Document Lo3Xk6vpm1NkDg54EVKzMwQb

11i PLAINTIFF'S +;[TTT 4 0 w P Original Papers F. H.GARDNER, M.D. Acta h o e m u 80: 61-64( 1988) Early Expressionof MCS2 (CD13)in the Cytoplasm of Blast Cells from Acute Myeloid Leukaemia M.S. Pombo de Oliveira. Esrela Matutes. Sudha Rani. Ricardo Morillo. Daniel Catovsky MRC Leukaemia Unit, Royal Postgraduate Medical School, London, U K - -Key Words. Acute myeloid leukaemia Cell markers Monoclonal antibodies Abstract. The expression of two myeloid antigens identified by the monoclonal antibodies (McAb) MCSZ (CDI3) and MY9 (CD33) was investigated in 136 cases of leukaemia. MCSZ was positive in blast cells of 78 of 88 (88.5%) and MY9 in 51 of 81 (64%) cases of acute myeloid leukaemia (AML)and chronic aranulocytiE leu- kaemia in myeloid blast crisis. One or other McAb, or both, were positive in all but 2 (2.3%) of these cases. MCS2 was more sensitive than MY9 to detect blasts of the myeloid lineage due to its most frequent reactivity in the cytoplasm of fixed cells by the immunoperoxidase (IP)technique compared with its membrane expression on cell suspensions by immunofluorescence(IF). MY9 was not suitable for tests on fixed cells. MCSZ was positive by IP but not by IF in 24% of AML, but the reverse was not obseived. This suggests that the antigen detected by MCSZ is expressed in myeloblasts first in the cytoplasm and later on the cell membrane, pattern which is similar to that of the early antigens CD3 and CD22 in T and B lineage lymphoblasts, respectively. MCSZ was always positive in FAB types of AML-involving myeloblasts (M I-M4), including cases of undiffer- entiated morphology (MO),whilst MY9 was more frequently positive in monocytic leukaemia (M5). On the other hand, MCSZ was positive in 4 of 33 cases of acute lymphoblastic leukaemia and MY9 in I . We conclude that both McAb, particularly MCS2, contribute to the better characterisation of myeloid leukaemias but that other tests are required to clarify the nature of the blasts when unexpected reactivities are observed. Introduction The classification of acute myeloid leukaemia (AML)has been based largely on morphological and cytochemical findings [I]. The availability of monoclonal antibodies (McAb) with myeloid specificity has provided new insights into the characterisation of A M L Although a relatively large number of McAb identifying myeloid-associated antigens has been described, only a small number react with early myeloid cells and appear to be of value for the phenotyping of A M L One of them is MY9 (CD33), an McAb that reacts with a p67 protein expressed in committed myeloid progenitor cells [2]. Another is MCS2 (CDI3). a reagent specific for cells at an early maturation stage along the myeloid pathway [3]. MCSZ has the'same specificity as MY7 (CD13) but appears to react with antigens of different molecular weight [4]. We have compared the pattern and reactivity of MY9 and MCSZ using immunofluorescence (IF) and immunoperoxidase (IP) methods and have shown that the IP technique on fixed cells significantly increases the sensitivity of MCSZ for the recognition of AML blast cells and suggests that the p 150 antigen detected by MCSZ is expressed in the cytoplasm of myeloblasts before its membrane expression. -- I Material and Methods Cells were obrained from peripheral blood (PB) and/or bone marrow (EM) from 136 patients studied at diagnosis. These included 65 case of AML, 33 of acute lymphoblastic leukaemia (ALL), 24 of chronic granulocytic leukaemia in blast crisis (CGLBC). 