Cancer Genetics and Cytogenetics 136 (2002) 101–107

0165-4608/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.PII: S0165-4608(02)00515-0

Fluorescence in situ hybridization for the study of cell lineage involvement in myelodysplastic syndromes

with chromosome 5 anomalies

Kristina Anderson

a,

*, Ingrid Arvidsson

a

, Björn Jacobsson

b

, Robert Hast

a

a

Division of Hematology, Department of Medicine, Karolinska Hospital, Karolinska Institutet, Stockholm, Sweden

b

Department of Clinical Pathology, Karolinska Hospital, Karolinska Institutet, Stockholm, Sweden

Abstract

Fluorescence in situ hybridization (FISH) with a locus-specific dual DNA probe (LSI EGR-1SO/D5S23SG) for chromosome 5 was used in combination with morphology to study bone marrow cell lin-eage involvement of the abnormal chromosomal clone in 13 patients with deletion 5q [del(5q)], either as asole aberration or as part of a complex karyotype, and in six cases with monosomy 5 by metaphase cytoge-netics, all with complex karyotypes including 2–6 marker chromosomes. In the monosomy 5 group, onlyone case displayed the expected one orange and one green (1O

�

1G) FISH pattern in a majority of thecells. The other five patients instead showed 1O

�

2G FISH signals in 17–86% of the bone marrow cells,which is the typical pattern for del(5q). In the del(5q) group, 26–98% of the bone marrow cells exhibited1O

�

2G FISH signals. All patients showed clonal involvement of the myeloid cell lineages, including themegakaryocytes in a few cases, whereas lymphoid cells generally exhibited the normal 2O

�

2G FISHpattern. No difference was seen between patients with 5q

�

syndrome, those with del(5q) and a complexkaryotype, and the monosomy 5 group. We were thus unable to confirm the recent suggestion that B-cellsare a part of the abnormal clone in MDS with del(5q). Furthermore, true monosomy 5 seems to be rare in

MDS. © 2002 Elsevier Science Inc. All rights reserved.

1. Introduction

In myelodysplastic syndromes (MDS), deletions of thelong arm of chromosome 5 have been reported in up to 27%of the cases [1]. The aberration, which may appear as a soleabnormality or as part of a complex karyotype, is among themost frequently reported chromosomal abnormalities inMDS [2,3], but the deleted segments vary in size and break-point localization. The most frequently reported breakpointsare 5q12

�

q14 and 5q31

�

q33 [4–7]. The consistent loss ofgenetic material from the long arm of chromosome 5 in my-eloid disorders has led to the speculation that the deleted re-gion may harbor a tumor suppressor gene of pathogenic im-portance for the development of leukemias [8]. In contrast,monosomy 5 usually appears as part of a complex karyo-type involving several other chromosomes and rarely as asole abnormality [9,10]. Monosomy 5 has been consideredless likely to be a primary karyotypic abnormality and notof pathogenic significance in MDS [11].

Clonal involvement of mature lymphoid cells in MDS hasbeen observed by some investigators [12–14], but not seen byothers [15–18]; however, these studies included a number ofdifferent chromosome abnormalities. Recent studies under-taken in an attempt to clarity the involvement of pluripotentstem cells have provided new insights into the biology ofMDS. Saitoh et al. reported that MDS clones with trisomy 8appear first in a committed myeloid progenitor cell, corre-sponding to CFU-GEMM [19]. Nilsson et al., on the otherhand, found that in MDS patients with del(5q) an extraordi-nary high proportion (

�

98%) of the pluripotent CD34

�

CD38

�

stem cells carried the aberration [20]. Nilsson et al.were also able to show for the first time that the abnormaldel(5q) clone in some cases involved early stem cells of lym-phoid commitment (CD

�

CD19

�

) [20]. It is likely thatdel(5q) is a primary event in the development of MDS, whilenumerical changes like trisomy 8 are acquired later.

We wanted to study the penetration of the abnormalchromosomal clone into the different bone marrow cell lin-eages in MDS patients with chromosome 5 anomalies, andcompare the outcome in patients with del(5q) as a sole ab-normality, del(5q) as part of a complex karyotype, andmonosomy 5, respectively, particularly regarding lymphoidcell involvement.

* Corresponding author. BMC B12, Klinikgatan 26, Department ofStem Cell Biology, Lund University, S-221 84 Lund, Sweden; Tel:

�

46-46-2223074; fax:

�

46-46-2223600.

