Comparative genomic hybridization in childhood acute lymphoblastic leukemia

Leukemia (1998) 12, 1638–1644  1998 Stockton Press All rights reserved 0887-6924/98 $12.00 http://www.stockton-press.co.uk/leu

BIO-TECHNICAL METHODS SECTION (BTS)

BTS Leukemia

Comparative genomic hybridization in childhood acute lymphoblastic leukemia ML Larramendy1,2, T Huhta1,2, K Vettenranta3, W El-Rifai1,2, J Lundin4, S Pakkala5,6, UM Saarinen-Pihkala3 and S Knuutila1,2 1

Department of Medical Genetics, Haartman Institute, University of Helsinki; 2Laboratory of Medical Genetics, 3Division of HematologyOncology and Stem Cell Transplantation, Hospital for Children and Adolescents, 4Clinical Research Institute, 5Transplantation Laboratory, and 6Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland

DNA copy number changes were studied by comparative genomic hybridization (CGH) on bone marrow samples obtained from 72 patients with childhood acute lymphoblastic leukemia (ALL) at diagnosis. The patients had been admitted to the Helsinki University Central Hospital (Finland) between 1982 and 1997. CGH showed DNA copy number changes in 45 patients (62.5%) with a mean of 4.6 aberrations per patient (range, 1 to 22). The results of CGH and chromosome banding analysis were generally concordant, but CGH facilitated specific karyotyping in 34 cases. DNA copy number gains were more frequent than losses (gains:losses, 6:1). Gains of DNA sequences affected almost exclusively whole chromosomes and were most commonly observed in chromosomes 21 (25%), 18 (22.2%), X (19.4%), 10 (19.4%) and 17 (19.4%). The most common partial gain was 1q31-q32 (8.3%). The most common gains of chromosomes 21, 18, X, 10, 17, 14, 4, 6 and 8 appeared concurrently. High-level amplifications of small chromosome regions were sporadic, detected only in two patients (2.8%). Chromosome 21 was involved in both cases. The most common losses were 9p22-pter (12.5%) and 12p13-pter (11.1%). No statistically significant association between the CGH findings and the diagnostic white blood cell count was observed. Keywords: comparative genomic hybridization; acute lymphoblastic leukemia; DNA copy number changes; gains; losses

Introduction Cytogenetic evaluation of childhood lymphoblastic leukemia (ALL) is of immediate clinical importance. Hyperdiploidy (modal chromosome number over 50) as well as some structural aberrations (eg t(12;21))1 appear to be associated with a favorable outcome,2–4 whereas some (eg t(9;22), t(4;11)) predict a high likelihood of failure with conventional therapy.3–5 Among patients with childhood ALL a comprehensive cytogenetic analysis of the malignant clone(s) remains problematic despite progress in methodology.6–8 Consequently, it has not been possible to reliably evaluate the prognostic and/or diagnostic significance that many of the chromosomal aberrations detected in lymphoid blasts may have. Not only the analysis of solid tumors, but also the study of hematologic neoplasms can benefit from comparative genomic hybridization (CGH)9–13 which may solve some of the problems encountered using standard cytogenetics. No cell Correspondence: S Knuutila, Laboratory of Medical Genetics, Helsinki University Central Hospital, PO Box 404 (Haartmanink 3, 4th floor), FIN-00029 HUCH, Helsinki, Finland; Fax: 358 9 1912 6788 The first two authors contributed equally to this work Received 9 February 1998; accepted 26 May 1998

cultures or mitotic cells are needed for CGH. Information of DNA copy number changes can be obtained even in cases with poor chromosome morphology or marker chromosomes. Importantly, unlike conventional cytogenetic analysis, CGH makes it possible to screen chromosome areas containing high-level gene amplifications that so far have been poorly evaluated.14,15 The aim of this study was to perform a comprehensive CGH analysis on patients with childhood ALL in a single institution setting. We report the DNA copy number changes detected in 72 pediatric ALL patients at diagnosis.

Materials and methods

Patients The study was performed on a comprehensive series of 72 diagnostic (ie prior to the initiation of chemotherapy) bone marrow samples from patients with childhood ALL. All patients were diagnosed and treated between 1982 and 1997 in the Hospital for Children and Adolescents at the Helsinki University Central Hospital. The diagnosis was based on morphological evaluation of bone marrow aspirates and biopsies, as well as on flow cytometric analysis of the marrow blast population. Some clinical parameters of the patients are summarized in Table 1.

Conventional cytogenetic analysis Standard chromosome banding analysis was performed on bone marrow cells after short-term culture.16 Karyotype abnormalities were described according to the specifications proposed by the International Standing Committee on Human Cytogenetic Nomenclature.17 The karyotype data included in this study was retrieved from clinical records.

