Acquisition of CD13and CD33Expression at Relapse on Acute ...

CASE REPORT

Acquisition of CD13and CD33Expression at Relapse onAcute Myeloid Leukemia Cells with an Unusual Phenotype:

MPO+CD13 CD33-Hiroshi Kawada, Yukinobu Ichikawa, Nobumasa Kobayashi* , Ryuki Fukuda,

Shuji Yonekura and Tadami Nagao

A case of acute myeloid leukemia (AML)with an unusual phenotype which was negative fora panel of myeloid antigens determined by flow cytometry, but was strongly positive formyeloperoxidase has recently been reported. Weherein describe a case of AMLwith this unusualphenotype at diagnosis; relapse occurred with the acquisition of CD13and CD33expressions.Morphological features of the blasts at relapse seemed to be more compatible with myeloblasts thanthose at diagnosis. These phenotypic and morphological changes are suggestive of asynchronousdifferentiation, clonal evolution or clonal change of leukemic cells.

(Internal Medicine 32: 733-736, 1993)

Key words: flow cytometry, phenotypic change

Introduction

Recently, flow cytometry (FCM) has become a routinediagnostic procedure in hematology. Myeloperoxidase (MPO)~~myeloid antigen (MAg)+ acute myeloid leukemia (AML) wasclassified as AML-MOby FAB group in 1991 (1). On the otherhand, MPO+MAg~AMLwas first described by Seshi et al in1 992 (2). Weencountered another case with MPO+MAg"AMLat diagnosis, but he relapsed with typical cell-surface pheno-types as AML.

Case ReportA 48-year-old man was admitted to our hospital with a 2-week history of headache. Physical examinations and neuro-logical examinations were unremarkable. Laboratory findingson admission were as follows: white blood cell count 145.2x 1 09/1 (neutrophils 0%, lymphocytes 1%, blast cells 99%); erythro-cyte count 1.77xlO12/l; hemoglobin 6. lg/dl; hematocrit 17.0%;platelet count 32x 109/l; serum lactate dehydrogenase 6 1 6 IU/1.Blast cells consisted of small immature type 1 blasts with highnuclear/cytoplasmic ratio. Auer rods were occasionally ob-

served in the blasts with round nuclei and some nuclear inden-tation. Cytochemically, the leukemic blasts (98 %) were stronglypositive for MPO(substrate: 4-chloro- l -naphthol) (Fig. 1A, C).

The karyotype was of normalmale type. Surface antigens of theleukemic cells in peripheral blood (PB) were analyzed by FCMusing a fluorescein-activated cell sorter (FACSTAR®,Beckton-Dickinson) and a panel of monoclonal antibodies (MoAbs)(Fig. 2A, 3A and Table 1). The cells were weakly positive forthe lineage non-specific marker, CD38, but they were negativefor lineage-specific markers. The patient was diagnosed to haveAML-M1according to the FABclassification. He was treatedwith BHAC-DMPregimen (3) and entered into a completeremission (CR). The first relapse of his disease occurred after a9-month remission period. He entered into a 2nd CR by

reinduction therapy with BHAC-AMPregimen (4), but herelapsed soon after the 2nd consolidation therapy. His diseasedid not respond to further treatments and he died of pneumonia.At 2nd relapse, the leukemic cells in bone marrow (BM)aspirates (blasts 32%) were examined by FCM, because fewblasts were found in PB. The leukemic cells were found to bepositive forboth CD33 and CD13 antigens (95.6% and 67.4%).The CD38 antigen was detected in 97.2% of the blasts, but theother markers were negative (Fig. 2B, 3B and Table 1). Mor-phologically, the blast cells were large cells with a low nuclear/cytoplasmic ratio, large irregular nuclei, and basophilic cyto-plasm compared with those observed at diagnosis. Auer rodswere recognized, and MPOwas still strongly positive (morethan 90% of the blasts) (Fig. IB, D). The karyotype was also

From the Fourth Department of Internal Medicine, Tokai University School of Medicine, Isehara and *Blood Transfusion Center of Tokai University Hospital,Bohseidai, Isehara

Received for publication June 4, 1993; Accepted for publication September 9, 1993Reprint requests should be addressed to Dr. Hiroshi Kawada, the Fourth Department of Internal Medicine, Tokai University School of Medicine, Bohseidai,