E-mail address

: kristina.anderson@stemcell.lu.se (K. Anderson).

102

K. Anderson et al. / Cancer Genetics and Cytogenetics 136 (2002) 101–107

2. Materials and Methods

2.1. Patients

Nineteen patients with MDS diagnosed at DanderydHospital between 1988 and 1998 entered the study. Thirteenpatients had del(5q), either as a sole aberration or as a partof complex chromosomal clone, and 6 had monosomy 5, inall cases with a complex karyotype. Clinical, cytogeneticand hematological descriptions of the patients are shown inTable 1. The median age for patients with del(5q) was 79years (range 43–84) and for the monosomy 5 group 66 years(range 52–79). Three patients with monosomy 5 were con-sidered to be therapy-related MDS (t-MDS), since they hadbeen treated with chemotherapy prior to the development ofMDS; two cases received melphalan for multiple myelomaand one different combination regimens for seminoma.

2.2. Metaphase cytogenetics

Bone marrow samples for chromosomal investigationswere obtained and analyzed at the time of diagnosis accord-

ing to routine procedures at the Department of Clinical Ge-netics, Karolinska Hospital.

2.3. Fluorescence in situ hybridization (FISH)

For detection of cell lineage involvement, we used ourMGG/FISH method that allows detection of chromosomalaberrations in morphologically defined cells [21]. In short,selected cells after MGG staining were photographed usingan epifluorescence microscope (Zeiss) and color CCD cam-era (DX-930, SONY). After destaining in methanol/aceticacid (3:1) and dehydration, the bone marrow smears werepostfixed in 1% neutrally buffered formalin (NBF) followedby microwave heat treatment twice in sodium citrate pH 6.0for 5 minutes. Smears were then treated with 0.01% protein-ase K (type VIII, Sigma, MO, USA) followed by subse-quent fixation in NBF.

A dual-color DNA probe (LSI EGR-1SO/D5S23SG, Vy-sis, Downers Grove, IL, USA) specific for band 5q31 of thelong arm and band 5p15.2 of the short arm on chromosome5 was used for hybridization according to the manufacturers

Table 1Clinical description, MGG/FISH, and metaphase cytogenetics in 19 MDS patients with del(5q) (patients 1–13) and monosomy 5 (patients 14–19)

Patientno. Age/sex Diagnosis

Blastsin BM%

Blasts

MGG/FISH

Cytogenetics

P/Mc/Mm B/S Mo Ly Ebl No. ofcells% (N) % (N) % (N) % (N) % (N) % (N)

1 70/F RA 4 — — 65 (92) 56 (9) 0 (47) — 156

a

46,XX,del(5)(q31)[10]2 82/F RA

�

5 100 (1) 100 (22) 100 (15) 100 (5) 0 (45) 100 (1) 100 46,XX,del(5)(q14)[33]3 43/F RA 2 60 (5) 86 (103) 91 (97) 71 (7) 0 (62) 61 (23) 305 46,XX,del(5)(q13q33)[23]/46,XX[3]4 84/F RA

�

5 93 (30) 95 (59) 100 (47) 66 (3) 0 (66) 79 (34) 277 46,XX,del(5)(q21)[18]5 70/F RA

�

5 86 (43) 93 (134) 96 (108) 100 (13) 0 (27) 82 (17) 376 46,XX,del(5)(q13q33)[30]6 82/F RAEB-t 20 100 (59) 100 (42) 93 (27) 100 (1) 0 (10) 100 (39) 191 46,XX,del(5)(q?)[33]7 82/F RA

�

5 89 (36) 96 (139) 97 (76) 100 (8) 0 (25) 89 (28) 257 46,XX,del(5)(q21)[20]8 58/F RAEB-t 26 97 (33) 100 (13) 73 (69) 0 (3) 0 (215) 100 (3) 456 46,XX,del(5)(q15)[10]/47,XX,

�

mar[2]/46,XX[3]9 73/M RAEB-t 20 88 (110) 75 (4) 46 (13) 100 (2) 0 (38) 90 (553) 696 43

�

46,XY,

�

3,

�

4,

�

5, del(5)(q?),

�

7,

�

8,

�

13,inc[29]/46,XY[2]10 80/M RA 4 86 (22) 91 (32) 80 (50) 75 (4) 0 (55) 4 (218) 386 47,XY,del(5)(q?),