Comparative genomic hybridization CGH was performed using direct fluorochrome-conjugated DNAs for all samples according to methods described previously with minor modifications.9,10 Briefly, blast DNA and reference DNA (genomic DNA from peripheral blood lymphocytes from normal donors) were labeled with fluorescein-iso-

Bio-technical methods section (BTS) ML Larramendy et al

Table 1 Clinical, laboratory and karyotypic characteristics and DNA sequence copy number changes at diagnosis from 72 patients with acute lymphoblastic leukemia

Patient No. (sex/age)a

Lab code Phenotype

WBC (×109/l)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

(M, 5.9) (M, 12.9) (F, 7.5) (M, 8.1) (M, 1.9) (F, 5.7) (M, 7.7) (F, 11.7) (M, 4.6) (F, 8.4) (M, 15.7) (F, 4.2) (F, 8.4) (F, 4.1) (F, 3.0) (M, 3.7) (M, 6.0) (M, 13.8) (M, 15.7) (M, 15.4) (F, 7.5) (M, 0.15) (F, 5.0) (M, 2.5)

950769 890694 950150 960759 950243 920463 871252 920029 910973 910721 870060 910266 930442 920805 930297 920183 920079 900988 920200 910553 950716 890380 930105 880936

T cell pre B pre B pre B pre B pre B pre B B cell pre B B cell B cell T cell? pre B pre B pre B pre B pre B pre B pre B pre B pre B pre B pre B —

25 26 27 28 29 30 31 32

(M, 15.7) (F, 13.8) (M, 9.1) (F, 4.8) (F, 8.5) (F, 3.7) (M, 14.2) (F, 0.9)

910371 960861 960605 960096 821150 960476 970797

pre B pre B T cell pre B pre B pre B pre B pre B

33 (F, 8.2)

870438

pre B

34 (F, 11.2)

920564

pre B

35 (F, 2.3)

970041

pre B

36 (M, 4.8) 37 (M, 10.3)

950434 900100

pre B pre B

38 (M, 6.6)

950153

pre B

6.3

39 (F, 2.0)

970431

pre B

51.2

40 (M, 2.0)

960459

pre B

41 (F, 1.8) 42 (F, 4.1) 43 (F, 11.7)

950369 910943 900320

pre B pre B pre B

99.2 4.8 5.8

44 (M, 5.7) 45 (M, 5.8)

950455 900856

pre B pre B

5.5 34.1

46 (F, 14.3)

930419

B cell

21

47 (F, 2.4)

920813

pre B

14.9

48 (F, 12.1) 49 (M, 3.0)

960105 910192

pre B pre B

62.7 13.2

50 (F, 4.0)

950294

pre B

4.3

Karyotype in bone marrow cellsb

15 11.3 165 36.3 4.7 19.8 5 11.4 36.1 112 8.5 2.1 5.6 12.4 6.8 21.6 6.4 113 50.6 20.3 4.1 193 1.7 6.7

46,XY[12] 46,XY[13] 46,XX[6] 46,XY[14] 46,XY[12] 46,XX[6] 46,XY[3] 46,XX[14] 46,XY[13] 46,XX[2] 46,XY[20] 46,XX[18] 46,XX[7] 46,XX[2] 46,XX[4] 46,XY[18] 46,XY[20] 46,XY[12] 46,XY[12] 46,XY[10] 46,XX[5] 46,XY[7] 46,XX[19] 46,XY,t(11;12)(q13;q24)[7] /46,XY[8] 19.3 46,XY,t(4;11)(q21;q?24)[8] 149.4 47,XX,+6[2]/46,XX[11] 4.8 − 4.4 − 10 46,−C,+D,−E+mar,inc[14]/46,XX[1] 21.8 − 13.2 − 235 45,XX,der(7;9)(q10;q10)[4] /46,XX[15] 48.9 44−47,+G,+mar,inc[cp18]/ 46,XX[15] 2.4 55−58,XX,+B,+?6,+C,+E,+G, inc[cp3]/46,XX[2] 3.9 53−55,XX,+B,+?9,+C,+C,+E,inc [cp8]/46,XX[4] 21.3 46,XY[11] 4.9 45−46,+mar,inc[cp3]/46,XY[3]

238

45−46,XY,−12,+1−2mar,inc[cp3]/ 46,XY[2] 45,XX,der(8)t(8;15)(q?;q?),del(9) (p22),del(12)(p12),del(13)(q14), −15[9]/46,XX[1] 46,XY[17]

