GENES, CHROMOSOMES & CANCER 4:331-336 (1992)

Hyperdiploidy Arising From Near-Haploidy in Childhood Acute Lymphoblastic Leukemia Norio Onodera, Norah R. McCabe, James B. Nachman, F. Leonard Johnson, Michelle M. Le Beau, Janet D. Rowley, and Charles M. Rubin Departments of Pediatrics (N.O., N.R.M., J.B.N., F.L.J., C.M.R.), Medicine (M.M.L., J.D.R, C.M.R.), and Molecular Genetics and Cell Biology U.D.R.). Univenity of Chicago, Chicago, Illinois. and Children's Medical Center, lwate Prefectural Kitakarni Hospital, Kitakarni, Iwate, Japan (N.O.)

Acute lymphoblastic leukemia (ALL) of childhood is frequently characterized by a hyperdiploid karyotype. Typically, most of the affected chromosomes in the abnormal clone are present in three copies. W e have studied two patients with hyperdiploid ALL whose leukemic cells were atypical in that all or most of the chromosomes were present in either two o r four copies, raising a suspicion that the observed karyotype arose through duplication of chromosomes in a precursor cell with a near-haploid chromosome number. Analysis of restriction fragment length polymorphisms confirmed that both cases arose from a near-haploid cell; all informative disomic chromosomes tested had loss of heterozygosity. Furthermore, the hyperdiploid karyotypes did not arise via a perfect haploid cell with exactly 23 chromosomes, because tetrasomic chromosomes remained heterozygous. These two patients probably are classified best as near-haploid cases, which often are observed to have a co-existing hyperdiploid clone with a duplicated chromosome set. The distinction between typical hyperdiploidy and hyperdiploidy arisingvia a near-haploid cell may be clinically important, because the prognosis for patients with a hyperdiploid karyotype is favorable in comparison t o that of patients with a near-haploid karyotype. Genes Chrorn Cancer 4:33 1-336 (I992). @

1992 Wiley-Liss, Inc.

INTRODUCTION

A hyperdiploid karyotype is observed in 23 to 39% of newly diagnosed children with acute lymphoblastic leukemia (ALL) (Kaneko et al., 1982; Heerema et al., 1985; Prigogina et al., 1988; Pui et al., 1988; Fletcher et al., 1989 Uckun et al., 1989). These patients are generally divided into subgroups, namely, those with a chromosome number of 4 7 4 9 and those with 2 50 chromosomes. The hyperdiploid 1 50 group (14 to 27% of childhood ALL) has a clear pattern of chromosomal gain. Nearly all of the patients have a chromosome number of 5 1 4 5 with a peak at 55 (Pui et al., 1989). Very few patients with 50 chromosomes have been described (Williams et al., 1982). Frequently gained chromosomes (in more than half of the cases) are chromosomes 4,6,10,14,17,18,20,21, and X (Williamset al., 1986),and typically the affected chromosomes are present in three copies. Chromosome 21 is consistently reported as the most frequently gained chromosome (Third International Workshop on Chromosomes in Leukemia, 1981; Heerema et al., 1985; Williams et al., 1986; Prigogina et al., 1988) and is often observed in multiple extra copies. The presence of hyperdiploidy ? 50 correlates strongly with good risk features, including age between 3 and 7 years, white blood cell count less than 10 x 109/L,French-American-British(FAB) Cooperative Group subtype L1 (Kaneko et al., 1982)and com0 1992 WILEY-LISS, INC.