�

8[1]/46,XY[12]11 56/M RAEB 13 84 (38) 91 (131) 98 (113) 80 (5) 0 (98) 81 (216) 522 46,XY,del(5)(q?),add(11),

�

14,

�

17,

�

mar[36]12 79/F RA

�

5 — 86 (14) 73 (102) 25 (4) 0 (28) 79 (14) 169 46,XX,del(5)(q22)[24]/47,XX,

�

8[34]/46,XX[8]13 83/F RA

�

5 100 (4) 94 (33) 100 (36) 100 (3) 0 (13) 90 (89) 182 46,XX,del(5)(q?),del(11)(q14)[17]/46,XX,del(2)(q33),del(5)(q?),del(11)(q14)[9]

14 66/F RARS 3 100 (6) 75 (8) 33 (12) — 0 (5) 89 (305) 44 45

�

48,XX,

�

4,

�

5,del(7)(q?),

�

8,

�

16,

�

17,del(19)(?),

�

1

�

6mar[33]/46,XX[2]

15

a

79/M MM/RARS

�

5 100 (9) 2 (62) 52 (23) 100 (2) 0 (18) 85 (128) 392 45,X,

�

Y,

�

4,

�

5,

�

12,

�

17,

�

5mar[23]16

a

63/M MM/RA 4 50 (10) 20 (15) 23 (11) 33 (6) 0 (16) 18 (28) 201 44,XY,

�

5,

�

7,

�

15,

�

17,

�

2mar[11]/46,XY[9]

17 71/F RARS 1 100 (15) 89 (18) 56 (16) 100 (4) 0 (30) 94 (223) 279 46,XX,

�

3,

�

5,

�

17,

�

3mar[5]/45,XX,

�

5,

�

17,

�

mar[3]/46,XX[14]18 57/F RA/AML 80 99 (326) 100 (10) 100 (4) — 0 (11) 100 (1) 362 43

�

45,XX,add(2)(q?),del(3)(p?),

�

5,

�

7,

�

2mar[9]19

b

52/M Ca/RAEB 8 7 (14) 4 (26) 15 (48) — 0 (10) 9 (274) 408 43

�

45,XY,del(3)(p?),

�

5,

�

7,inc[32]/46,XY[2]

The outcome of MGG/FISH is presented as the percentage of the cells exhibiting one orange and two green (1O

�

2G) signals within different cell lineages.

Abbreviations

: B, bands; Ca, carcinoma; Ebl, erythroblasts; Ly, lymphocytes; Mc, myelocytes; Mm, metamyelocytes; MM, multiple myeloma; N, numberof cells; Pts, patients; S, segmented neutrophils.

a

Peripheral blood.

b

Patients previously treated for antecedent malignancy.

K. Anderson et al. / Cancer Genetics and Cytogenetics 136 (2002) 101–107

103

instructions. After hybridization, the slides were washed in 1

�

SSC for 5 minutes at 72

�

C. Finally, the bone marrow slideswere mounted in DAPI (4, 6-diamidino-2-phenylindole 1

�

g/ml, Sigma).

The hybridized slides were evaluated with an epifluores-ence microscope (Zeiss) equipped with the appropriate filtersets. Between 100 and 2000 cells were evaluated in eachcase and the number of nuclei with 0, 1, 2, or

3 orange (O)and green (G) signals were counted. The normal FISH pat-tern is 2O

�

2G FISH signals, whereas 1O

�

2G is typicalfor a cell with del(5q), and 1O

�

1G for cases with mono-somy 5.

2.4. Control specimens

The specificity of the dual-color probe for chromosome 5was tested on bone marrow smears from 15 control subjects.

Between 35 and 2000 cells were counted in each controlsample in order to establish a normal range. The percent-ages of control cells showing different combinations of or-ange and green fluorescent spots are shown in Table 2. Themean percentage (

SD) of control cells with a normal dis-omic FISH pattern was 93.8% (

4.8), and with 1O

�

2GFISH signals 2.8% (

2.2).

