, Aegerter P, Auvert B, Meloni G, Goldstone AH, Burnett AXCA, Korbling M, Herve P, Maraninchi D, et al. (1990) Autologous bone marrow transplantation for acute myelocytic leukemia in first remission: a European survey of the role of marrow purging. Blood 75 (8), 1606-1614. Rubin J, Malley V, Ball ED (1994) A combination of anti-CD15 monoclonal antibody PM-81 and 4-hydroperoxy-cyclophosphamide augments tumor cytotoxicity while sparing normal progenitor cells. Journal of Hematotherapy; 3 (2), 121-127. Gidali J, Feher I, Kovacs P (1994) Comparative heat sensitivity of murine and human hemopoietic progenitors and clonogenic leukemia cells. Stem Cells 12 (5), 533-538. Lemoli RM, Igarashi T, Knizewski M, Acaba L, Richter A, Jain AXMD, Levy J, Gulati SC (1993) Dye-mediated photolysis is
.
capable of eliminating drug-resistant (MDR + ) tumor cells. Blood 81 (3), 793-800. [20] Nimgaonkar M, Kemp A, Lancia J, Ball ED (1996) A combination of CD34 selection and complement-mediated immunopurging (anti-CD15 monoclonal antibody) eliminates tumor cells while sparing normal progenitor cells. Journal of Hematotherapy 5 (1), 39-48. [21] Laporte JP, Douay L, Lopez M, Labopin M, Jouet JP, Lesage S, Stachowiak, J, Fouillard L, Isnard F, Noel-Walter MP, et al. (1994) One hundred twenty-five adult patients with primary acute leukemia autografted with marrow purged by mafosfamide: a 10-year single institution experience. Blood 84 (11), 3810-3818. [22] Selvaggi KJ, Wilson JW, Mills LE, Cornwell GG3, Hurd D, Dodge WXGR, Martin SE, McMillan R, Miller W, et al. (1994) Improved outcome for high-risk acute myeloid leukemia patients using autologous bone marrow transplantation and monoclonal antibody-purged bone marrow. Blood 83 (6), 1698-1705.
AC133 expression on acute myeloid leukemia blasts: correlation to FAB and to CD34 expression and possible implications for peripheral blood progenitor cell purging in AML F. Fauth *, E. Weidmann, H. Martin, B. Schneider, S. Sonnhoff, D. Hoelzer Department of Hematology, Klinikum der Johann Wolfgang Goethe Uni6ersita¨t, Medizinische Klinik III-Ha¨matologie/Onkologie, Knochenmarktransplantation, Theodor-Stern-Kai 7, 60590 Frankfurt, Germany Received 2 February 2000; accepted 12 August 2000
Abstract AC133 is an antigen expressed on CD34 + hematopoietic progenitor cells. In acute myeloid leukemia (AML) it is expressed on leukemic blasts of most FAB subtypes. However, few data are available regarding coexpression of other surface antigens. We measured AC133 expression on AML blasts from 28 consecutive patients at initial diagnosis (n= 26) or at diagnosis of first relapse (n=2) and on 26 leukapheresis products from 14 patients. In AML AC133 correlated with CD34 expression (Spearman r= 0.4711, P= 0.0114) and even stronger with combined CD34/CD33 expression (Spearman r= 0.5083, P= 0.0068). In leukapheresis products AC133 expression correlated with CD34 expression (Spearman r= 0.7495, P = 0.002) and the yield of the obtained amount of CD34+ cells (Spearman r= 0.6484, P= 0.0121). In conclusion AC133 expression is closely related to CD34 expression in AML. In leukapheresis products AC133 provides an additional marker for selection of PBPC autografts in AC133AML. © 2001 Elsevier Science Ltd. All rights reserved. Keywords: AC133; Surface antigens; Acute myeloblastic leukemia; Autologous peripheral blood progenitor cell transplantation
1. Introduction AC133 has been first described as an antigen expressed on certain subsets of hematopoietc progenitor cells such as most of CFU-GM and a small proportion of BFU-E [1] in normal bone marrow, early myeloid progenitors such as CD34+/CD38 − cells, CD34+ / CD33 + cells and precursors of dendritic cells [2], but only a small proportion of CD10+ and CD19 + B-cell precursors [3]. It has also been described to be expressed on the AML cell line MUTZ-2 [4]. Abbre6iations: AML, acute myeloid leukemia; aPBPCT, autologous peripheral blood progenitor cell transplantation; BFU-E, burst forming unit — erythroid; CD, cluster differentiation; CFU-GM, colony forming unit — granulocytic monocytic; FAB, French American British classification for acute myeloid leukemias; G-CSF, granulocyte — colony stimulating factor; IL, interleukin; SCF, stem cell factor; PBPCT, peripheral blood progenitor cell transplantation. * Corresponding author. Tel.: +49-69-63017410; fax: + 49-6963016089. E-mail address: [email protected] (F. Fauth).