mon ALL antigen (CALLA, CD10)-positiveearly preB-cell phenotype (Pui et al., 1988). Moreover, patients with hyperdiploidy ? 50 have the longest diseasefree survival compared with any other cytogenetic group (Bloomfield et al., 1989). A near-haploidkaryotype is a relatively rare occurrence in childhood ALL. The presenting features in this group including age, white blood cell counts, and immunophenotype are similar to those of childhood ALL in general (Pui et al., 1990).However, recognition of near-haploidcases is important, because the available data suggest that these patients have a relatively high risk of leukemic relapse (Brodeur et al., 1981;Pui et al., 1987, 1990; Callen et al., 1989).In patients with a near-haploid clone, a second hyperdiploid clone is often observed. Cytogenetic evidence favors the view that the hyperdiploid clone results from duplication of the chromosomes in the near-haploid line. We have studied two patients with hyperdiploid karyotypes, who were atypical in that all or most chromosomes were present in either two or four copies; the trisomies usually seen in hyperdiploid ALL were not observed. We show by an analysis of restriction fragment length polymorphismsthat in these two cases a near-haploid cell was the precursor to the Received September 23, 1991; accepted December 2, 1991. Address reprint requests to Dr. Charles M. Rubin, Department of Pediatrics, University of Chicago, 5841 S. Maryland Ave., MC4060, Chicago, IL 60637.

332

ONODERA ET A L

hyperdiploid cell. These two patients probably are classified best as near-haploid cases. METHODS Patients

Patient 1 was a 5-year-old black female who presented with a 1-month history of fever. Her white blood cell count was 33,200/mm3with 85% blasts. A bone marrow examination revealed 90% lymphoblasts and FAB subtype L1. Immunophenotyping demonstrated an early B lineage cell type. She was treated with Children’s Cancer Study Group protocol 105 regimen 2A, which is designed for patients with an intermediate prognosis. The protocol includes intensive induction and consolidation with delayed intensification (Gaynon et al., 1987). She completed 30 months of this therapy and has been in continuous complete remission for 43 months. Patient 2 was an 18-year-oldwhite male who presented with several days of abdominal pain. The white blood cell count was 3,100/mm3 with 14% blasts. A bone marrow examination revealed 83% lymphoblasts and FAB subtype L1. Immunophenotyping demonstrated an early B lineage cell type. He also was treated with Children’s Cancer Study Group protocol 105 regimen 2A. He relapsed 17 months after diagnosis while on maintenance therapy; the bone marrow had 61% blasts at that time. A second remission was achieved. He subsequently relapsed again and died of progressive disease 28 months after diagnosis. Cytogenetic Analysis

Cytogenetic analysis was performed using a trypsin-Giemsa banding technique. Cells were obtained from bone marrow and/or peripheral blood before initiation of therapy. Metaphase cells from direct preparations and/or short-term (24 and 48 hour) unstimulated cultures were examined. Chromosomal abnormalities were described according to the International System for Human Cytogenetic Nomenclature (TSCN, 1985). Molecular Analysis

DNA was extracted from cryopreserved buffy coat cells from bone marrow or mononuclear cells from peripheral blood. Isolation of DNA, restriction endonuclease digestion, electrophoresis, Southern blotting, radiolabeling of probes, and DNA hybridization were performed according to standard procedures. Probes

The probes used to detect DNA restriction fragment length polymorphisms are described in Table 2.

Additional probe information and references are listed by the 10th International Workshop on Human Gene Mapping (Kidd et al., 1989). Probes p2.1, p21-4U, and pPW511-1H and pPW228C were gifts from Dr. Ellen Solomon, Dr. Gordon D. Stewart, and Integrated Genetics (Framingham, MA), respectively. Probe CRIL427 was purchased from CollaborativeResearch Inc. (Bedford, MA). The remaining probes were obtained from the American Type Culture Collection (Rockville, MD); cYNA13, pYNH24, pEFD64.1, pJCZ67, and pMHZl0 were deposited by Drs. Ray White and Yusuke Nakamura, and pCMW-1, p79-2-23 and 2% were deposited by Drs. David Ledbetter, Michael Litt, and A. Millington-Ward, respectively. RESULTS Karyotypes

Karyotypes are presented in Table 1. Patient 1had 50 chromosomes in the abnormal clone at diagnosis due to double gains of chromosomes 18 and 21. Patient 2 had two abnormal clones detected at diagnosis. One clone was near-haploid with 30 chromosomes, and the other had 60 chromosomes and represented a nearly precise doubling of the near-haploid set of chromosomes. At relapse, the near-haploid clone was not observed in Patient 2; however, the single abnormal clone had 60 chromosomes and was almost identical to the previously observed hyperdiploid clone. Analysis of Restriction Fragment Length Polyrnorphisrns