3. Results

3.1. Deletion 5q patients

Metaphase cytogenetics showed del(5q) as a single aber-ration in 7 of 14 patients, and as part of a complex chromo-somal clone in the other 6 cases (Table 1). In 12 patientsdel(5q) was present in 36–100% (median 100) of themetaphases, while in one case the deletion was seen in only1 of 13 (8%) analyzed metaphases (Table 1). In the latterpatient, FISH analyses showed 1O

�

2G signals in 38% ofthe bone marrow cells which was comparable to the out-come of FISH in the other 12 cases with del(5q) (Fig. 1). 1O

�

1G FISH signals were found in between 0 and 4% (median0) of the bone marrow cells (Fig. 1), while other combina-tions of FISH signals were rare (

�

1%) in patients withdel(5q).

Penetration of the del(5q) clone into the different cell lin-eages in the bone marrow was studied using MGG-FISH

Table 2Mean percentage (

SD) of different combinations of FISH signals using the LSI EFR-1SO/D5S23SG probe for chromosome 5 on bone marrow smears of 15 control subjects

Combinations of fluorescent spots

1O

�

1G 1O

�

2G 2O

�

2G 2O

�

1G Others

Mean (SD) 0.6 (0.9) 2.75 (2.2) 93.8 (4.8) 1.27 (1.4) 1.6 (3.3)

35–2000 cells were scored in each experiment.

Fig. 1. Outcome of FISH using a dual-color probe for chromosome 5 in 19 patients with del(5q) and monosomy 5. The figure shows the percentages of thebone marrow cells with different combination of FISH signals (�, 1O � 1G; , 1O � 2G; �, 2O � 2G).

104

K. Anderson et al. / Cancer Genetics and Cytogenetics 136 (2002) 101–107

(Table 1). The median number of blast cells exhibiting 1O

�

2G FISH signals was 89% (range 60–100), of other im-mature myeloid cells 94% (range 75–100), of stabs and seg-mented neutrophils 93% (range 46–100%), of monocytes80% (range 0–100) and of erythroblasts 86% (range 4–100).Fig. 2 demonstrates MGG and FISH staining of bone mar-row erythroid cells from a monosomy 5 patient. In a fewcases MGG-FISH also showed involvement of megakaryo-cytes into the del(5q) clone (Fig. 3). In contrast, lymphoidcells and plasma cells showed only the normal 2O

�

2GFISH pattern in all patients with del(5q) (Table 1).

3.2. Monosomy 5 patients

All six patients with monosomy 5 by metaphase cytoge-netics also had complex chromosomal aberrations, includ-ing numerical and structural changes of several other chro-mosomes and 2–6 marker chromosomes (Table 1). Themedian percentage of bone marrow cells showing 1O

�

1GFISH signals was 4% (range 0–65) (Fig. 1). Only one pa-tient displayed the typical FISH pattern for monosomy 5(1O

�

1G) in the majority of the bone marrow cells,whereas the other five cases had

�

10% (Fig. 1). The latterfive patients instead showed 1O

�

2G FISH signals (Fig.2A) in 17–86% of the bone marrow cells (Fig. 1). Othercombinations of FISH patterns were seen in

�

1% of thebone marrow cells, except in two cases with t-MDS where3O

�

3G FISH signals were seen in 3% and 28% of thebone marrow cells (Table 1), respectively.

When MGG-FISH was used to study bone marrow celllineage involvement, it appeared that the predominant FISHpatterns were detected in the myeloid, but not in the lymphoidcells (Table 1). Among the five patients showing mainly 1O

�

2G FISH signals, the median percentage of blast cellsshowing the aberration was 90% (range 50–100), of other im-mature myeloid cells 73% (range 20–100), of neutrophils53% (range 23–100), and of erythroblasts 77% (range 18–100) (Table 1). The lymphoid cells were not involved in anyof the cases. The patients with a majority of 1O

�

1G FISHsignals showed a similar penetration of the abnormal cloneinto only the myeloid cell lineages (Table 1). In the twot-MDS patients, both having myeloma, where 3O

�

3G FISHsignals were found in a subset of bone marrow cells, trisomy5 was restricted to the plasma cells (Table 1).