Isehara, Kanagawa 259-1 1

Internal Medicine Vol. 32, No. 9 (September 1993) 733

Kawadaet al

Fig. 1. Morphological and cytochemical observations of the blasts at diagnosis (A, C) and at relapse (B, D). A, B: May-Giemsa staining of blasts(x250), C, D: cytochemical staining for myeloperoxidase (x250).

u

U

CO

CD

rs'(7)

Forward Scatter(A)

c3

c73

Forward Scatter(B)

Fig. 2. Cytogramsof flow cytometric studies. Gate settings used to analyze leukemiccells are shown. (A) at diagnosis (peripheral blood), (B) at relapse (bone marrow).

unchanged.

Discussion

We describe a case with MPO+MAg~AMLwho relapsedwith the acquisition of CD13and CD33expressions. His blast

cells were examined by using PB or BM; different CD13andCD33MoAbswere employed at the time of diagnosis andrelapse. However, CD33 MoAbs including My9 and Leu-M9MoAbsrecognize the same or closely related epitopes (5). Atleast five distinct epitopes of CD13antigen are shown, but theepitopes recognized by My7and Leu-M7 MoAbsseem to be

734 Internal Medicine Vol. 32, No. 9 (September 1993)

Phenotypic Change of MPO+CD1 3~CD33"AML

CO,-IQO

CD20

I

CD33

COH

8

CD3

co3o

CD10

Q

O

CD20

Fig. 3. Double immunofluorescence analysis ofleukemic cells. Blast cells were negative for CD13 and CD33 antigens, but arelatively small numberof cells were positive for CD38antigen at diagnosis (A). The blasts became positive for CD13 and CD33 antigen, and almost all of the cells were positivefor CD38 antigen at relapse (B).

almost identical (6). Our preliminary study using other AMLpatients also demonstrated similar reactivities of the two CD1 3MoAbs(unpublished observations). In addition, no significantdifference was found in the cell-surface marker expressions ofblast cells between the two different sources of samples (PB vs.BMaspirates) obtained from ten leukemia patients (data notshown). Furthermore, other AMLcases examined around thesame time by FCMusing the MoAbswere definitely positivefor CD13 and/or CD33 antigens. Thus, the discrepancies in thepercentages of CD13+ and CD33+cells examined at diagnosisand relapse are not explanable by the reactivity of MoAbsorsample sources employed.

Wedo not believe that this case represents a secondary

malignancy following cytotoxic chemotherapy because mostpatients with secondary leukemias have non-random chromo-somal abnormalities (7). Furthermore, Auer rods are reported tobe rare and MPOis weakly positive in the secondary AML.Atypical features indicating trilineage dyshematopoiesis werenot observed in the present case (8).The cell-surface antigens, i.e. CD13and CD33, are shown tobe expressed in very early stages ofmyeloid differentiation (9),whereas MPOis detectable at a relatively late stage of the

process (10). The minimally differentiated AMLwhich isnegative for MPOand expresses MAgwas recently classifiedas AML-MOby FAB group (1). In the present case, leukemiccells at diagnosis showed an unusual phenotype (MPO+CD1 3~CD33~) similar to the case reported by Seshi et al (2), but hisleukemic cells expressed both CD13and CD33antigens atrelapse. The commonfeatures between our case and their casewere as follows: male, no organomegalyor lymphadenopathy,high white blood cell count, high levels of serum lactatedehydrogenase, diagnosed as AML-M1, and no chromosomalabnorm ality.

In the present case, the percentages of CD38+leukemic cellswere relatively low at diagnosis, but those ofCD33+ and CD38+leukemic cells were high at relapse (more than 95%). The

patterns of surface phenotype were relevant to the evidencethat the expression of CD38 is followed by CD33 expressionin the myelomonocytic differentiation (1 1). Morphologicalfeatures of the blasts at relapse seemed to be more compatiblewith myeloblasts than those at diagnosis.The present case indicates that the MPO+CD13"CD33"leukemic cells can express typical AMLphenotypes undercertain circumstances. The mechanismof phenotypic changes

Internal Medicine Vol. 32, No. 9 (September 1993) 735

Kawadaet al

Table 1. Characteristics of Leukemic Cells

At diagnosis At relapse

Source of sample PB BM(% blast) (99) (32)

Karyotype 46XY(20/20) 46XY(20/20)

Cell-surface marker % Positive

CDla (0KT6) 0.5 N.D.