DNA copy number changesc Losses

Gains

− − − − − − − − − − − − − − − − − − − − − − − −

− − − − − − − − − − − − − − − − − − − − − − −

NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NN NB

− − − 13q21−qter 9p21−pter 9p 12p 7p, 9p

− − − − − − − −

NB NA N? LO LO LO LO LO

9pter−q22

21

LG

9p22-pter

4, 6, 10, 11, 12, 14, 17, 18, 21, 22 4, 8, 9, 10, 14, 18, X

LG LG

17q, Xq23−qter 7p15−pter

LG LG

Xq25−q26

LG

15q13−q15

LG

12p11

LG

10p12-pter 10, 17q, X 8q22−qter, 21

LG LG LG

Xq23−q26 1q24-qter

LG GP

1q24−qter

GP

1q24−qter

GP

5q31−qter 16, Xq21.3−qter

GP GP

15q11−q15, 21, X

GP

13q21−qtr Xpter−q13 11q21−qter, 12p13−pter 6q24−qter, 12p

9p22−pter, 12p12−pter, 13q14−qter 9p21−pter, 12p13−pter, 17p 47−48,+2−4mar,inc[cp5]/46,XX[7] 12p 49,+C,+C,inc[2]/46,XX[2] 9p 44−46,−C,+G,inc[cp4]/47,XX, 9p, 12p +21c[3] 46,XY[7] 6q, 8p, 12p, Xp 46,XY,der(19)t(1;19)(q2;p1)[9]/ − 46,XY[3] 46,XX,del(3)(q?26),add(9)(p?21), − der(19)t(1;19)(q23;p13)[8] 46,XX,der(19)t(1;19)(q2;p1)[6]/ − 46,XX[4] 45−46,XX,−5,+1−2mar[8]/46,XX[3] − 46,XY,del(3)(?q24),inc[cp10]/ − 46,XY[10] 48−50,XX,+?X,+G,inc[cp10]/ − 46,XX[6]

Category

(Continued)

1639

Bio-technical methods section (BTS) ML Larramendy et al

1640

Table 1

Continued

51 (M, 3.1) 52 (M, 4.7)

920970 910014

pre B pre B

53 (F, 9.7)

840776

T cell

54 (F, 3.2)*

901164

pre B

65.8

55 (F, 11.8)

950223

pre B

14.9

56 57 58 59 60 61 62

(F, 2.7) (F, 6.2) (M, 0.6) (M, 2.8) (F, 4.9) (F, 5.4) (M, 5.9)

900380 900585 900987 890606 821096 900827 910047

pre B B cell pre B pre B pre B pre B pre B

153 4.4 140 181 39.7 13.6 9.3

63 (M, 1.7)

900521

pre B

10.6

64 (M, 2.2) 65 (F, 1.1)

960697 890528

pre B pre B

27.7 5.6

66 (M, 3.8)*

880813

pre B

17.1

67 (M, 1.9)

910762

pre B

4.2

68 (F, 8.0)

901253

pre B

2.6

69 (F, 2.4)

970148

pre B

2.4

70 (M, 2.4)

970092

pre B

47.8

71 (F, 13.7)* 950346

pre B

4.5

72 (F, 4.0)*

pre B

20.2

930400

6.2 7.6 176

42−47,+mar,inc[cp15]/46,XY[2] 54−55,XY,+X,+4,+6,+?8,+?10, +15,+?20,+mar[cp7] 46,XX,add(19)(p1)[3]/46,XX, t(8;14)(q24;q11)[2]/46,X,−X, +mar[7]/46,XX[6] 64−66,XX,+1,+3,+4,+5,+6,+8,+10, +?10,+11,+12,+16,+17,+18, +?18,+19,+21,+22,+mar,inc[11]/ 46,XX[9] 47,XX,+5,t(9;22)(q34;q11)[10]/ 46,XX[5] − 46,XX[18] 47,XY,+X[7]/46,XY[9] 48,XY,+10,+21[18] 47−50,+G,inc[cp10] 52−53,XX,+X,+mar,inc[cp20] 56,XY,+X,+8,+8,+10,+10,+18, +19,+20,+21,+22[10]/46,XY[10] 54−56,XY,+X,+6,+?8,+14,+17, +18,+20,+21,+21,+mar[cp8] 48−51,XY,+?4,inc[cp2]/46,XY[7] 55,XX,+X,+4,+6,+?10,+17,+18, +21,+21,+mar[5]/46,XX[6] 60−61,XY,+X,+3,+4,+6,+?8, +?9,+14,+16,+17,+18,+21,+21, +3mar[cp10]/46,XY[7] 51−58,XY,+X,+6,+8,+9,+10,+13, +15,+18,+21[cp10]/46,XY[5] 56−59,XX,+?4,+6,+?10,+?17, +?18,+?21,inc[cp3]/46,XX[2] 46,XX[8]

− −

9p13−qter, 17, 21, 22 1q31−q32, 4, 6, 8, 10, 15, 17, 18, X 1q21−q23, 6

−

GP GP GP

−

1q21−qter,2pter−q22, 3, 4, 5, 6, 8, 10, 11, 12, 14, 16, 17, 18, 21, 22

GP

−

5

GS

− − − −

21 14 X 10, 21 15, 21 (q22−qter) 4, 14, 17, 18, X 8, 10, 18, 21, X

GS GS GS GS GS GM GM

−

6, 14, 17, 18, 21

GM

− −

4, 6, 14, 17, 18, 21, X 4, 6, 10, 17, 18, 21, X

GM GM

−

3, 6, 8, 14, 17, 18, 21, X

GM

−

6, 7, 8, 10, 14, 17, 18, 21 4, 6, 10, 14, 17, 18, 21, X 4, 5, 8, 10, 11, 12q, 14, 18, 21 4, 6, 8, 10 15, 17, 18, 21, X 4, 6, 8, 9, 10, 14, 17, 18, X 1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, X