4. Discussion

Previous studies on chromosome 5 abnormalities havereported that the indolent 5q

�

syndrome presents impair-ment in only the erythroid and megakaryocytic maturation,whereas the myelomonocytic differentiation was not mark-edly influenced [6]. In contrast, cases with del(5q) and pro-gressive MDS or AML have been found to demonstrate anarrest in a trilineage precursor cell [16]. Using our MGG-FISH method, which allows the evaluation of clonal in-volvement of different bone marrow cell lineages [21], wefound that all 13 patients with del(5q) by metaphase cytoge-

netics showed the typical 1O � 2G FISH pattern in bothmature and immature granulocytic cells, monocytes anderythroblasts, and that there were no differences betweenpatients with del(5q) was the sole aberration and those withcomplex chromosomal clones. In a few cases we could alsodemonstrate megakaryocytes containing 8O � 16G FISHsignals, a FISH pattern characteristic for del(5q) in a cellwith a ploidy number of 16N. Clonal involvement of themegakaryocytes has been difficult to evaluate by FISH be-cause of their size and the variable ploidy. van Lom et al. re-cently showed megakaryocytes to be part of the abnormalchromosomal clones in MDS with monosomy 7 and trisomy8 [22], and Gordon et al. demonstrated the same in MDSwith del(5q) [23].

Fig. 2. (Upper panel) FISH and (lower panel) May-Grünwald-Giemsacombined staining of bone marrow erythroid cells from a monosomy 5patient. A dual-DNA probe was used to identify the long (O) and short (G)arms of chromosome 5.

K. Anderson et al. / Cancer Genetics and Cytogenetics 136 (2002) 101–107 105

In contrast to the present findings of clonal cells in allmyeloid cell lineages, we did not observe lymphoid cellsexhibiting del(5q) in any of the patients. The result agreeswith an earlier study in MDS with del(5q) [18], but is atvariance with two recent reports [13,20]. Both these studiesclearly showed CD19� cells with del(5q), but only in a mi-nority of the patients studied. It is unclear why del(5q) isdifficult to detect in mature lymphocytes, even though it hasbeen shown that the aberration is present in the great major-ity of the pluripotent CD34�CD38� stem cells in MDS[20]. One explanation might be that there is a commitmentof the transformed pluripotent stem cells to the myeloid celllineage, not allowing maturation along the lymphoid celllineage [14]. It is also possible that small subpopulations ofB cells may be missed when investigated by routine mor-phology and FISH, since most lymphoid cells in the periph-eral blood or the bone marrow are T cells. The sensitivity ofthe FISH analysis to detect a small subset of abnormal B-cells can be increased combining FISH with immunopheno-typing [13] or using FACS separated subpopulations ofcells [20]. Therefore, estimation of the true frequency of

clonal involvement of mature B cells in MDS must awaitlarger studies using immunophenotyping.

In MDS patients with monosomy 5 by metaphase cyto-genetics, we found only one case in which FISH showed amajority of bone marrow cells with the typical pattern 1O �1G. Five of six patients instead exhibited 1O � 2G FISHsignals in most bone marrow cells, indicating that only apart of chromosome 5 was lost. The findings are in line withrecent reports in MDS with monosomy 5 or 7 by metaphasecytogenetics, showing that certain parts of the respectivechromosomes are involved in unbalanced translocations to avariety of partner chromosomes [5,9,24–26]. Although ourFISH technique did not allow for a more detailed character-ization of the parts of chromosome 5 that were lost, it wasevident that the 5q31 region was deleted in all cases, andthat part of the short arm was retained. All patients in themonosomy 5 group showed a complex karyotype bymetaphase cytogenetics, including several marker chromo-somes. It is conceivable that the marker chromosomes con-tained fragments of chromosome 5, either as unbalancedtranslocations, dicentrics or insertions, giving rise to a FISH

Fig. 3. FISH of a bone marrow megakaryocyte from a patient with del(5q). A dual-color DNA probe was used to identify the long (O) and short (G) arms ofchromosome 5 showing 8O � 16G FISH signals.

106 K. Anderson et al. / Cancer Genetics and Cytogenetics 136 (2002) 101–107

picture similar to del(5q). This is of interest since structuralchanges involving 5q31 are more likely to be of importancein the pathogenesis of MDS than a loss of the whole chro-mosome 5.

In a recent study, Andersen and Pedersen-Bjergaard [27]reported an increased frequency of dicentric chromosomesin t-MDS and t-AML patients after previous treatment withalkylating agents. A significant difference was observed be-tween the two largest groups of dicentrics, dic(1;7) anddic(5;17), respectively, where the former were observedalone or together with one or two other cytogenetic abnor-malities, whereas the latter were part of complex karyotypicabnormalities in all cases. Similarly, we found in the presentstudy that four of six patients in the monosomy 5 group ex-hibited monosomy 17 as part of their complex karyotypes,and that two of the patients had previously been treated withalkylating agents.

