The Presence of Lymphoid-Associated Antigens in Adult Acute Myeloid Leukemia is Devoid of Prognostic Relevance

Transcription

1 The Presence of Lymphoid-Associated Antigens in Adult Acute Myeloid Leukemia is Devoid of Prognostic Relevance Francesco Lauria,a Donatella Ra.spadori,b Maria Alessandra Ventura,b Damiano Rondelkh Nicoletta Testoni,b Patrizia Tosi,h Mariagrazia Michieli, Daniela Damiani: Maria Rosa Motta,h Sante Turab Istituto di Scienze Mediche, Universith di Milano, Milan, Italy; %tituto di Ernatologia L. e A. Serhgnoli, Universith di Bologna, Bologna, Italy; Dipartirnento di Scienze Mediche e Morfologiche, Universith di Udine, Udine, Italy Key Words. mabs AML surface antigens CD7 CD34 CR Abstract. The immunophenotype of 110 adult patients with diagnosis of acute myeloblastic leukemia (AML) was analyzed using a wide panel of monoclonal antibodies (mabs). Leukemic blasts were tested by applying direct immunofluorescence analysis and dual-fluorescence staining, and two groups of patients were identified: 56/110 (51%) expressing only myeloid antigens (MyIAML) and 54/110 (49%) expressing both myeloid and lymphoid antigens (Ly/AML). CD13 and CD33 were expressed in almost all FAB subtypes, whereas CD14, frequently expressed in M4 and M5 subtypes (70%), was rarely expressed in MO + M1 cases (9%). On the contrary, CD34, expressed in 77% of MO + M1 cases, was practically absent in M3 and M5 subtypes (6% and 7%, respectively). CD2 and CD7 antigens were found in 34% and 42% of patients respectively, whereas B cell-associated antigens, such as CDlO and CD19, were found in 31% and 18% of patients. Cytogenetic abnormalities characteristically present in AML patients were also analyzed and, except for t(8;21) which was found in both groups of patients, the other abnormalities were frequently found in cases coexpressing lymphoid-associated antigens. Finally, the complete remission (CR) rate, survival and event-free survival were analyzed according to the presence of lymphoid markers and also of some specific antigens such as CD7 and CD34. The only prognostic difference was represented by CD34 patients who showed a reduction in the CR rate compared with CD34- patients (65% versus 82%) QI = 0.05) which became more evident when the mean intensity of fluorescence was considered. Correspondence: Dr. Donatella Raspadori, Istituto di Ematologia L. e A. Serhgnoli, Via Massarenti Bologna, Italy. Received January 26, 1995; provisionally accepted February 18, 1995; accepted for publication March 26, OAlphaMed Press /95/$5.00/0 In conclusion, mabs conjugated with fluorochromes and analyzed by more sophisticated cytometers allow the identification of a higher number of AML cases bearing lymphoid-associated antigens, but this phenotypic coexpression is not associated with biologically distinct forms of leukemia. Introduction Immunological phenotyping is regarded as an essential tool for the diagnosis and classification of acute leukemias [ 1,2] such as in acute lymphoblastic leukemia (ALL) where it may also play a relevant role in predicting response to therapy and clinical outcome [3]. In recent years, the application of a wide panel of monoclonal antibodies (mabs) has allowed the identification of leukemias characterized by the presence of blast cells coexpressing both myeloid and lymphoid markers [4-81. The recognition of cases with an inappropriate antigenic expression, termed biphenotypic or hybrid, may be useful toward shedding new light on the biological aspects of acute leukemias, in terms of clinico-therapeutical implications. Several studies have so far been published [9-121, but the frequency, significance and prognostic role of these hybrid leukemias still remain a matter of debate. These conflicting results may be due to heterogeneous and/or selected cases and to technical reasons, such as the use of direct or indirect immunofluorescence, quality of cytometers and different criteria for positive antigen expression [ STEM CELLS 1995;13:

2 429 Lymphoid-Associated Antigens in AML The introduction of multidimensional flow cytometric analysis and of mabs conjugated with fluorochromes, which have increased the sensitivity and specificity of leukemic cell detection, prompted us to reassess the immunophenotype of acute myeloblastic leukemia (AML) cases observed at our institution. Herein, the immunophenotyping analysis conducted in 110 cases of adult AML is presented and correlated with the morphological and cytogenetic aspects, and with the clinical outcome. Patients and Methods One hundred and ten consecutive adults with newly diagnosed AML were classified according to FAB criteria [ 161 and studied with adequate immunophenotypic and cytogenetic analysis. All patients except those older than 70 years underwent the standard Daunomycin-3d + Ara-C-7d continuous infusion chemotherapy. Isolation of Leukemic Cells Leukemic cells were obtained following Lymphoprep (Nycomed Pharma AS, Oslo, Norway) gradient separation from bone marrow and/or peripheral blood of 1 10 patients. After washing twice, the cells were resuspended in RPMI 1640 medium (Seromed Biochrome KG, Berlin, Germany) supplemented with heat-inactivated fetal bovine serum (Seromed Biochrome KG) and 10% dimethylsulfoxide and frozen in liquid nitrogen. Immunophenotyping Immunological characterization was performed on fresh (60% of cases) and/or cryopreserved leukemic cells. For the latter, the blasts were thawed, washed twice, incubated at 37 C overnight, and tested for viability with propidium iodide staining. Immunophenotyping analysis was carried out only if viable cells were >90%. All leukemic samples were incubated with a panel of fluorescein (FITC) and phycoerythrin (PE)- conjugated mabs reactive with myeloid and lymphoid antigens (CD13, CD14, CD33, CD34, CD2, CD7, CD4, CDlO, CD19, CD4la, glycophorin A; Becton Dickinson, San Jose, CA) employed in flow cytometric analysis (FACScan, Becton Dickinson). Appropriate isotype controls were performed using IgG, FITC + IgG2, PE mabs. Lymphoid-associated antigens were considered present when more than 20% of the blastic cells were positive, and the multiantigenic expression was further confirmed by dual-fluorescence analysis performed using PE-conjugated antisera against myeloid antigens and FITC-conjugated antisera against lymphoid antigens. Data acquisition and analyses were performed using the LYSYS II Software (Becton Dickinson, San Jose, CA); in all cases 10,000 events were acquired. The analysis was restricted to the blast population by gating on forward and side light scatter parameters, excluding the majority of lymphocytes, monocytes and myeloid granulated precursors. Furthermore, for each antigenic analysis, the mean intensity of fluorescence (MIF), expressed as the ratio of sample mean channelxontrol mean channel, was measured. For intracytoplasmic detection of CD3, MP07 and CD22 antigens, cytospin preparations of leukemic blasts were fixed in periodate lysine paraformaldehyde for 20 min at 4 C and subsequently incubated with anti-mp07 (DAKO, Glostrup, Denmark), CD22 and CD3 mabs (Becton Dickinson, San Jose, CA) using an indirect immunofluorescence technique. TdT determination was performed with a commercially available immunofluorescence kit (Immunotech SA, Marseille, France) and was considered positive if 20% of the cells had nuclear fluorescence. Cytogenetic Analysis At diagnosis, the chromosome analysis of bone marrow cells was performed on short-term cultures. Synchronization with methotrexate and bromodeoxyuridine was performed in some cases. The banding techniques used were R-banding with acridine orange and G-banding with Wright s stain. In each case, at least 15 cells were analyzed. Chromosome aberrations were described in accordance with the International System for Human Cytogenetic Nomenclature [ 171. Statistics Since the purpose of this study was the evaluation of the relationship between the immunophenotype of leukemic blasts and the response to therapy, patients who died prior to day 14 were excluded from the analysis for remission, while all patients were included in survival and event-free survival analysis. Differences in response rates among subgroups of patients were analyzed by Chi-square test. Life tables for survival and event-free survival were