GM

− −

−

− −

57−58,XY,+4,+6,+10,+15,+17, − +18,+21,+21,inc[cp4] 54−55,XX,+4,+?8,+?8,+5−6mar[cp2]− /46,XX[6] 50−67,XX,+C,+C,+C,+E,+G,inc − [cp3]/46,XX[9]

GM GM GM GM GM

a

Age at diagnosis in years; *for CGH analysis, confidence intervals for ratio profiles corrected according to ploidy level. − = no evaluable metaphases were obtained. c High-level amplifications in bold. NN, normal CGH, normal karyotype; NB, normal CGH, balanced translocation only; NA, normal CGH, abnormal karyotype; N?, normal CGH, no karyotype data available; LO, only loss/es in CGH; LG, loss/es and gain/s in CGH; GP, partial gain/s but not loss/es detected by CGH; GS, only gain/s (⭐3) detected by CGH; GM, multiple gains (⭓3) not loss/es detected by CGH. b

thiocyanate (FITC)-conjugated dCTP and dUTP (DuPont, Boston, MA, USA), and Texas-red-conjugated dCTP and dUTP (DuPont) by nick translation to obtain fragments ranging from 600 to 2000 bp as reported previously.18 The hybridization mixture consisted of 400 ng of blast DNA, 400 ng of reference DNA, and 10 ␮g of unlabeled Cot-1 DNA (Gibco BRL, Life Technologies, Gaithersburg, MD, USA) dissolved in 10 ␮l of hybridization buffer (50% formamide, 10% dextran sulphate, 2 × SSC). The hybridization mixture was denatured at 75°C for 5 min and hybridized to a slide with normal metaphase spreads denatured in 70% formamide/2 × SSC (pH 7) at 68°C for 2 min. Hybridization was performed at 37°C for 48 h. Afterwards the slides were washed three times in 50% formamide/2 × SSC (pH 7), twice in 2 × SSC, and once in 0.1 × SSC at 45°C, followed by 2 × SSC, 0.1 M NaH2PO4-0.1 M Na2HPO4-0.1% NP40 (pH 8), and distilled water at room temperature for 10 min each. After air-drying, the slides were counterstained with 4′,6-diamidino-2-phenylindole (DAPI)

(Sigma Chemical, St Louis, MO, USA), and mounted with an antifading medium (Vectashield; Vector Laboratories, Burlingame, CA, UA).

Digital image analysis The hybridizations were analyzed using an Olympus fluorescence microscope and the ISIS digital image analysis system (MetaSystems, Altlussheim, Germany) based on an integrated high-sensitivity monochrome charge-coupled device (CCD) camera and automated CGH analysis software. Three-color images (red for reference DNA, green for blast DNA, and blue for counterstaining) were acquired from 8–12 metaphases per sample. Only metaphases of good quality with strong uniform hybridization were included in the analysis. Chromosomes not suitable for CGH analysis were excluded (ie chromosomes heavily bent, overlapping, or with overlying artefacts). Unless

Bio-technical methods section (BTS) ML Larramendy et al

otherwise indicated, chromosomal regions were interpreted as overrepresented when the corresponding ratio exceeded 1.17 (gains) or 1.5 (high-level amplification), and as underrepresented (losses) when the ratio was less than 0.85. If necessary, the threshold values were corrected according to the ploidy level as recommended elsewhere.19 This resulted in lower and upper threshold values of 0.57 and 0.78 for pseudotriploid, and 0.43 and 0.58 for pseudotetraploid blast cells, respectively. The ploidy level was chosen as the closest round integral to the number of chromosomes revealed by the karyotype analysis. In each CGH experiment, a negative (peripheral blood DNA from normal controls) and a positive (tumor DNA with known copy number changes) control were included and run simultaneously with the blast samples. All results were confirmed using a 99% confidence interval. Briefly, intraexperiment standard deviations for all positions in the CGH ratio profiles were calculated from the variation of the ratio values of all homologous chromosomes within the experiment. Confidence intervals for the ratio profiles were then computed by combining them with an empirical interexperiment standard deviation and by estimating the error probability based on the t-distribution.

Metaphase-fluorescent in situ hybridization (metaphase-FISH) Chromosome painting using fluorescein isothiocyanate (FITC)conjugated DNA probes specific to chromosomes 6 and 10 (Cambio, Cambridge, UK) was performed to confirm the chromosomal aberrations in two patients (Nos 26 and 69). Hybridizations and washings were performed according to the supplier’s instructions. The analysis was performed using a Zeiss Laborlux fluorescence photomicroscope with Zeiss filters 02 (FITC) and 09 (DAPI).