We compared nonleukemic and leukemic tissue from each patient. The mononuclear fraction of the peripheral blood obtained during remission from Patient l , and cryopreserved buffy coat cells from a bone marrow aspirate obtained during remission from Patient 2 were used as constitutional nonleukemic tissue. Bone marrow obtained at diagnosis from Patient 1 and at relapse from Patient 2 was used as leukemic tissue. We examined DNA polymorphisms on chromosomes 1, 2, 3, 7, 11, 15, and 16, each of which was present in two copies (disomic) in the hyperdiploid leukemic clones of both patients. We also examined DNA polymorphisms on chromosome 9, which was disomic in patient 1 and tetrasomic in patient 2, and chromosome 21, which was tetrasomic in both patients. The probes listed in Table 2 were used. Probes for chromosomes 1,2,7,9,11,15,16,and 21 were informativefor Patient 1,and probes for chromosomes 1, 2, 3, 9, 11, 16, and 21 were informative for Patient 2. Informative probes were those revealing two different alleles in the nonleukemic tissue.

333

HYPERDIPLOID ALL

TABLE I. Cytogenetic Studies of Two Patients With Acute Lymphoblastic Leukemia and a Hyperdiploid Karyotype State of Disease

Source

No. Metaphase Cells Examined

Karyotype

I

Diagnosis

Bone marrow and blood

32

2

Remission Diagnosis

Bone marrow Bone marrow

20 40

46,XX(8 I %)/50,XX, + 18, + 18, + 2 I , +21 (19%) 46,XX( 100%) 46,XY(72%)/3O,XY, + 5, 8,+ 9, 14,+ 19.+21 (13%)/60,XY,+X, +Y,+5,+5,+8,+8,+9,+9, + 14,+ 14,+ IS,+ 19.+21,+21

Relapse

Bone marrow

21

(15%) 46,XY(90%)/60,XY, +8,+8,+9,+9,+

Remission

Bone Marrow

21

20, 2 I, + 2 I (10%) 46,XY( 100%)

Patient

+

+

14,+ + X, + Y,14,+ + 5, 19, + 5,

+ +

TABLE 2. Analysis of D N A Polymorphisms in Two Patients With Acute Lymphoblastic Leukemia and a Hyperdiploid Karyotype Probes Chromosomal Location

Iq 2P 3 7q 9q I lp13-pI5 I5pter-q I 3 I6q24 2 Ipter-q2 I. I 21ql I 21ql I 21ql 1.2-q21 2 Iq22.3

Locus Symbol

Name

PIC Value

D IS74 D2S44 D2S42 D7S396 D9SI I D I IS150 D I5S24 D I6S7 D2 IS26 D2lSllO D2 IS52 D21SI D21SI 12

cYNA I 3 pYNH24 pEFD64. I pJCZ67 pMHZ I0 p2. I pCMW- I p79-2-23 26C p2 I-4u pPW51 I-IH pPW228C CRLL427

0.96b 0.97b 0.85b 0.80b 0.72b 0.90b 0.85b 0.7Eb 0.56 0.30 0.25 0.36 0.93b

Loss of Heterozygosity in Leukemic DNA” Patient I

Patient 2

Yes Yes

Yes Yes Yes

Yes Yes Yes Yes Yes No

No No

-

No Yes Yes No No No No No

aMissing data (-) indicates that the probe was not informative in that patient. bProbe is of the VNTR (variable number of tandem repeat) type.

For both patients all informative probes for disomic chromosomes revealed a loss of heterozygosity in the leukemic tissue, i.e., a loss of one of the two alleles observed in the non-leukemic DNA (Table 2, Fig. 1). In some instances the leukemic DNA had one band of full intensity and another of weak intensity; the weak band represents residual normal cells. We studied several examples of tetrasomic chromosomes, namely, chromosome 21 in both patients using five probes from various positions on the chromosome, and chromosome 9 in Patient 2. In contrast to the results obtained for the disomic chromosomes, all informative probes demonstrated two polymorphic bands of equal intensity in leukemic tissue indicating maintainance of heterozygosity (Table 2, Fig. 2).

DISCUSSION

We had clues from the cytogenetic analysis that the two patients studied here did not have typical hyperdiploidy, despite counts of 50 or 60 chromosomes per cell. Patient 1 had 50 chromosomes due to double gains of chromosomes 18 and 21. This suggested the possibility that the hyperdiploid clones arose from doubling of chromosomes from a near-haploid cell with the karyotype 25,X, + 18,+ 21. For Patient 2 we were more certain of this sequence of cytogenetic evolution, because near-haploid and hyperdiploid cells had been identified in a sample obtained at diagnosis, although only the hyperdiploid clone was observed at relapse. Nearly all chromosomes were disomic or tetrasomic in the hyperdiploid cells.

334

ONODERA ET A L

A similar patient has been reported in whom one chromosome was studied by analysis of a chromosomal polymorphism (Stamberg et al., 1986). The patient’s karyotype was 52,XX, + 8, + 8, + 10,+ 10,+ 21, +21; thus all chromosomes were present in two or four copies, and a hypothetical intermediate cell may have had the near-haploid karyotype 26,X, + 8, + 10, +21. In the leukemic cells it was observed that chromosome 15 had become homozygous for a polymorphism on the short arm. In as many as 90% of near-haploid cases, there is a second abnormal clone with a hyperdiploid karyotype that represents a doubling of the chromosomes in the near-haploid cell line (Oshimura et al., 1977; Shabtai et al., 1979; Kaneko et al., 1980; Brodeur et al., 1981; Misawa et al., 1985; Callen et al., 1989; Pui et al., 1990); this was observed in Patient 2 at diagnosis. The evidence for this interpretation is that the hyperdiploid clone usually contains exactly two copies of all chromosomes in the near-haploid clone, including chromosomes with polymorphisms and structural abnormalities. The near-haploid clone contains at least one copy of each chromosome plus a second copy of

Fig. I. Restriction fragment length polymorphism analysis of chromosome 2 in leukemic cells (lanes a) and nonleukemic cells (lanes b) from Patients I and 2. D N A was digested with Mspl and hybridized with probe pYNH24. which recognizes a VNTR polymorphism on chromosome 2. The faint bands in the lanes containing D N A from leukemic cells represent residual normal bane marrow cells present in the samples.

For both patients we were able to confirm that the hyperdiploid clones arose from near-haploid cells. In all informative chromosomes tested that were present in two copies, loss of heterozygosity was present; this observation is consistent with a uniparental origin for the two chromosomal homologs. The widespread loss of heterozygosity is most consistent with development of haploidy or near-haploidy in a cell, followed by duplication of the chromosomes. Thus our patients probably are classified best with near-haploid cases. We have studied 15 patients in total with hyperdiploid ALL using multiple restriction fragment length polymorphisms, and only these two patients had widespread loss of heterozygosity (N. Onodera, N.R. McCabe, C.M. Rubin, unpublished observations). but significant incidence of Thus there is a near-haploid cases presenting as hyperdiploid cases.

Fig. 2. Restriction fragment length polymorphism analysis of chromosome 2 I in leukemic cells (lanes a) and nonleukemic cells (lanes b) from Patients I and 2. D N A was digested with Mspl and hybridized with probe pPW228C.

335

HYPERDIPLOID ALL

certain chromosomes. The additional chromosomes above the haploid set are nonrandom. Two copies of chromosome 21 and two sex chromosomes are present in 80-90% of cases. Other common disomies in the near-haploid clone involve chromosomes 18 (65% of cases), 10 (&YO), 14 (45%), 6 (IS%), and 8 (15%) (Callen et al., 1989). Our results also show that a perfect haploid cell with exactly 23 chromosomes does not occur as an intermediate step in the development of the near-haploid or corresponding hyperdiploid clone. In chromosomes present in four copies in the hyperdiploid cells, the parental contributions were equal in dosage. If a perfectly haploid cell was a precursor to the nearhaploid and hyperdiploid cells, then the chromosomes present in two copies in the near-haploid cell or in four copies in the hyperdiploid cell would be of a single parental origin; based on our results, this is not the case. Absence of a perfectly haploid precursor cell is also obvious in males with near-haploid karyotypes in which both the X and the Y chromosomes often are retained; the karyotype for Patient 2 at diagnosis is an example. The mechanisms leading to a near-haploid karyotype are unknown. Also unknown is the relationship between a near-haploid karyotype and the pathophysiology of the leukemic process. It is possible that one or more recessive mutant alleles of tumor suppressor genes are unmasked as a result of near-haploidy. As discussed above, patients with near-haploid karyotypes often have hyperdiploid clones. In some of these patients, such as Patient 1 at diagnosis and Patient 2 at relapse, only hyperdiploid cells are observed. It is important to distinguish these cases from the more typical hyperdiploid patients, because the prognosis for these two groups is very different. This may be accomplished by one of several methods. Extensive cytogenetic analysis in which large numbers of metaphase cells are analyzed may reveal the nearhaploid cells. DNA flow cytometry, which assesses metaphase and interphase cells, may also detect a near-haploid population of cells (Pui et al., 1990). Finally, demonstration of widespread loss of heterozygosity by an analysis of DNA polymorphisms may point to the near-haploid origin of the hyperdiploid cells. ACKNOWLEDGMENTS

This work was supported in part by grant CA42557 0.D.R.) from the National Institutes of Health and grant DE-FG02-86ER60408 U.D.R.) from the Department of Energy. M.M.L. is a Scholar of

the Leukemia Society of America, and C.M.R. is a Pew Scholar in the Biomedical Sciences. We thank Dr. Manuel 0. Diaz for helpful discussions; Drs. Ellen Solomon and Gordon D. Stewart for the gifts of probes p2.1 and p21-4U, respectively; and Integrated Genetics (Framingham, MA) for the gift of probes pPW511-1H and pPW228C. We also thank Peter Rubinelli for expert technical assistance in the molecular analysis. REFERENCES Bloomfield CD, Seeker-WalkerLM, Goldman AI, Van Den Berghe H, de la Chapelle A, Ruutu T, Alimena G, Garson OM, Golomb HM, Rowley JD, Kaneko Y, Whang-Peng J, Prigogina E. Philip P, Sandberg AA, Lawler SD, Mitelman F (1989) Six-year follow-up of the clinical significance of karyotype in acute lymphoblastic leukemia. Cancer Genet Cytogenet 40171-185. Brodeur GM, Williams DL, Look AT, Bowman WP, Kalwinsky DK (1981) Near-haploid acute lymphoblastic leukemia: A unique subgroup with a poor prognosis? Blood 5814-19. Callen DF, Raphawl K, Michael PM, Garson OM (1989) Acute lymphoblastic leukemia with a hypcdiploid karyotype with less than 40 chromosomes:The basis for division into two subgroups. Leukemia 3:74%752. Fletcher JA, Kimball VM, Lynch E, Donnelly M, Pavelka K, Gelder RD, Tantravahi R, Sallan SE (1989) Prognostic implications of cytogenetic studies in an intensively treated group of children with acute lymphoblastic leukemia. Blood 74213C2135. Gaynon PS,Steinherz PG, Reaman GH, Bleyer WA, Sather H, Hammond GD (1987) Strategies for treatment of children with acute lymphoblastic leukemia and unfavorable presenting features. Haematol Blood Transfusion 30167-172. Heerema NA, Palmer CG, Baehner RL (1985) Karyotypic and clinical findings in a consecutive series of children with acute lymphocytic leukemia. Cancer Genet Cytogenet 17:165179. ISCN (1985) An International System for Human Cytogenetic Nonmenclature. Hamden DG, Klinger HI?,eds. Published in collaboration with Cytogenet Cell Genet, Karger, Basel; also in Birth Defects: Original Article Series, Vol. 21, No. 1. New York March of Dimes Birth Defects Foundation. Kaneko Y, Sakurai M (1980) Acute lymphocytic leukemia (ALL) with near-haploidy: A unique subgroup of ALL? Cancer Genet Cytogenet 21S18. Kaneko Y, Rowley JD.Variakojis D, Chilcote RR, Check I, Sakurai M (1982) Correlation of karyotype with clinical features in acute lyrn phoblastic leukemia. Cancer Res 422918-2929. Kidd KK, Bnwcock AM, Schmidtke J, Track RK, Riccinti F, Hutching G, Bale A, Pearson P, Willard HF (1989) Report of the DNA committee and catalogs of cloned and mapped genes and DNA polymorphisms. Cytogenet Cell Genet 513522-947. Misawa S, Oguma N, Testa JR (1985) A case of acute lymphoblastic leukemia with severe hypcdiploidy. Cancer Genet Cytogenet 16137143. Oshimura M, Freeman AI, Sandberg AA (1977) Chromosomes and causation of human cancer and leukemia. XXIII. Near-haploidy in acute leukemia. Cancer 4011&1148. Prigogina EL, Puchkova GP, Mayakova SA (1988) Nonrandom chromosomal abnormalities in acute lymphoblastic leukemia of childhood. Cancer Genet Cytogenet 3218S-203. hi CH, Williams DL, Raimondi SC, Rivera GK, Look T, Dodge RK, George SL, Behm FG, Crist WM, Murphy SB (1987) Hypodiploidy is associated with a poor prognosis in childhood acute lymphoblastic leukemia. Blood 70247-253. Pui CH, Williams DL, Roberson PK, Raimondi SC, Behm FG, Lewis SH, Rivera GK, Kalwinsky DK, Abromowitch M, Crist WM, Murphy SB (1988) Correlation of karyotype and immunophenotype in childhood acute lymphoblastic leukemia. J Clin Oncol 6 : W l . Pui CH, Raimondi SC, Dodge RK, Rivera GK, Fuchs LAH, Abromowitch M, Look TA, Furman WL. Crist WM, Williams DL (1989)Prognostic importance of structural chromosomal abnormalities in children with hyperdiploid ( > 50 chromosomes) acute lymphoblastic leukemia. Blood 731%5%1967. Pui CH, Carroll AJ, Raimondi SC, Land VJ, Crist WM, Schuste JJ, Williams DL. Pullen DJ, Borowitz MJ, Behm FG, Look AT (1990) Clini-

336

ONODERA ET A L

cal presentation, karyotypic characterization, and treatment outcome of childhood acute lymphoblastic leukemia with near-haploid or hypodiploid < 45 line. Blood 75:117&1177. Shabtai F, Lewinski UH, Har-Zahav L, Gafter U. Halbrecht I, Djaldeti M (1979) A hypodiploid clone and its duplicate in acute lymphoblastic leukemia. J Clin Pathol 721018-1024. Stamberg J, Shende A, Lanzkowsky P (1986) Somatic shift to homozygasity for a chromosomal polymorphism in a child with acute lymphoblastic leukemia. Blood 67:350-353. The Third International Workshop on Chromosomes in Leukemia (1981) Cancer Genet Cytogenet 495142.

Uckun FM, Gajl-Peczalska KJ, Provisor AJ, Heerema NA (1989) Immunophenotype-karyotype associations in human acute lymphoblastic leukemia. Blood 73271-280. Williams DL, Tsiatis A, Brcdeur GM, Look AT, Melvin SL, Bowman WP, Kalvinsky DK, Rivera G, Dahl GV (1982) Prognostic importance of chromosome number in 136 untreated children with acute lymphoblastic leukemia. Blood 60:864-871. Williams DL, Look AT, Harber J, Murphy SB. Kalwinsky DK, Rivera GK, Melvin SL, Stass S. Dahl GV (1986) Chromosomal kanslocations play a unique role in influencing prognosis in childhood acute lymphoblastic leukemia. Blood 68205212.