3 Lauria et al. 430 constructed by the method of Kaplan and Meier [ 181, with differences compared by the log-rank test. Early death or failure to enter remission was considered an event at zero time. Patients who underwent bone marrow transplantation were censored at the time of the procedure. Results One hundred and ten patients, morphologically and cytochemically diagnosed as AML, entered this study. Detailed clinical, hematological and morphological data of the patients are reported in Table I. There were 63 males and 47 females with a median age of 45 years (range 14-82). According to the FAB criteria, 3 patients had an MO cytotype, 19 MI, 27 M2, 17 M3, 31 M4 and 13 M5. lmmunnphenotypic Analysis Leukemic blasts of all cases were tested with a wide panel of mabs, and the overall results showed that 56 of the 110 patients (51%) expressed only myeloid antigens (MyIAML), and (49%) expressed myeloid and lymphoid antigens (LyIAML). Lymphoid antigens Table I. Clinical and morphological characteristics of the two groups of AML patients with or without the expression of lymphoid antigens No. of patients Median age (range) Male/Female FAB morphology MO M1 M2 M3 M4 M5 Results CR' CCR' MyIAML (14-82) 32/ /41 (77%) 15/31 (48%) LyIAML (16-66) 3 1/ /46 (72%) 15/33 (45%) 'CR = Complete remission; *CCR = Continuous complete remission Fig. 1. Dual-fluorescence analysis showing blast cells expressing CD7 antigen on the total number of leukemic cells. On the right side of the figure, CD19 is expressed only in 41% of myeloid blasts. when present, could be expressed on the surface of all leukemic cells (67%) or only on a proportion of them (33%) (Fig. 1). Six cases were also TdT positive. Expression of Myeloid and Lymphoid- Associated Antigens All AML patients expressed at least one of the myeloid-associated antigens. As reported in Table 11, CD 13 and CD33 were highly and consistently expressed in almost all subtypes of AML while CD14 was frequently expressed in M4 (60%) and M5 (70%) subtypes. On the contrary, CD34 was rather frequently present in MO + M1 (77%) and M4 (74%), and practically absent in M3 and M5 subtypes (6% and 7%, respectively). The expression of T and B lymphoid markers, in correlation with the FAB cytotype, is summarized in Table 11. Fifty-four patients expressed lymphoid markers: 20 of them (37%) expressed only T markers and 34 (63%) expressed both T and B lymphoid markers; no case coexpressed B lymphoid markers only. In particular, CD2, CD4 and CD7 antigens were progressively more expressed from MO + M1 to M5 with mean values of 34%, 39% and 42%, respectively (Table 11). Of the 48 CD7 positive patients, 20 also coexpressed CD34 antigen, while the sole expression of CD7 was recorded in only 6 of them. Concerning the B lymphoid-associated antigens, CDlO and CD19 were highly detected in M4 and M5 subtypes. Nine out of the 13 cases with M5 cytotype fall within the LyIAML group (Table 11). Finally, intracytoplasmic CD3 and CD22 antigens were found to be negative in all cases. In order to clarify the role of fluorescence analysis in the variability of lymphoid antigen

4 43 1 Lymphoid-Associated Antigens in AML Table 11. Correlation between FAB cytotypes and antigenic expression CD13 CD33 CD14 CD34 CD2 CD4 CD7 CDlO CD19 91% 86% 9% 77% 14% 18% 40% 14% 9% 88% 100% 15% 52% 18% 26% 33% 15% 10% 94% 94% 41% 6% 35% 35% 23 % 12% 6% 100% 96% 60% 74% 42% 55% 55% 45% 29% 84% 92% 70% 7% 61% 61% 61% 70% 38% 91% 94% 39% 43% 34% 39% 42% 31% 18% 'No. of patients expression, blasts of four AML patients were tested by applying both indirect and direct immunofluorescence using the same fluorochrome (FITC) and the same mab clone. As shown in Figure 2, though the proportion of positive cells was similar with both techniques, fluorescence intensity results were remarkably higher with direct technique. Cytogeizetics Cytogenetic analysis was performed in 41/56 patients expressing only My/AML and in 52/54 patients showing both myeloid and lymphoid antigens, and the cytogenetic pattern was also correlated