14 of chronic lymphocytic leukaemia and non-Hodgkin lymphomas. Diagnose were established by morphological criteria on May-Griinwald-Giemsa-stainedBP and 'BM films and the cyto- I 61 de Oliveira/Matutes/ Rani/Morill3Catovsky chemical reactions of Sudan black B. a-naphthyl-acetate-estemse dnd acid phosphatase, as previously described [SI.The diagnosis of ALL was based on cytology and cytochemistry supported by immunological evidence of the presence of E- or T-cell anrigens. The diagnosis of AML was based on the FAB criteria [I]as well as on the absence of lymphoid markers. All samples of ALL and AML analysed had more than 50%blast cells. Cell Markers Mononuclear cells from BP and/or BM were isolated by Lym- phoprep (Nyegaard). An indirect IF technique was performed on cell suspensions using McAb, followed by a second layer of a fluorescein-isothiocyanate-conjugatedgoat antimouse F(ab)2 serum (Cappel). The cells were washed in phosphate-buffered saline (PBS) resuspended in PBS/glycerol (SO/SO) mounted on slides and examined under a Zeiss 14 fluorescence microscope. A control in which the first layer of McAb was replaced by non-immune mouse ascites was always included. AB Serum at a 2% dilution was used throughout to block non-specific binding to Fc receptors. The following mouse McAb were used: GRBl (IgGZa), gift from Prof. Garrido-Torres. against class I I major histocompatibility complex antigen. MY9 (Coulter-clone. U K lgC2, CD33) 121 and MCSZ (Nicherei Corporation, Japan: IgG 1, CDI 3) reactive with imma- ture myeloid and monocytic cells [3]; B4 (Coklter-clone, UK: CD19) against B-lineage blasts. and B-lymphocytes; OKB-CALLA (Orthoclone, UK) and/or Nu- I (Nicherei Corporation, Japan) both McAb against the common ALL antigen (CDIO); antiglyco- Table 1. Expression of MCSZ in AML and CGL-BC tested by two methods ~ Numberof IF+ IP+ IF+/IP+ IF-/IP+ IF-/IPcases tested 50 32 44 32 12 6 phorin A (LICR LONIRIO) used to recognise erythroid precurSOTS. (gift from Dr. Edwards; C17; IgG1. CDw41). against pla~elet glycoprotein (GP) llla (gifl Of Dr. Tetteroo). and ANSI (IgG. CDw42. Dakopatu. UK), which recognises an antigen on CPlb. were used to demonstrate megakaryoblasts. Terminal deoxynucle- otidyl transferase (TdT)was assessed by an indirect IF method de- scribed by Bollum [6] using a rabbit anti-TdT serum (Sero-lab. UK). MCS2 was also studied in all c a m by the IP reaction on fixed cells. Cytoccntrifuged slides were fixed for 10 min with acetone; endogenous peroxidase was inhibited by immersing the slides in chloroform for IO min at room temperature. After PBS washing for 5 min, a first incubation was made with 2% AB serum, followed by the McAb MCS2 for 30 min used at a I :20 dilution of the I :100diluted ascites used for the IF method. Two layers of honeradish-per- oxidase-conjugated antibody were employed, rabbit antimouse and swine antirabbit (Dakopatts. UK). After washing, the peroxi- dase reaction was carried out with 3,3'-diamino-benzidine-tetrahy- drochloride (Sigma, UK)in PBS for IO min in the dark. The slides were counterstained with Harris' hematoxylin lor 2 min, dried and mounted in DPX and analysed by comparing with a negative control. Results A comparison between MCS2 investigated by IF on cells suspensions and by IP on fixed cells was made in samples from 50 cases of A d L and CGL-BC (table I). In 76% of cases both methods gave similar results:+ 64% both werepositive and in 12% both tests were negative. In 12 cases (24%) MCSZ was posi- tive in >20% of the blasts cells when tested by IP whilst it was completely negative by IF. In addition, when both tests were positive, the percentage of reactive cells was 10-80% higher by IP than by IF.In 12 Table 11. Expression of myeloid antigens detected by MCSZ and MY9 in leukaemiccells Leukaemia type Number of cases tested MCS2 + I MY9 + MCSZ +/ MY9+ AML CGL-BC ALL NHVCLL 6 9 59 38 35 24' 19 13 13 332 4I 1 14 0 0 D Total I36 82 51 - -NHL Non-Hodgkin lymphomas: CLL chronic lymphocytic leukaemia. ' MCS2 assessed by IFand/or I P. ' MY9 was not investigated in 6 cases and MCS2 was not tested in 1. MY9 was not studied in 2 CGL-BC and 6 ALL 49 MCS2 +/ MY9- ~ ~~~ MCS2-/ MY9 + MCS2-/ MY9- 20 2- 2- 50 0 20 27 00 14 ~- 27 2 47 ..., .. .__. Cytoplasmic Expression of MCS2 in AML 63 cases who were positive by IF, MY9 was tested by IP Table 111. Reactivity of MCS2 and MY9 in FAB types of A M L with three different fixatives using various dilutions, but no positive results were obtained by IP. The re- FAB type MCSZ MY9 sults that follow will refer to MCSZ tested by IF or IP and to MY9 by IF only. The results obtained with both McAb on the blast cells from the 136 cases studied are summarised in table 11. Overall 88.5% of cases of AML and CGL-BC were positive with MCS2, whereas only 64% of cases were positive with MY9; 60% of cases were positive with both McAb; 31% of cases were positive with MCSZ and negative with MY9 and 2.5% were negative with MCS2 and positive with MY9: Overall one tested n -MO'(n IO) MI (n- 16) M2.M3.M4(n-18) -M 5 a + b ( n - 1 3 ) M6. M 7 (n 8) 9 16 I8 13 a positive' n% 9 loo 16 100 I8 100 IO 77 6 75 Total(n-65) 64' 59 92 tested positive' n n ?'a IO 4 40 15 10 67 17 IO 59 I I IO 91 6 4 67 59 38 64 or other McAb were positive in 93% of all cases of myeloid leukaemia. It is of interest that the only cases 'I Positive cases: > 20%positive blast cells. M O cases with undifferentiated morphology, negative Sudan which were MCS2- and/or MY9-negative involved Black E. negative lymphoid markers and positive myeloid anti- megakaryoblasts: 2 had AML (M7)and 4 had CGL in gen(s), confirmed by peroxidase at the eiectron-microscopic level ~ megakaryoblasticcrisis. The results in 65 cases of AML analysed according ('5 cases). 8 cases ( I MO, 4 M I , 2 MS and I M7) were negative by IF and positive by IP. to the FA9 classification are shown in table 111. The group of very immature myeloblastic leukaemia de- signated MO was always MCS2-positive, whereas MY9 was positive in only 40% of cases. Overall MCSZ was more often positive than MY9 in most Table IV. C a s e of ALL with myeloid positive markers (Oh of FAB types, except for M5 (table 111). Although 2 positive blasts) cases of A M (M7)were negative (see above), 3 others were positive in 20-68% of blasts. Evidence for Cases Age FAB Ia CD19 CDlO CD3 MCS2 MY9 TdT megakaryocytic differentiation in these 5 cases was provided by McAb against platelet GP and the plate- I 2 let peroxidase reaction at the ultrastructural level. 3 36 L2' 90 92 98 n.d. 20 19 20 86 U 90 60 60 neg. 40 neg. 90 20 L2 100 47 85 neg. 60 n.d. 60 Of 33 cases of ALL (table IV), 27 were of B-lineage 4 68 L2 neg. ntg. neg. 6 9 60 neg. 67 and 6 had T-ALL; in 4 of the 25 cases of adult-ALL, the blasts were positive with one or other McAb in 19-609'0 of cells. Three (cases 1-3) had c-ALL and one (case4) T-ALL (table IV). ' Rearrangement of J n Ig gene demonstrated: other cases not 'tested. E-rosette test 59%and OKTl7 positive in 70% of blasts. Discussioo We have compared the reactivity of two McAb, MCSZ (CD13) and MY9 (CD33) in a series of leukaemic samples which involved blast cells. Although a number of reports have appeared, including one from our own group (71, describing the specificity and sen- sitivity of these reagents for myeloid precursors, almost all studies have been carried out on cell suspen- sions by IF microscopy or flow cytometry. We have, in addition, used systematically MCSZ by I P on fixed cells and showed that this method is more sensitive than IF to demonstrate this antigen in early myelo- blastic leukaemia. It would appear that as other early antigens, CD22 in the B-lineage [ 8 , 9 ]and CD3 and in the T-lineage [8, 91, the antigen detected by MCS2 is expressed first in the cytoplasm and later on the cell membrane. Unfortunately, we could not demonstrate the same phenomenon for the protein identified by MY9, as we were unable to show reactivity with this McAb by IP in 12 cases known to be MY9-positive by IF,although we are aware that others showed positiv- ity with alkaline-phosphatase-labelled antibodies [IO]. As a result of its greater sensitivity by IP (table I), MCSZ was positive in a higher proportion (88.5%) of - 64 de Oliveira/ MatutcslRani/Morilla/Catovsky cases of A M L and CGL-BC than MY9 (64%).How- ever, as a result of the preferential reactivity of MCS2 in A M L cases involving myeloblasts (Ml-M4), including poorly differentiated ones (MO) and of MY9 for monocytic leukaemia (MS),the use of both McAb can recognise 97.7% of these myeloid leukaemias. The diagnostic value ofMCSZ by IP was particularly important in AML cases designated MO where light mi- croscopy, cytochemistry and lymphoid markers are negative. All the 9 cases of MO tested were MCS2-pos- itive whilst only 4 of them were positive with MY9. These cases are of interest because some of them (as well as cases of MI) may be TdT-positive (1 I] and have been considered as ALL by some authors [12]. In 5 of these cases of MO studied by ultrastructural cytochemistry, we have shown myeloperoxidase activity in granules and inmembranous structures confirming their early myeloblastic nature. Despite the greater sensitivity and specificity of MCS2 for the myeloid lineage, 4 (I2.s'~)of bonafide cases of ALL (16% of adult ALL) showed reactivity with MCSZ and one with MY9. Our limited studies so far with immuno- globulin gene rearrangement (table IV) suggest that these are otherwise typical cases of ALL, confirming observations of Sobol et al. [12]. However, further systematic studies by electron microscopy with sensitive methods for myeloperoxidase [13) will be necessary to ascertain whether there is also evidence of myeloid structures on these biphenotypic cases. Nevertheless, the demonstration of myeloid antigens in adult-ALL cases appears to be clinically significant and is associated with a worse prognosis(l2]. Acknowledgements The authors are grateful to Drs. J.D. Griffin, C.I. Edwards, P. Tettcroo. Prof. Gatrido-Torres and Mr. K. Sasaki from Nicherei Co.for their gift ofmonoclonal antibodies. S.R. was supportedby a I'cllowship from the British Commonwealth. Griffin. J.D.: Linch. D.: Sabbath. K.;Larcom, P.:Schlouman, S.F.: A monoclonal antibody reactive with normal and leukemic human myeloid progenitor cell^. Leuk. Res. 8: 521-534 ( 1984). Drexler. H.G.: Sagawa. K.: Menon, M.; Minowada, J.: 'Pan- myeloid' reagent: the monoclonal antibody MCS2 in the routine immunodiagnostic service of leukemia phenotyping. Gann 76: 235-239( 1985). Drexler. H.G.:Sagawa. K.: Menon, M.; Minowada. J.: Reactiv- ity pattern o f 'myeloid monoclonal antibodies' with emphasis on MCS2. Leuk. Res. 10: 17-23 (1986). Dacie, J.V.; Lewis, S.M.: Practical haematology; 6th ed. (Churchill-Livingstone. Edinburgh 1984). Bollum. F.J.: Terminal deoxynucleotidyl transferase as a hematopoieticcell marker. Blood 54: 1203-1215 (1979). Matutes. E.; Rodriguez, B.; Polli. N.; Tavara de Castro, J.; Par- nira. A.; Andrews. C.; Griffin, J.D.; Tindle. RW.; Catovsky, D.: Characterizationof myeloidleukemiaswith monoclonal antibodies3C5 and My9. HemaLOncology 3.- 179-186(1985). Campana. D.: Thompson. J.S.: Amloc. P.: Brown. S; Janossy, G.: The cytoplasmic expression o f CD3 antigens in normal and malignant cells of the T lymphoid lineage. J. Immun. 138: 648-655 (1987). 9 Rani, S.;De Oliveira. M.S.: Catovsky. D.: Different membrane and cytoplasmic expression of CD3 and CD22 in lymphoid malignancies(submitted). 10 Hanson. C.A.; Gajl-Pazalska. J.; Parkin. J.L. BNnning. R.D.: lmmunophenotyping o f acute myeloid leukemi8 using mono- clonal antibodies and the alkaline phosphrtase-antialkaiine .phosphatasetechnique. Blood 70: 83-89(1987). II Parreira, A.: De Oliveira. M.S.P.; Matutes. E.Foroni, L;Ca- tovsky. D.: Terminal deoxynucleotidyl tranrfenrc positive acute myeloid leukaemia. An d a t i o n with immature myeloblastic leukaemia. Br. J. Haemat. 68: (in press, 1988). 12 Sobol. R.E.; Mick, R.: Royston. 1.: Davey, F.R.: Ellison, R.R.; Newman, R.; Cuttner. J.; Griffin. J.D.; Collins, H.; Nelson, D.A.: Bloomfield. C.D.: Clinical importance of myeloid antigen expression in adult acute lymphoblastic leukemia. New Engl. J. Med. 316: I I1 1 - 1 I I7 (1987). 13 Matutes. E.1 Foroni. L.; Amin. S.; De Olivein. M.P.: Brito- Babapulle. V.: Lunatto, L.; Catovsky, D.:'kudo-lymphoid' leukaemia with unusual features: ultastructural. immunological. cytogenetic and molecular studies. Eur. J. Haemat. 38: 303-309 (1 987). References I Bennett. J.M.: Catovsky. D.: Daniel, M.T.; Flandrin. G.;Gal- ton. D.A.G.: Gralnick, H A : Sultan, G.: Proposed revisedcriteria for the ciusification of acutemyeloid leukemia. Ann. intern. Med. 103: 620625 (1985). Received: October 14,1987 Accepted: November 20,1987 DanielCatovsky. MD MRC kukaemia Unit Royal Postgraduate Medical School Ducane Road London WlZOHS(UK) I