CD34 + /AC133+ progenitor cells isolated from leukapheresis products of chemotherapy and G-CSF mobilized high-risk breast cancer patients were shown to be expandable ex vivo using SCF, IL-3 and IL-11 to erythroid precursors and CFU-GM by Vavrova et al. [6]. In leukemic blasts AC133 is present not only on myeloid but also on lymphoid blasts and on a subset of B-lymphoid precursor cells and CML cells [5]. Buhring et al. [7] found that 15/18 tested CD34+ and 4/5 CD34 − AML samples expressed AC133. Horn et al. [8] demonstrated coexpression of AC133 and CD34 in 24/30 (80%) AML cases. In a small (n= 6) number of patients Miraglia et al. [9] found three of five CD34+ samples to express AC133. The function of AC133 on a molecular level is not clear. A structure different from other surface antigens on hematopoietic cells and homology with prominin, a protein located at the apical surface of murine epithelium have been demonstrated [9,10].
0145-2126/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 5 - 2 1 2 6 ( 0 0 ) 0 0 1 1 2 - 0
192
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
To investigate the potential role of AC133 as a diagnostic marker in AML and as a marker for selection of peripheral blood stem cells in autologous peripheral blood progenitor cell transplantation (aPBPCT) in AML we measured the expression of AC133, CD34 and other surface antigens on human leukemic blast cells in consecutive 28 patients with de novo AML at initial diagnosis and on leukapheresis products (n =26) from 14 patients (13 AML, one secondary AML).
2. Patients and methods
2.1. Patients Twenty-eight consecutive patients with AML were evaluated. All 28 patients were diagnosed to have AML de novo, two of them in first relapse. Median age was 55 (18 –76) years, gender distribution was 14 females and 14 males. At the time of initial diagnosis leukemic cells were obtained from bone marrow aspiration, or peripheral blood when marrow was not aspirable. All patients were studied prior to treatment. Informed consent was obtained for use of aliquots of leukapheresis products. Other analyses were performed from blood or bone marrow samples taken for routine procedures.
2.2. Measurement of surface antigen expression Mononuclear cells were stained by double color direct immunofluorescence phycoerythrin (PE) and fluorescein-isothiocyanate (FITC)-labeled with a series of monoclonal antibodies (MoAb) and studied immediately. The following combinations were used (FITC/ PE): CD3/HLA DR; CD10/CD19; CD41/CD13; CD14/CD11c; CD34/CD33; CD15/CD11b; CD34/ AC133; POX. A gate was set primarily on the subpopulation of malignant cells by CD34 positivity and side scatter. In a second step the percentage of positive cells in this gate was measured for the tested antigens. Positivity was defined by using isotype controls, any intensity equal or lower than in isotype control was regarded as negative. Analysis of phenotypical characteristics was performed using a FACScalibur flow cytometer and the Cellquest software (Becton Dickinson, Heidelberg, Germany). The AC133 (PE) MoAb was purchased from Miltenyi Biotec, Bergisch Gladbach, Germany, the POX (FITC) antibody was purchased from DAKO Inc. Denmark, all others were purchased from Becton Dickinson, Heidelberg, Germany. These were CD3, CD10, CD14, CD15, CD34 and CD41a (FITC) and HLA-DR, CD11c, CD13, CD19, and CD33 (PE).
2.3. Classification FAB type was determined by morphological criteria. Additionally cytogenetics and immunological analysis were evaluated. However, in case of a cytogenetic aberration clearly associated with a certain FAB type (e.g. inv (16)(p13q22) with FAB M4Eo or t(8;21) with FAB M2, the patient was scored according to karyotype. If there was a difference between morphology and immunology, the patient was scored after morphological criteria. Immunological phenotype was not respected for FAB classification.
2.4. Leukaphereses Twenty-six leukaphereses from 14 patients with AML (one secondary AML, 13 de novo AML) were performed at the local blood bank. Patients were mobilized with G-CSF which was started with rising WBC and continued until the day of the last leukapheresis in a dose of 5–10 mg/kg BW. All patients were at complete morphological remission at this time after having received chemotherapy according to ongoing protocols. No blast cells were detected in the leukapheresis samples. We measured 26 leukaphereses from 14 patients. Median mononuclear cell yield was 1.76× 1010. Median CD34 + cell yield was 1.24× 108 per leukapheresis and 2.80×108 per patient. Immunophenotyping was done by double staining cells with the following monoclonal antibodies (FITC/ PE): CD34/CD19, CD45/CD34 and CD34/AC133. The first two panels are used on a routine basis in leukapheresis products in our department. Positivity was defined by using isotype controls as described above. Early progenitor cells were gated by forward and sideward scatter and the antigens were measured by staining with the moAbs and studied immediately. Correlation of AC133 to CD34 expression was calculated on the basis of medium values of all leukaphereses of each patient.
2.5. Statistical analysis All shown data are median values except those stated differently. For statistical analysis, the Mann –Whitney test for non-paired observations was used. Correlations were calculated with the Spearman non-parametric correlation test.
3. Results The FAB distribution in the investigated patients was M0 =1, M1 = 4, M2 = 5, M3 = 0, M4 = 9, M4Eo =4, M5 =3, M6 = 2, M7 =0. Most patients had a normal
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
193
Table 1 AML blasts: patient’s gender, age and expression of CD34 and AC133 No.
Gender (m/f)
Age (years)
FAB
CD34 expression (%)
AC133 expression (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
f f f m f m m f m f m m m f m f m m m f f f m f m f m f
73 19 70 75 58 68 36 68 60 54 40 51 38 58 56 24 44 43 25 69 70 50 65 49 51 73 77 40
M4 M4Eo M5 M2 M1 M4 M4Eo M4 M4 M0 M4 M4Eo M4 M1 M6 M2 M2 M2 M2 M4 M4 M4Eo M4 M1 M6 M5 M5 M1
10.91 29.76 0.15 65.66 95.71 0.20 48.85 0.41 0.15 95.53 65.50 12.70 15.40 54.60 74.28 85.08 95.27 67.86 38.66 3.14 0.55 91.98 0.18 91.82 19.24 67.78 21.97 0.18
11.18 25.42 0.68 50.46 95.92 0.83 50.88 0.55 0.12 0.57 78.50 3.76 0.28 54.92 67.92 26.16 25.53 54.64 53.90 0.89 0.42 0.06 89.88 55.69 0.47 30.23 5.15 0.12
karyotype on cytogenetic evaluation. Four patients had an inv (16); of these, two presented with an additional marker chromosome. Five patients presented with deletions or complex abnormalities. Five patients had less than 5000 WBC/ml at initial diagnosis. For more details see Table 1.
3.2. Correlation of CD34 expression to FAB morphology Expression of CD34 expression in AML M4/M5 patients (n= 12) was significantly lower than in AML M1/M2 patients (n=9, P= 0.0056). Comparing AML
3.1. Correlation of AC133 expression to other surface antigens in AML patients There was a significant correlation between the expression of AC133 and CD34 in the AML group (n= 28; Spearman r = 0.4515; P = 0.0159). This correlation was stronger when AC133 expression was correlated to combined expression of CD34 and CD33 (n=27; Spearman r =0.4940; P =0.0088). The correlation is shown in Fig. 1. Regarding expression of AC133 and expression of POX (n = 27, Spearman r = 0.3383, P=0.0844) a trend towards significance could be observed. In the CD34+ subgroup no correlation between AC133 and a single antigen was found. AC133 expression did not correlate with any other of the evaluated surface antigens (n =27).
Fig. 1. This graph shows the correlation of AC133 and CD34/CD33 expression on AML blast cells.
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
194
Table 2 This table shows the gender, age and FAB subtype of the patients that underwent leukapheresis, the number of leukaphereses, the CD34 yield per leukapheresis, the total amount of CD34+ cells collected and the percentages of CD34 and AC133 expression of mononuclear cells in the leukapheresis products Patients
Leukapheresis products
No.
Gender
Age
FAB
No. of LA
CD34+ MNC (/LA)
CD34 yield (/LA)
CD34 (%)
AC133 (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
m f m f m m m m m m m f m f
52 54 36 58 44 38 47 56 56 60 40 54 63 44
M2 M0 M4Eo M1 M2 M4 M3 M5 M6 M4 M4 M4 M2 M4
2 2 2 2 1 2 1 1 2 2 2 3 2 2
8.20×107 1.22×108 1.16×108 6.98×108 3.14×108 5.36×108 6.77×108 4.42×108 1.06×108 9.81×107 1.29×108 1.01×108 2.10×108 7.51×107
1.64×108 2.44×108 2.31×108 1.40×109 3.14×108 1.07×109 6.77×108 4.42×108 2.11×108 1.96×108 2.57×108 3.04×108 4.21×108 1.50×108
1.04 3.29 2.00 8.42 4.07 7.77 15.81 3.14 3.05 1.24 1.85 1.62 5.38 0.96
0.73 3.00 2.20 6.76 0.58 6.99 8.69 2.57 3.55 0.80 1.09 1.54 3.71 1.12
M4/M5 patients with AML M0 + M1 + M2 + M6 patients (n= 12), this difference was even more significant (P = 0.0018). In AML M4Eo median CD34 expression was 39.31 (12.7 –91.82)% (n = 4) differing significantly from AML M4 patients (n =9, P =0.0336).
3.3. Correlation of AC133 expression to FAB morphology A lower AC133 expression in the AML FAB M4/M5 patients (n=12) than in the patients with AML FAB M1/M2 could not be demonstrated in this study (n= 9, P= 0.255). Median AC133 expression in AML M4Eo patients was 38.15 (3.76 – 55.69)% (n =4). This was not significantly different from AML M4 patients (n= 9, median 0.83 (0.12 –89.88)%, P =0.1986)
3.4. AC133 expression in leukapheresis products In the leukapheresis products, there was a positive correlation of AC133 expression to CD34 expression (Spearman r= 0.7495, P =0.002), to the amount of CD34 + cells in the leukapheresis (Spearman r= 0.6352, P = 0.0147) and to the CD34 yield as a whole (Spearman r =0.6484, P =0.0121) (26 aphereses from 14 patients). CD19 did not correlate with AC133 expression. The general data of the leukapheresis products are summarized in Table 2.
3.5. Correlation of AC133 expression to initial WBC, cytogenetics and age No correlation was found between AC133 expression and initial WBC, cytogenetics or age (data not shown).
4. Discussion AC133 is an antigen expressed on leukemic blasts as well as on normal hematotpoietic progenitor cells. However, its value as a marker and its function in leukemic cells is still unclear. By analyzing data from AML samples and leukapheresis products we investigated, wether it was possible to define by AC133 expression a cell population whose presence correlates with other important risk factors in AML such as FAB morphology, age, karyotype or initial WBC. Our second aim was to evaluate its role as a possible selection alternative for collection of peripheral blood rogenitor cells in autologous peripheral blood progenitor cell transplantation (aPBPCT) in AML. Expression of AC133 and CD34 correlated well in our study, stronger correlation was achieved comparing AC133 with combined CD34/CD33 expression, whereas CD33 expression alone did not correlate with AC133 expression. These data are in line with findings of Snell et al. [11] who found a correlation between AC133 and CD34 expression in normal progenitor cells indicating that this feature is not distinctive between normal and leukemic progenitor cells. In contrast to the data of Buhring et al. [7] we found that in 6/7 CD34 − samples (expression B 1% of cells) AC133 expression was also below 1%. Of these six samples five showed M4/M5 morphology, one showed M1 morphology. These results show that AC133 is expressed on CD34 + cells when they are already CD33+ but disappears when these cells become CD34−. In this study no AC133− subpopulation of CD34+ cells in AML blasts could be defined.
F. Fauth et al. / Leukemia Research 25 (2001) 191–196
We conclude that AC133 expression defines a short time period in differentiation of AML blasts and that AC133 is rather a surrogate marker for CD34+ / CD33 + cells but a marker to define different populations of blast cells in AML. Regarding correlation of AC133 to FAB morphology our results could not show a trend towards different expression of AC133 in AML M4/M5 and AML M1/ M2. These data contrast with data from Miraglia et al. [9] who found bright expression of AC133 only in AML FAB M4/M5 but are in line with the findings of Horn et al. [8] and Buhring et al. [7], who observed AC133 expression in all FAB subtypes in 30 patients. A larger series of patients might be required to settle this topic. However, since we observed a significantly lower expression of CD34 in AML M4/M5 compared to other morphologies, which is well in line with previous publications [12 –14], any difference in AC133 expression could be explained by the correlation of AC133 and CD34 expression. This means that AC133 is but a surrogate marker for CD34 expression in terms of immunological characterisation of FAB subtypes in AML. Our second aim was to evaluate whether AC133 may be a potential tool for purging AML autografts for autologous peripheral blood progenitor cell transplantation (aPBPCT). In several studies the effect of purging autografts with cytotoxic agents, iron-bound monoclonal antibodies (beads) or physical methods have been investigated in vivo and in vitro [15 – 20]. For bone marrow purging with mafosfamide a beneficial effect on survival has been shown in several studies [15,21] as could be demostrated for bone marrow purging with monoclonal antibodies [22]. Because CD34+/AC133 + cells have repopulating capacity of in cell culture [6] and in animal models [1,2] whereas CD34+/AC133 − cells have not [2], these cells should be a target cell population in leukapheresis for aPBPCT. The correlation between AC133 and CD34 expression found in the evaluated leukaphereses confirms the findings of the literature. AC133 could therefore be used for positive selection of early hematopoietic progenitor cells as an alternative of CD34 or for elimination of leukemic progenitors by negative selection in AC133− AML. From the results of this study we conclude that AC133 expression on CD34+ myeloid blasts in AML is well correlated to CD33 expression of the respective cells but is rather a surrogate marker for the combined expression of CD34 and CD33 than it is able to define a specific subpopulation of AML blasts. AC133 expression on G-CSF-mobilized peripheral blood progenitors in AML patients in CR correlates with CD34 expression, indicating its usefulness as an alternative to CD34 for selecting peripheral blood progenitors for autografts or purge them from leukemic progenitors by negative selection in AC133− AML.
195
References [1] Yin AH, Miraglia S, Zanjani ED, Almeida-Porada G, Ogawa M, Leary AGXOJ, Kearney J, Buck DW (1997) AC133, a novel marker for human hematopoietic stem and progenitor cells. Blood 90 (12), 5002-5012. [2] de Wynter EA, Buck D, Hart C, Heywood R, Coutinho LH, Clayton A, Rafferty JA, Burt D, Guenechea G, Bueren JA, Gagen D, Fairbairn LJ, Lord BI, Testa NG (1998) CD34 + AC133 + cells isolated from cord blood are highly enriched in long- term culture-initiating cells, NOD/SCID-repopulating cells and dendritic cell progenitors. Stem Cells 16(6), 387-396. [3] Buhring HJ, Seiffert M, Bock TA, Scheding S, Thiel A, Scheffold A, Kanz L, Brugger W. (1999) Expression of novel surface antigens on early hematopoietic cells. Ann N Y Acad Sci 872, 25-38. [4] Kratz-Albers K, Zuhlsdorp M, Leo R, Berdel WL, Buchner T, Serve H (1998) Expression of a AC133, a novel stem cell marker, on human leukemic blasts lacking CD34-antigen and on a human CD34+ leukemic line:MUTZ-2 [letter]. Blood 92(11), 4485-4487. [5] Waller CF, Martens UM, Lange W. Philadelphia chromosomepositive cells are equally distributed in AC133 + and AC133fractions of CD34 + peripheral blood progenitor cells from patients with CML (1999) [letter] Leukemia 1999; 13(9):14661467. [6] Vavrova J, Filip S, Vokurkova D, Blaha M, Vanasek J, Jebavy L (1999). Ex vivo expansion CD34 +/AC133 +-selected autologous peripheral blood progenitor cells (PBPC) in high-risk breast cancer patients receiving intensive chemotherapy [In Process Citation]. Hematol Cell Ther 41(3), 105-112. [7] Buhring HJ, Seiffert M, Marxer A, Weiss B, Faul C, Kanz L, Brugger W (1999) AC133 antigen expression is not restricted to acute myeloid leukemia blasts but is also found on acute lymphoid leukemia blasts and on a subset of CD34 + B-cell precursors [letter]. Blood 94 (2), 832-833. [8] Horn PA, Tesch H, Staib P, Kube D, Diehl V, Voliotis D (1999) Expression of AC133, a novel hematopoietic precursor antigen, on acute myeloid leukemia cells [letter]. Blood 93 (4), 1435-1437. [9] Miraglia S, Godfrey W, Yin AH, Atkins K, Warnke R, Holden JT, Bray RAX, Waller EK, Buck DW. (1997) A novel fivetransmembrane hematopoietic stem cell antigen: isolation, characterization, and molecular cloning. Blood 90 (12), 5013-5021. [10] Corbeil D, Roper K, Weigmann A, Huttner WB. (1998) AC133 hematopoietic stem cell antigen: human homologue of mouse kidney prominin or distinct member of a novel protein family? [letter]. Blood 91 (7), 2625-2626. [11] Snell V, Jackson E, Buck D, Andreeff M. (1998) Expression of the AC133 antigen in leukemic andnormal progenitors. Blood 92 119a (abstr, suppl 1). [12] Batinic D, Tindle R, Boban D, Tiefenbach A, Rajic L, Labar B, Nemet D, Boranic M (1989) Expression of hematopoietic progenitor cell-associated antigen BI-3C5/CD34 in leukemia Leukemia Research 13 (1), 83-85 [13] Harada N, Okamura S, Kubota A, Shimoda K, Ikematsu W, Kondo S, Harada M, Niho Y (1994) Analysis of acute myeloid leukemia cells by flow cytometry, introducing a new light-scattering classification Cancer Research in Clinical Oncology 120 (9), 553-557 [14] Sperling C, Buchner T, Creutzig U, Ritter J, Harbott J, Fonatsch C, Sauerland C, Mielcarek M, Maschmeyer G, Loffler H, et al. (1995) Clinical, morphologi, cytogenetic and prognostic implications of CD34 expression in childhood and adult de novo AML Leukemia an Lymphoma 17 (5-6), 417-426 [15] Gorin NC, Labopin M, Meloni G, Korbling M, Carella A, Herve P, Burnett, A, Rizzoli V, Alessandrino EP, Bjorkstrand B,
196
[16]
[17]
[18]
[19]
F. Fauth et al. / Leukemia Research 25 (2001) 191–196 et al. (1991) Autologous bone marrow transplantation for acute myeloblastic leukemia in Europe: further evidence of the role of marrow purging by mafosfamide. European Co-operative Group for Bone Marrow Transplantation (EBMT). Leukemia 5 (10), 896-904. Gorin NC, Aegerter P, Auvert B, Meloni G, Goldstone AH, Burnett AXCA, Korbling M, Herve P, Maraninchi D, et al. (1990) Autologous bone marrow transplantation for acute myelocytic leukemia in first remission: a European survey of the role of marrow purging. Blood 75 (8), 1606-1614. Rubin J, Malley V, Ball ED (1994) A combination of anti-CD15 monoclonal antibody PM-81 and 4-hydroperoxy-cyclophosphamide augments tumor cytotoxicity while sparing normal progenitor cells. Journal of Hematotherapy; 3 (2), 121-127. Gidali J, Feher I, Kovacs P (1994) Comparative heat sensitivity of murine and human hemopoietic progenitors and clonogenic leukemia cells. Stem Cells 12 (5), 533-538. Lemoli RM, Igarashi T, Knizewski M, Acaba L, Richter A, Jain AXMD, Levy J, Gulati SC (1993) Dye-mediated photolysis is
.
capable of eliminating drug-resistant (MDR + ) tumor cells. Blood 81 (3), 793-800. [20] Nimgaonkar M, Kemp A, Lancia J, Ball ED (1996) A combination of CD34 selection and complement-mediated immunopurging (anti-CD15 monoclonal antibody) eliminates tumor cells while sparing normal progenitor cells. Journal of Hematotherapy 5 (1), 39-48. [21] Laporte JP, Douay L, Lopez M, Labopin M, Jouet JP, Lesage S, Stachowiak, J, Fouillard L, Isnard F, Noel-Walter MP, et al. (1994) One hundred twenty-five adult patients with primary acute leukemia autografted with marrow purged by mafosfamide: a 10-year single institution experience. Blood 84 (11), 3810-3818. [22] Selvaggi KJ, Wilson JW, Mills LE, Cornwell GG3, Hurd D, Dodge WXGR, Martin SE, McMillan R, Miller W, et al. (1994) Improved outcome for high-risk acute myeloid leukemia patients using autologous bone marrow transplantation and monoclonal antibody-purged bone marrow. Blood 83 (6), 1698-1705.