with the expression of some specific antigens, such as CD2, CD7, CD19 and CD34. Normal karyotypes were observed in both groups of patients, as well as the t(8;21) which was unrelated to the CD19 expression. %IB CR Fig. 2. Fluorescence profile of CD7 lymphoid associated antigen obtained with direct (A) and indirect (B) immunofluorescence technique on blasts of one AML case. The intensity of fluorescence is remarkably higher with direct technique (MIF 9.8 versus 3.4, respectively). w Other clonal abnormalities characteristically associated with AML, including t(15;17), 1 lq, 5q abnormalities, +8, inv(3) and inv/del( 16), were almost exclusively found in AML patients expressing lymphoid markers. Results and Prognostic Aspects Out of 110 patients, eight were excluded due to an ineligibility for any treatment because of age. Fifteen additional patients were not considered for remission analysis because they died during the induction therapy (prior to day 14), but were regularly included in survival and eventfree survival analysis. Thirty-one out of 41 patients (77%) expressing only myeloid antigens achieved complete remission (CR), and 15 of them are still in CR with a median duration of 9.1 months (Table I). In the group of patients coexpressing myeloid and lymphoid antigens, 33/46 (72%) achieved CR, and 15 of them are still in CR with a median duration of 8.3 months (Table I). It should be noted that four patients in the MyIAML group and eight in the other group, following CR, were submitted to autologous bone marrow transplantation, and all of them but two (one in the first group and one in the second group) are still in CR. The CD2, CD4, CD 10 and CD 19 antigenic expression was evaluated, but no prognostic implication was found. Moreover, CR rates were similar also in CD7' (75%) and CD7- (70%) patients, but different in CD34- and CD34' cases (82% versus 65%) (p = 0.05). Among the latter cases, the CR rate was significantly lower in patients showing a higher fluorescence intensity (MIF > 20) with respect to those showing a MIF < 20 (48% vesus 88%) (p = 0.01). Finally, all groups of patients had an identical survival and

5 Lauria et al. 432 event-free survival except for those with CD34 expression (MIF > 20) which showed a significant reduction in both survival and event-free survival compared with those with a MIF < 20. Discussion The presence of T and B lymphoid markers on myeloid blasts has been variably reported in several studies, and correlations have been attempted with morphology, cytogenetic and clinical outcome. Generally, in early immunophenotypic studies, a lower proportion of AML cases coexpressing lymphoid antigens had been reported, [6, 13, 191, whereas, more recently, larger numbers of patients have been evaluated and an increased number of patients showing both myeloid and lymphoid antigens has been described [9, 10, 12, 14, 15,201. In this present study, unlike several previously published reports, the expression of myeloid and lymphoid-associated antigens was evaluated by using conjugated mabs and by regularly applying the dual-staining technique and correlating the results with the FAB morphology, cytogenetic findings, response to therapy and survival and event-free survival. Unexpectedly, AML cases expressing only myeloid antigens were 5 1 % while the remaining 49% coexpressed both myeloid and lymphoid antigens. CDI 3 and CD33 antigens were expressed in almost all cases while CD34, which is considered a common antigen shared by immature precursors and variably associated with a poor prognosis [21,22], was, in our series, detected in 43% of cases. In agreement with previously published results [21-231, we confirmed a significant difference (p = 0.05) in CR rates between CD34 and CD34- patients (65% versus 82%, respectively); this difference, however, did not affect survival or event-free survival. Lymphoid-associated antigens, such as CD2, CD4 and CD7, unlike previous contributions [24-271, were highly expressed on myeloid blasts in our series. In particular, CD7 which has been associated with biasts at an early step in hematopoietic