In two patients, we observed the rare FISH pattern 3O �3G signals in 3 and 28%, respectively, of the bone marrowcells. Both patients had previously been treated for multiplemyeloma. Trisomy 5 is among the most frequent chromo-somal aberrations in multiple myeloma [28], and usingMGG-FISH we could show that the trisomy 5 clone was re-stricted to the plasma cells. Thus, FISH visualized two dif-ferent clones, one with a gain of chromosome 5 associatedwith the myeloma, and another linked to the MDS whereparts of chromosome 5 had been lost. Whether or not theseseemingly unrelated findings have a common origin is un-clear.

Acknowledgments

Supported by the Cancer Society in Stockholm (98:111,99:151, 00:117), the Swedish Cancer Society (4148-B98–01XAB, 4148-B99–02XBB), and the Karolinska InstitutetFunds. Dr Anderson (Jakovleva), Tartu University, Estoniais the recipient of a personal grant from the Swedish Institue(Visby Programme) and the Royal Swedish Academy ofSciences.

References

[1] Mufti GJ. Chromosomal deletions in the myelodysplastic syndrome.Leuk Res 1992;16:35–41.

[2] Kantarmjian HM, Keating J, Walters RS, Smith TL, Cork A, Mc-Credie KB, Freireich EJ. Therapy-related leukemia and myelodys-plastic syndrome: clinical, cytogenetic and prognostic features. J ClinOncol 1986;4:1748–57.

[3] Le Beau MM, Albain KS, Larson RA, Vardiman JW, Davis EM,Blough RR, Golomb HM, Rowley JD. Clinical and cytogenetic corre-lations in 63 patients with therapy-related myelodysplastic syndromesand acute nonlymphocytic leukemia: further evidence for characteris-tic abnormalities of chromosomes no. 5 and 7. J Clin Oncol 1986;4:325–45.

[4] Van den Berghe H, Vermaelen K, Mecucci C, Barbieri D, Tricot G.The 5q� anomaly. Cancer Genet Cytogenet 1985;17:189–255.

[5] Johansson B, Martens F, Mitelman F. Cytogenetic deletion maps ofhematologic neoplasms. Genes Chromosomes Cancer 1994;8:205–18.

[6] Nagarajan L. Molecular analysis of the 5q� chromosome. LeukLymphoma 1995;17:361–6.

[7] Heim S, Mitelman F. Cancer cytogenetics. 2nd edition. New York:Wiley-Liss, 1995.

[8] Le Beau MM. Deletions of chromosome 5 in malignant myeloid dis-orders. Cancer Surv 1992;15:143–59.

[9] Kakazu N, Taniwaki M, Horiike S, Nishida K, Tatekawa T, Nagai M,Takahashi T, Akaogi T, Inazawa J, Ohki M, Abe T. Combined spec-tral karyotyping and DAPI banding analysis of chromosome abnor-malities in myelodysplastic syndrome. Genes Chromosomes Cancer1999;26:336–45.

[10] Mitelman F, Johansson B, Martens F, editors. Mitelman Database ofChromosome Aberrations in Cancer: Available at: http://cgap.nci.nih.gov/Chromosomes/Mitelman.

[11] Pedersen-Bjergaard J, Pedersen M, Roulston D, Philip P. Differentgenetic pathways in leukemogenesis for patients presenting with ther-apy-related myelodysplasia and therapy-related acute myeloid leuke-mia. Blood 1995;86:3542–52.

[12] van Lom K, Hagemeijer A, Smit EME, Hählen K, Groeneveld K, Lö-wenberg B. Cytogenetic clonality analysis in myelodysplastic syn-drome: monosomy 7 can be demonstrated in the myeloid and lym-phoid lineage. Leukemia 1995;9:1818–21.

[13] Jaju RJ, Jones M, Boultwood J, Kelly S, Mason DY, Wainscoat JS,Kearney L. Combined immunophenotyping and FISH identifies theinvolvement of B-cells in 5q� syndrome. Genes Chromosomes Can-cer 2000;29:276–80.

[14] Miura I, Kobayashi Y, Takahashi N, Saitoh K, Miura AB. Involve-ment of natural killer cells in patients with myelodysplastic syndromecarrying monosomy 7 revealed by the application of fluorescence insitu hybridization to cells collected by means of fluorescence-acti-vated cell sorting. Br J Haematol 2000;110:876–9.

[15] Gerritsen WR, Donohue J, Bauman J, Jhanwar SC, Kernan NA, Cas-tro-Malaspina H, O’Reilly RJ, Bourhis JH. Clonal analysis of myelo-dysplastic syndrome: monosomy 7 is expressed in the myeloid lin-eage, but not in the lymphoid lineage as detected by fluorescent insitu hybridization. Blood 1992;80:217–24.

[16] Kroef MJ, Fibbe WE, Mout R, Jansen RP, Haak HL, Wessels JW,Van Kamp H, Willemze R, Landegent JE. Myeloid but not lymphoidcells carry the 5q deletion: polymerase chain reaction analysis of lossof heterozygosity using mini-repeat sequences on highly purified cellfractions. Blood 1993;81:1849–54.

[17] Bernell P, Jacobsson B, Nordgren A, Hast R. Clonal cell lineage in-volvement in myelodysplastic syndromes studied by fluorescence insitu hybridization and morphology. Leukemia 1996;10:662–8.

[18] Kroef MJPL, Bolk MJW, Muus P, Wessels JW, Bewerstock GC,Willeme R, Landegent JE. Mosaicism of the 5q deletion as assessedby interphase FISH is a common phenomenon in MDS and restrictedto myeloid cells. Leukemia 1997;11:519–23.

[19] Saitoh K, Miura I, Takahashi N, Miura AB. Fluorescence in situ hy-bridization of progenitor cells obtained by fluorescence-activated cellsorting for the detection of cells affected by chromosome abnormalitytrisomy 8 in patients with myelodysplastic syndromes. Blood 1998;92:2886–92.

[20] Nilsson L, Åstrand-Grundström IB, Arvidsson I, Jacobsson B, Hell-ström-Lindberg E, Hast R, Jacobsen SEW. Isolation and characteriza-tion of hematopoietic progenitor/stem cells in 5q-deleted myelodys-plastic syndromes: evidence for involvement at the hematopoieticstem cell level. Blood 2000;96:2012–21.

[21] Jacobsson B, Bernell P, Arvidsson I, Hast R. Classical morphology, es-terase cytochemistry and interphase cytogenetics of peripheral bloodand bone marrow smears. J Histochem Cytochem 1995;44:1303–9.

[22] van Lom K, Houtsmuller AB, Wim LJ, van Putten WL, Slater RM,Lowenberg B. Cytogenetic clonality analysis of megakaryocytes inmyelodysplastic syndrome by dual-color fluorescence in situ hybrid-ization and confocal laser scanning microscopy. Genes ChromosomesCancer 1999;25:332–8.

[23] Godon C, Talmant P, Garand R, Accart F, Bataille R, Avet-Loiseau

K. Anderson et al. / Cancer Genetics and Cytogenetics 136 (2002) 101–107 107

H. Deletion of 5q31 is observed in megakaryocytic cells in patientswith myelodysplastic syndromes and a del(5q), including the 5q�

syndrome. Genes Chromosomes Cancer 2000;9:350–2.[24] Lessard M, Herry A, Berthou C, Léglise MC, Abgrall JF, Morice P,

Flandrin G. FISH investigation of 5q and 7q deletions in MDS/AMLreveals hidden translocations, insertions and fragmentations of thesame chromosomes. Leuk Res 1998;22:303–12.

[25] Mohr B, Bornhäuser M, Thiede C, Schäkel U, Schaich M, Illmer T,Pascheberg U, Ehninger G. Comparison of spectral karyotyping andconventional cytogenetics in 39 patients with acute myeloid leukemiaand myelodysplastic syndrome. Leukemia 2000;14:1031–8.

[26] Castro PD, Liang JC, Nagarajan L. Deletions of chromosome5q13.3 and 17p loci cooperate in myeloid neoplasms. Blood 2000;95:2138–43.

[27] Andersen MK, Pedersen-Bjergaard J. Increased frequency of dicen-tric chromosomes in therapy-related MDS and AML compared to denovo disease is significantly related to previous treatment with alky-lating agents and suggests a specific susceptibility to chromosomebreakage at the centromere. Leukemia 2000;14:105–11.

[28] Gould J, Alexanian R, Goodacre A, Pathak S, Hecht B, Barlogie B.Plasma cell karyotype in multiple myeloma. Blood 1988;71:453–6.