CD2 (Leu-5b) 1. 1 0.4

CD3 (Leu-4) 9.7 1.0

CD4 (Leu-3a) 0.9 0.6

CD5 (Leu- 1) N.D. 0.6CD7 (Leu-9) 0.9 1.3

CD8 (Leu-2a) 0.2 0.4

CD 1 0 (CALLA) 0.2 0.0

CDlib (Leu-15) 0.4 4.4

CD1 3 (*My7,Leu-M7) 1. 1 67.4CD14 (Leu-M3) 0.2 2.3

CD15 (Leu-Ml) N.D. 2.1

CD16 (Leu-1 lc) N.D. 0.2

CD19 (Leu-12) 0. 1 0.4

CD20 (Leu- 16) 0.4 0.4CD33 (**My9,Leu-M9) 1.2 95.6

CD34 (HPCA-1) N.D. 4. 1

CD36 (OKM5) 0.4 5.5

CD38 (Leu-17) 36. 1 97.2

CD41b (TP80) N.D. 1.4

HLA-DR 0.2 6.9

Glycophorin A N.D. 1.5Kappa chain N.D. 0.2Lambda chain N.D. 0. 1

*'**Although My7and My9 were used in the initial study, Leu-M7 and Leu-M9were employed at the time of relapse. PB: peripheral blood, BM: bonemarrow, ND: not done

might be related to asynchronous gene expression, clonalevolution or clonal change in leukemic cells.

References

1) Bennett JM, Catovsky D, Daniel M-T, et al. Proposal for the recognitionof minimally differentiated acute myeloid leukemia (AML-MO).Br J

Haematol 78: 325, 1991.

2) Seshi B, Kashyap A, BennettM. Acute myeloid leukemia with an unusualphenotype: Myeloperoxidase (+), CD1 3 (-), CD14 (-), and CD33 (-). BrJ Haematol 81: 374, 1992.

3) Kato S, Kawashima Y, Ohno R, YamadaI. Treatment of adult acutenonlymphocytic leukemia with BHAC-DMPtherapy. Acta Haematol

Jpn 46: 1077, 1983.

4) YamadaK. Aclarubicin in acute leukemia in adults. Gann Monogr CancerRes36: 121, 1989.

5) Peiper SC, Leboeuf RD, Hughes CB, et al. Report on the CD33 clusterworkshop : B iochemical and genetic characterization of gp67. in : LeukocyteTyping IV, Knapp W, Dorkin B, Gilks WR, et al. Eds. Oxford UniversityPress, Oxford, 1990, p. 814.

6) Favaloro EJ, Bradstock KF, Kabral A, Grimsley P, Zowtyj H, Zola H.Further characterization of human myeloid antigens (gpl60, 95; gpl50;

gp67): Investigation of epitopic heterogeneity and non-haemopoieticdistribution using panels ofmonoclonal antibodies belonging to CD- I lb,

CD-13 and CD-33. Br J Haematol 69: 163, 1988.

7) Rowley JD, Golomb HM, Vardiman JW. Nonrandom chromosomeabnormalities in acute leukemia and dysmyelopoietic syndromes in

patients with previously treated malignant diseases. Blood 58: 759, 198 1.8) Vardiman JW, Coelho A, Golomb HM, Rowley J. Morphologic andcytochemical observations on the overt leukemic phase of therapy-related

leukemia. AmJ Clin Pathol. 79: 525, 1983.9) Griffin JD, Davis R, Nelson DA, et al. Use of surface marker analysis to

predict outcome of adult acute myeloblastic leukaemia. Blood 68: 1232,1986.

10) Koeffler HP, Ranyard J, Pertcheck M. Myeloperoxidase: Its structure andexpression during myeloid differentiation. Blood 65: 484, 1985.

ll) Terstappen LWMM,Huang S, Safford M, Lansdorp PM, Loken MR.Sequential generations of hematopoietic colonies derived from singlenonlineage-committed CD34+CD38"progenitor cells. Blood 77: 1218,

1991.

736 Internal Medicine Vol. 32, No. 9 (September 1993)