differentiation [25, 261 and with a poor prognosis [24-271, in our series was observed in an elevated number of cases (43%) without any prognostic implication. In fact, according to our results, CD7 and CD7- patients had a similar CR rate (75% versus 70%), as well as similar survival and event-free survival rates. When CD7 patients were further analyzed and subdivided according to the simultaneous presence of the CD34 antigen, a reduced CR rate was found in CD7+/CD34+ patients compared to CD7+/CD34- cases (65% versus 80%), suggesting that the lower probability of achieving CR was related to the CD34 and not to the CD7 antigenic expression. The increased incidence of lymphoid-associated antigens reported in our series may be attributed to different reasons ranging from the different clones employed to different immunofluorescence analysis. We clearly demonstrated that direct analysis, combined with routine dual staining, produced better detection of lymphoid antigens on myeloid cells (Fig. 2). In fact, when these methodologies were regularly applied by other investigators, the expression of lymphoid antigens on AML blasts was observed in 40% of cases [12] and the expression of CD7 was reported in 31.8% of cases by Del Poeta et al. [28]. Furthermore, the findings reported by Kifu et al. [24] which showed only 19% of CD7 AML patients with a poorer prognostic outcome, may be explained by the fact that the patients were studied with a less sensitive technique (indirect immunofluorescence) and that 35/40 CD7 cases also coexpressed the CD34 antigen which probably accounts for the worse prognosis, as in our experience. On the other hand, the use of a less adequate technique (microscopic fluorescent analysis) for the detection of TdT+ AML cases may explain the low incidence of these cases in our study. According to the present immunophenotypic study and clinical correlation, it appears that the presence of lymphoid-associated antigens on myeloid blasts does not necessarily identify patients with early or immature phenotype and/or with a poorer prognosis. However, the identification of these hybrid cells may be useful for the detection of minimal residual disease; in fact the persistence, even at very low concentrations, of these unusual cells in patients conventionally defined in CR is associated with a significant reduction in the duration of CR [ 101. With regard to the cytogenetic studies, we confirmed previous observations [ 14,291 which demonstrated that the clonal cytogenetic abnormalities characteristically associated with AML are more frequently found in cases displaying both myeloid and lymphoid antigens. This observation is intriguing, but its explanation remains

6 433 Lymphoid-Associated Antigens in AML unclear. According to our findings, none of the major cytogenetic abnormalities were specifically correlated with a single T or B lymphoid antigen, despite previous reports claiming a relationship with some cytogenetic pattern, i.e., t(8;zl) and CD19 expression [lo, 301 or chromosome 5 involvement and CD7 expression [31]. In conclusion, the use of direct immunofluorescence has allowed the identification of an increased number of AML patients expressing lymphoid-associated antigens. However, this finding appears to have no clinical relevance, while the presence of CD34 antigen, represents a more reliable poor prognostic factor in AML patients. Acknowledgments This research was supported in part by MURST 40%-60%. References 1 Chan LC, Pegram SM, Greaves MF. Contribution of immunophenotype to the classification and differential diagnosis of acute leukaemia. Lancet 1991 ;1: Foon KA, Todd RF 111. Immunologic classification of leukemia and lymphoma. Blood 1986;68: Greaves MF, Janossy G, Pet0 J et al. Immunologically defined subclasses of acute lymphoblastic leukaemia in children: their relationship to presentation features and prognosis. Br J Haematol 198 1;48: Stass SA, Mirro J Jr. Lineage heterogeneity in acute leukaemia: acute mixed-lineage switch. Clin Haematol 1986;15: Greaves MF, Chan LC, Furley AJW et al. Lineage promiscuity in hemopoietic