Statistical analysis The association between CGH aberrations and clinical features was tested by the ␹2 test or Fisher’s exact test in the case of low entries. Comparison of the white cell counts in different DNA copy number change categories was made using the Mann–Whitney U test.

Results

Gains and high-level amplifications A high proportion of the gains (87.3%) were observed to affect whole chromosomes and less frequently chromosome arms (2.2%) or chromosomal bands (10.5%) (Table 1, Figure 1). Gains of DNA sequence copy number were most commonly observed affecting whole chromosomes X (19.4%), 10 (19.4%), 14 (18.1%), 17 (19.4%), 18 (22.2%), and 21 (25.0%). The most frequent partial gains were observed at Xq25-q26 (25.0%), 17q (22.2%), 10p12-pter (20.8%), 6p (18.1%), 8q22qter (15.3%), and 1q31-q32 (8.3%). High-level amplifications were seen in two patients (patients 60 and 62). The amplification of 21 was seen twice (patients 60 and 62) and amplified chromosomes 8, 10 and 18 in one case (patient 62) (Figure 1).

Losses of DNA Losses were observed to affect equally chromosome arms (50.0%) and chromosomal bands (50.0%) but never whole chromosomes (Table 1, Figure 1). The most common regions with losses were 9p (6.9%) and 12p (6.9%). The most frequently affected genomic regions were 9p22-pter (12.5%) and 12p13-pter (11.1%). In three patients (4.2%), 9p22-pter and 12p13-pter were simultaneously affected (patients 39, 40 and 43).

Association analysis of changes detected by CGH A combination analysis of the DNA copy number changes detected by CGH was performed (Table 2). The most frequent association was observed between gains in chromosomes 17 and 18 (13 patients) and, at lower frequency, between simultaneous gains in chromosome 18 and chromosome 4, 6, 10, 14 or 21, and between chromosomes 17 and 6 (12 patients) (Table 2). Losses of DNA copy number changes at 9p22-pter or losses at 12p13-pter occurred as frequently as gains in chromosome 21. No statistically significant correlations between the grouped CGH findings (category of DNA copy number changes) and white blood cell counts at diagnosis could be demonstrated. However, we observed that the trend in patients, for whom CGH revealed one to three gains but no losses of DNA sequences (GS patients, Table 1), was to have the highest white blood cell count at diagnosis (white blood cell count median value 89.8 and 12.8, for GS and the rest, respectively), but the differences were not statistically significant (P = 0.09).

Overview of DNA sequence copy number changes The CGH results for all the patients are shown in Table 1. CGH revealed that 45 (62.5%) of the 72 ALL patients showed extensive changes with a mean of 4.6 aberrations per patient (range, 1 to 22). The other 27 patients (37.5%) without DNA copy number changes include 23 patients with a normal karyotype (31.9%; patients 1–23). Two of these patients carried balanced translocations as sole aberrations (2.8%; patients 24 and 25), one patient had an aberrant clone in 15.4% of the cells (1.4%; patient 26), and, finally, one patient had no available diagnostic karyotype (1.4%; patient 27). Gains were more frequent than losses (gains:losses, 6:1). All chromosomal regions with an increased or decreased DNA sequence copy number are summarized in Figure 1.

Similarities and discrepancies between chromosome banding and CGH analysis The chromosome banding results of the patients are shown in Table 1. Chromosomal aberrations were detected in 39 patients (54.2%). Among the patients with abnormal karyotypes, two (2.8%) had balanced chromosomal translocations as sole alterations (patients 24 and 25). Normal karyotype was found in 28 patients (38.9%; patients 1–23, 36, 40, 44, 57 and 69), and in five (6.9%; patients 27, 28, 30, 31 and 56) no evaluable metaphases were obtained for analysis. Table 3 shows the similarities and discrepancies between the chromosome banding and CGH analysis. There was only one case (patient 26) in which the CGH result was normal even though

1641

Bio-technical methods section (BTS) ML Larramendy et al

1642

Figure 1 Summary of gains and losses of DNA sequence copy number in 72 childhood acute lymphoblastic leukemia patients analyzed by CGH at diagnosis. Losses are shown on the left side and gains on the right of each chromosome. Each line represents a genetic alteration seen in one patient. High-level amplifications of small chromosomal regions are marked with thick lines. Asterisk under a bar denotes constitutional chromosomal alteration.

Table 2 The most frequent simultaneous/concurrent DNA sequence copy number gains in 72 childhood acute lymphoblastic leukemia patients

+21 +18 +X +10 +17 +14 +4

+6

+8

Total

10 12 8 9 12 9 9

7 10 7 9 7 7 7 7

18 16 14 14 14 13 12 12 10

+21 +18 +X +10 +17 +14 +4 +6 +8

12 8 10 11 9 8 10 7

11 12 13 12 12 12 10

9 9 7 9 8 7

9 6 10 9 9

10 10 12 7

9 9 7

9 7

7

Total

18

16

14

14

14

13

12

12

12

8 11

10 12 9

11 13 9 9

9 12 7 6 10

8 12 9 10 10 9

10

chromosome banding revealed a numerical aberration, trisomy 6. The result was confirmed by metaphase-FISH with chromosome 6 specific probe (Figure 2). FISH indicated that the frequency of trisomic cells was as low as 6%, which fully explains the normal CGH finding because the frequency was beyond the detection level (50%) of CGH. For patient 69, who had a normal karyotype but whose CGH results showed several gains including chromosome 10, metaphase-FISH using chromosome 10 specific probe indicated that 3% of mitotic cells had trisomy 10. Discussion More than one half of our cases with childhood ALL showed DNA copy number changes. We demonstrate here that the CGH analysis provides support for standard chromosome

Table 3 Similarities and discrepancies between CGH and chromosome banding analyses

Findings G-banding unsuccessful/CGH successful CGH unsuccessful/G-banding successful Normal in both techniques G-banding showed balanced translocation/ CGH normal Identical gains/losses in both techniques G-banding normal/GGH abnormal CGH normal/G-banding abnormal CGH gave additional information to G-banding G-banding gave additional information to CGH

No. of cases 5 (7) 0 23 (32) 2 (3) 7 5 1 25 8

(10) (7) (1) (35) (11)

Figure 2 Patient 26: metaphase-FISH with chromosome 6 specific probe revealed trisomy 6 in two out of 35 metaphases studied (A), whereas the fluorescent-to-blast reference ratio profile of CGH showed no change (B).

Bio-technical methods section (BTS) ML Larramendy et al

banding analysis. In as many as 34 cases, CGH gave additional information for chromosome banding analysis. In 17 cases (patients 29, 33–35, 37, 38, 41–43, 50, 51, 60, 61, 64, 71 and 72) chromosome banding analysis revealed the chromosome number but seriously failed in detecting chromosomes that were gained, missing or involved in marker chromosomes. In these cases CGH helped to recognize the trisomies and facilitated the selection of FISH probes for patient follow-up.20 CGH also provided information on four patients without any available karyotype data (patients 27, 28, 30 and 31) and on five patients with normal karyotypes after banding analysis (patients 36, 40, 44, 57 and 69). Because CGH appears to complement routine cytogenetics effectively, we are currently performing CGH analysis routinely on all patients with childhood ALL at diagnosis. In some cases, however, chromosome banding provided information that was not detected by CGH. For one case, chromosome banding analysis indicated trisomy 6 but CGH was normal. Because our FISH analysis showed that the frequency of trisomic cells was low (6%), the most evident explanation is that CGH is unsensitive to changes that occur at low frequencies. Furthermore, for eight cases, chromosome banding showed additional aberrations not seen in CGH. This may result from tumor heterogeneity. These restrictions of CGH and especially its inability to reveal balanced translocations clearly indicate the importance of chromosome banding analysis as the basic diagnostic tool in ALL. DNA copy number gains of whole chromosomes were seen in 39% of all cases. The most common gains (19–25%) were found in chromosomes 21, 18, X, 10 and 17 and, at lower frequencies, in chromosomes 14, 4, 6 and 8 (14–18%). The detected gains agree well with the results we obtained using chromosome banding, previous findings14,21,22 and those presented in two recent CGH reports.23,24 The most frequent gains of whole chromosomes (21, 18, X, 10, 17, 14, 4, 6 and 8) appear concurrently. This might indicate interaction between oncogenes and other genes in these chromosomes. Partial gains of chromosomes were rarely observed in our material. However, gains in 1q affecting the minimal common region of 1q31-q32 were the most frequently observed in six out of the 72 ALL patients (8.3%) (patients 45–47, 52, 54 and 72). The gain at 1q24-qter, detected by CGH as the sole aberration in three of these six patients (patients 45–47), is most probably due to an unbalanced t(1;19)(q24;p12), a wellknown aberration in ALL.14 An unbalanced translocation may be overlooked unless the quality of chromosome banding is good. In these cases CGH supports cytogenetics. There were two recurrent losses affecting 9p and 12p. These losses had not been observed in a recent comprehensive CGH analysis.24 However, these changes are well-known through chromosome banding analysis,3,14 but the banding analysis revealed the loss at 9p only in one out of nine cases (patient 39) and the loss at 12p in two out of eight (patients 38 and 39). Furthermore, CGH analysis helped to identify the deleted area, making it possible to narrow it down to 9p22-pter and 12p12-p13, respectively. The chromosomal band 9p21 harbors two adjacent tumor suppressor genes, CDKN2 (p16) and CDKN2B (p15). These genes have been reported to be deleted in 14–40% of ALL (reviewed in Quesnel et al25). CDKN2 is deleted more often, but it is frequently codeleted with CDKN2B. Okuda et al26 found the deletions of CDKN2/CDKN2B to be associated with poor event-free survival. At least two other tumor suppressor genes at 9p21 have been found to be lost in ALL: IFNalpha27–29 and MTAP.27

The chromosomal band 12p13 contains the ETV6 (TEL) gene, involved in translocation (12;21)(p13;q22) which is the most frequent chromosomal abnormality in childhood ALL.30,31 Patients with this translocation seem to have a favorable prognosis.1 Cave´ et al32 found ETV6 to be deleted in 23% of 215 cases. The deletion of ETV6 and t(12;21) were associated in most patients. In these patients one allele is thus altered by translocation and the normal allele is lost. Another putative tumor suppressor gene at 12p13 is KIP1, which has also been found to be deleted in childhood ALL.33 Compared to the molecular studies our lower frequencies of losses at 9p and 12p are logical considering the methodological restrictions of CGH in detection of deletions smaller than 3–5 Mb.10 However, it seems that a large proportion of losses at 9p and 12p are also detectable by CGH. Because these regions contain genes significant for ALL leukemogenesis, it is important to identify the changes accurately. CGH appears to add positively and significantly to the tools available for comprehensive cytogenetic analysis of childhood ALL. The evaluation of the diagnostic and/or prognostic relevance of individual CGH findings in the lymphoid blasts, however, awaits further studies on a larger series of patients. Given the wealth of detailed cytogenetic information obtained through CGH, we feel its inclusion in the armamentarium used in the primary cytogenetic evaluation of childhood ALL appears justified.

Acknowledgements This work was supported by the Sigrid Juse´lius Foundation, the Finnish Cancer Society, and the Research Institute of the Helsinki University Central Hospital in Finland; the National Council of Scientific and Technological Research (CONICET), and the National University of La Plata in Argentina.

References 1 McLean TW, Ringold S, Neuberg D, Stegmaier K, Tantravahi R, Ritz J, Koeffler HP, Takeuchi S, Janssen JW, Seriu T, Bartram CR, Sallan SE, Gilliland DG, Golub TR. TEL/AML-1 dimerizes and is associated with a favorable outcome in childhood acute lymphoblastic leukemia. Blood 1996; 88: 4252–4258. 2 Secker-Walker LM. Prognostic and biological importance of chromosome findings in acute lymphoblastic leukemia. Cancer Genet Cytogenet 1990; 49: 1–13. 3 Raimondi SC. Current status of cytogenetic research in childhood acute lymphoblastic leukemia. Blood 1993; 79: 3316–3324. 4 Pui C-H, Christ WM, Look AT. Biology and clinical significance of cytogenetic abnormalities in childhood acute lymphoblastic leukemia. Blood 1990; 76: 1449–1463. 5 Hunger SP. Chromosomal translocations involving the E2A gene in acute lymphoblastic leukemia: clinical features and molecular pathogenesis. Blood 1996; 87: 1211–1224. 6 Williams DL, Harris A, Williams KJ, Brosius MJ, Lemonds W. Direct bone marrow chromosome technique for acute lymphoblastic leukemia. Cancer Genet Cytogenet 1984; 13: 239–257. 7 Stewart EL, Secker-Walker LM. Detection of the chromosomally abnormal clone in acute lymphoblastic leukemia. Cancer Genet Cytogenet 1986; 23: 25–35. 8 Webber LM, Garson OM. Fluorodeoxyuridine synchronization of bone marrow cultures. Cancer Genet Cytogenet 1983; 8: 123– 132. 9 Kallioniemi A, Kallioniemi O-P, Sudar D, Rutovitz D, Gray JW, Waldman F, Pinkel D. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 1992; 258: 818–821. 10 Kallioniemi O-P, Kallioniemi A, Piper J, Isola J, Waldman FM,

1643

Bio-technical methods section (BTS) ML Larramendy et al

1644

11

12 13 14 15

16

17 18

19 20

21 22 23

Gray JW, Pinkel D. Optimizing comparative genomic hybridization for analysis of DNA sequence copy number changes in solid tumors. Genes Chromosom Cancer 1994; 10: 231–243. du Manoir S, Speicher MR, Joos S, Schro¨ck E, Popp S, Do¨hner H, Kovacs G, Nicoud MR, Lichter P, Cremer T. Detection of complete and partial chromosome gains and losses by comparative genomic in situ hybridization. Hum Genet 1993;90: 590–610. Monni O, Joensuu H, Franssila K, Knuutila S. DNA copy number changes in diffuse large B-cell lymphoma – comparative genomic hybridization study. Blood 1996; 87: 5269–5278. El-Rifai W, Elonen E, Larramendy ML, Ruutu T, Knuutila S. Chromosomal breakpoints and changes in DNA copy number in refractory acute myeloid leukemia. Leukemia 1997; 11: 958–963. Heim S, Mitelman F. Cancer Cytogenetics, 2nd edn. Wiley-Liss: New York, 1995. Monni O, Joensuu H, Franssila K, Klefstrom J, Alitalo K, Knuutila S. BCL2 overexpression associated with chromosomal amplification in diffuse large B-cell lymphoma. Blood 1997; 90: 1168– 1174. Knuutila S, Vuopio P, Elonen E, Siimes M, Kovanen R, Borgstro¨m GH, de la Chapelle A. Culture of bone marrow reveals more cells with chromosomal abnormalities than the direct method in patients with hematologic disorders. Blood 1981; 58: 369–375. Mitelman F (ed). ISCN. An International System for Human Cytogenetic Nomenclature. S Karger: Basel, 1995, pp 1–114. El-Rifai W, Larramendy ML, Bjo¨rkqvist A-M, Hemmer S, Knuutila S. Optimization of comparative genomic hybridization using fluorochrome conjugated to dCTP and dUTP nucleotides. Lab Invest 1997; 77: 699–700. du Manoir S, Schro¨ck E, Bentz M, Speicher MR, Joos S, Ried T, Lichter P, Cremer T. Quantitative analysis of comparative genomic hybridization. Cytometry 1995; 19: 27–41. El-Rifai W, Ruutu T, Vettenranta K, Elonen E, Saarinen UM, Volin L, Knuutila S. Follow-up of residual disease using metaphase-FISH in patients with acute lymphoblastic leukemia in remission. Leukemia 1997; 11: 633–638. Raimondi SC, Roberson PK, Pui C-H, Behm FG, Rivera GK. Hyperdiploid (47–50) acute lymphoblastic leukemia in children. Blood 1992; 79: 3245–3252. Sandberg AA. The Chromosomes in Human Cancer and Leukemia. Elsevier: New York, 1990. Karhu R, Siitonen S, Tanner M, Keina¨nen M, Ma¨kipernaa A, Leht-

24

25 26

27

28

29

30

31

32

33

inen M, Vilpo JA, Isola J. Genetic aberrations in pediatric acute lymphoblastic leukemia by comparative genomic hybridization. Cancer Genet Cytogenet 1997; 95: 123–129. Paszek-Vigier M, Talmant P, Me´chinaud F, Garand R, Harousseau J-L, Bataille R, Avet-Loiseau H. Comparative genomic hybridization is a powerful tool, complementary to cytogenetics, to identify chromosomal abnormalities in childhood acute lymphoblastic leukemia. Br J Haematol 1997; 99: 589–596. Quesnel B, Preudhomme C, Fenaux P. p16ink14a gene and hematological malignancies. Leuk Lymphoma 1996; 22 :11–24. Okuda T, Shurtleff S, Valentine M, Raimondi S, Head D, Behm F, Curcio-Brint A, Liu Q, Pui C, Sherr C. Frequent deletion of p16INK4a/MTS1 and p15INK4b/MTS2 in pediatric acute lymphoblastic leukemia. Blood 1995; 85: 2321–2330. Dreyling M, Bohlander S, Le Beau M, Olopade O. Refined mapping of genomic rearrangements involving the short arm of chromosome 9 in acute lymphoblastic leukemias and other hematologic malignancies. Blood 1995; 86: 1931–1938. Hirai H, Ogawa S, Hangaishi A, Takahashi T, Kurokawa M, Mitani K, Ueda R, Yazaki Y. Recent progress in molecular mechanisms of leukemogenesis: the cyclin-dependent kinase 4-inhibitor gene in human leukemias. Leukemia 1997; 11: 358–360. Taceuchi S, Koike M, Seriu T, Bartram C, Slater J, Park S, Miyoshi I, Koeffler H. Homozygous deletions at 9p21 in childhood acute lymphoblastic leukemia detected by microsatellite analysis. Leukemia 1997; 11: 1636–1640. Romana S, Poirel H, Leconiat M, Flexor M-A, Mauchaffe M, Jonveaux P, Macintyre E, Berger R, Bernard O. High frequency of t(12;21) in childhood B-lineage acute lymphoblastic leukemia. Blood 1995; 86: 1263–1269. Liang D, Chou T, Chen J, Shurtleff S, Rubnitz J, Downing J, Pui C, Shih I. High incidence of TEL/AML1 fusion resulting from cryptic t(12;21) in childhood B-lineage acute lymphoblastic leukemia in Taiwan. Leukemia 1996; 10: 991–993. Cave´ H, Cacheux V, Raynaud S, Brunie G, Bakkus M, Cochaux P, Preudhomme C, Lai J, Vilmer E, Grandchamp B. ETV6 is the target of chromosome 12p deletions in t(12;21) childhood acute lymphocytic leukemia. Leukemia 1997; 11: 1459–1464. Ho¨glund M, Johansson B, Pedersen-Bjergaard J, Marynen P, Mitelman F. Molecular characterization of 12p abnormalities in hematologic malignancies: deletion of KIP1, rearrangement of TEL, and amplification of CCND2. Blood 1996; 87: 324–330.