differentiation and leukemia. Blood 1986;67: Drexler HG. Classification of acute myeloid leukemia. A comparison of FAB and immunophenotyping. Leukemia 1987;1: Mirro J Jr, Kitchingman GR. The morphology, cytochemistry, molecular characteristics and clinical significance of acute mixed-lineage leukemia. In: Scott CS, ed. Leukaemia Cytochemistry. Chichester: Ellis Horwood, 1989: Paietta E. Ambiguous phenotypes in acute leukemia. Leuk Lymphoma 1990;2: Catovsky D, Matutes E, Buccheri V et al. A classification of acute leukaemia for the 1990s. Ann Hematol 1991;62: Reading CL, Estey EH, Huh YO et al. Expression of unusual immunophenotype combinations in acute myelogenous leukemia. Blood 1993;81 : Drexler HG, Thiel E, Ludwig WD. Acute myeloid leukemias expressing lymphoid-associated antigens: diagnostic incidence and prognostic significance. Leukemia 1993;7: Buccheri V, Matutes E, Dyer MJS et al. Lineage commitment in biphenotypic acute leukemia. Leukemia 1993;7: Pui CH, Raimondi SC, Head DR et al. Characterization of childhood acute leukemia with multiple myeloid and lymphoid markers and at relapse. Blood 1991;78: Ball ED, Davis RB, Griffin JD et al. Prognostic value of lymphocyte surface markers in acute myeloid leukemia. Blood 1991 ;77: Terstappen LWMM, Safford M, Konemann S et al. Flow cytometric characterization of acute myeloid leukemia. Part 11. Phenotypic heterogeneity at diagnosis. Leukemia 1991;5: Bennett JM, Catovsky D, Daniel MT et al. Proposed revised criteria for the classification of acute myeloid leukemia. Ann Intern Med 1985; 103~ Harnden DG, Klinger HP, eds. An international system for human cytogenetic nomenclature. Basel: Karger, Original article series, Birth Defects 1985;21: Kaplan EL, Meier P. Non parametric estimation for incomplete observations. J Am Stat Assoc 1958;53: I. 19 Ludwig WD, Bartram CR, Ritter J et al. Ambiguous phenotypes and genotypes in 16 children with acute leukemia as characterized by multiparameter analysis. Blood 1988;71: Adriaansen HJ, Van Dongen JJM, Kappers- Klunne MC et al. Terminal deoxynucleotidyl transferase positive subpopulations occur in the majority of ANLL: implications for the detection of minimal residual disease. Leukemia 1990;4: Geller R, Zahurak M, Hurwitz CA et al. Prognostic importance of immunophenotyping in adults with acute myelocytic leukaemia: the significance of stem-cell glycoprotein CD34 (My 10). Br.I Haematol 1990;76: Myint H, Lucie NP. The prognostic significance of the CD34 antigen in acute myeloid leukaemia. Leuk Lymphoma 1992;7:

7 Lauria et al Borowitz MJ, Gockerman JP, Moore JO. Clinicopathologic and cytogenic features of CD34 (My 10) positive acute nonlymphocytic leukemia. Am J Clin Pathol 1989;91: Kita K, Miwa H, Nakase K et al. Clinical importance of CD7 expression in acute myelocytic leukemia. Blood 1993;81: Lo COCO F, De Rossi G, Pasqualetti D et al. CD7 positive acute myeloid leukaemia: a subtype associated with cell immaturity. Br J Haematol 1989;73: Kondo S, Okamura S, Harada N et al. CD7 positive acute myeloid leukemia: further evidence of cellular immaturity. J Cancer Res Clin Oncol 1992;118: Jensen AW, Hokland M, Jorgensen H et al. Solitary expression of CD7 among T-cell antigens in acute myeloid leukemia: identification of a group of patients with similar T-cell receptor p and 6 rearrangements and course of disease suggestive of poor prognosis. Blood 1991;78: Del Poeta G, Stasi R, Venditti A et al. Prognostic value of cell marker analysis in de novo acute myeloid leukemia. Leukemia 1994;8: Cuneo A, Michaux JL, Ferrant A et al. Correlation of cytogenetic patterns and clinicobiological features in adult acute myeloid leukemia expressing lymphoid markers. Blood 1992;79: Kenkichi K, Nakase K, Miwa H et al. Phenotypical characteristics of acute myelocytic leukemia associated with the t(8;21)(q22;q22) chromosomal abnormality: frequent expression of immature B-cell antigen CD19 together with stem cell antigen CD34. Blood 1992;2: Lutz D, Kasparu H, Nowotny H et al. Association of gp40/cd7+ acute myeloblastic leukemia and chromosome 5 aberrations. Haematol Blood Transfusions 1990;33: