Cancer Genetics and Cytogenetics 140 (2003) 73–77

0165-4608/03/$ – see front matter © 2003 Elsevier Science Inc. All rights reserved.PII: S0165-4608(02)00631-3

Short communication

Chromosomal imbalances in pleomorphic rhabdomyosarcomas and identification of the alveolar rhabdomyosarcoma-associated

PAX3-FOXO1A

fusion gene in one case

Anthony Gordon

a,b

, Aidan McManus

a,b

, John Anderson

c

, Cyril Fisher

d

, Syuiti Abe

e

, Takayuki Nojima

f

, Kathy Pritchard-Jones

b

, Janet Shipley

a,

*

a

Section of Molecular Carcinogenesis, Molecular Cytogenetics Laboratory, Institute of Cancer Research, Sutton, Surrey, UK SM2 5NG, UK

b

Section of Pediatric Oncology, Molecular Cytogenetics Laboratory, Institute of Cancer Research, Sutton, Surrey, UK SM2 5NG, UK

c

Unit of Molecular Haematology, Institute of Child Health

1

, London WC1N 1EH, UK

d

Department of Anatomic Pathology, Royal Marsden Hospital, London, SW3 6JJ, UK

e

Chromosome Research Unit, Faculty of Science, Hokkaido University, Sapporo, Japan

f

Department of Pathology, Kanazawa Medical University, Uchinada, Japan

Received for publication 21 March 2002; received in revised form 28 May 2002; accepted 29 May 2002

Abstract

Rhabdomyosarcomas (RMS) are soft tissue sarcomas resembling developing skeletal muscle, and pleo-morphic rhabdomyosarcomas (PRMS) are a rare nonpediatric entity. Little molecular cytogenetic infor-mation exists for PRMS, and their relationship to other subtypes of rhabdomyosarcoma and other sarco-mas is unclear. Chromosomal imbalances were determined in seven well-characterized cases of PRMSusing comparative genomic hybridization. The smallest overlapping regions of gain were 1p22

�

p33(71%), 7p (43%), 18/18q (43%), and 20/20p (43%), and the regions of loss were 10q23 (71%),15q21

�

q22 (57%), 3p, 5q32

�

qter, and 13 (all 43%). Four of the seven cases had amplicons involving theregions 1p21

�

p31, 1q21

�

q25, 3p12, 3q26

�

qtel, 4q28

�

q31, 8q21

�

q23/8q, and 22q. These regions aredistinct from those frequently associated with the alveolar subtype, whereas the embryonal subtype with-out anaplasia is rarely associated with amplification events other than gain/amplification of 8q material.The regions of imbalance appeared more similar to those reported for malignant fibrous histiocytomas(MFH) and osteosarcomas, consistent with the suggestion that PRMS can be considered part of the spec-trum of MFH. In addition, one of the cases classified as PRMS showed evidence for the presence of a

PAX3-FOXO1A

fusion gene, which is characteristic of the alveolar subtype of RMS. © 2003 Elsevier

Science Inc. All rights reserved.

1. Introduction

Rhabdomyosarcomas (RMS) are the most common formof pediatric soft tissue sarcoma, although they are compara-tively rare in adults, comprising only 2–5% of adult soft tis-sue sarcomas. Histologically, RMS resembles developingskeletal muscle and can be divided into two main subtypestermed alveolar and embryonal. An additional subtypetermed pleomorphic rhabdomyosarcoma (PRMS) is de-

scribed, which is generally not found in children, althoughaccounts for more than 50% of the cases of RMS in adultsover 40 years of age [1]. PRMS is found more frequently inmen and predominantly in the large muscles of the extremi-ties. Like the alveolar subtype, PRMS generally has a poorprognosis because it is highly aggressive and frequently me-tastasizes to sites such as the brain and lung [2].

There is little published data on the cytogenetics and mo-lecular genetics of PRMS. The karyotypes reported include

89

�

92,XXYY,

�

2,

�

3,

�

5,

�

14,

�

15,

�

add(17)(p11),

�

22,

�

8

�

10mar,inc[4]/46,XY[2] from a 17-year-old male;

50

�

56,X,

�

X,

�

Y,

�

4,

�

5,add(11)(p15),15,

�

18,

�

18,

�

20,

�

20,

�

22,

�

22,

�

r,inc[cp23] from a 54-year-old male; and 52,XX,add(6)(p24),

�

8,

�

del(10)(q23)

�

2,

�

11,add(18)(q21)

�

2,

�

add(16)(p13),

�

add(19)(q13),

�

20,

�

22,

�

mar[3

�

5]. A sin-gle PRMS cell line, HS-RMS-1, has been established with apseudotetraploid complex and highly incomplete karyotype

* Corresponding author. Molecular Cytogenetics, Haddow Laborato-ries, Institute of Cancer Research, 15 Cotswold Road, Sutton, Surrey SM25NG, UK. Tel.:

�

44-20-8770-4273; fax:

�

44-20-8770-7290.

1

Formerly the Section of Pediatric Oncology, Institute of CancerResearch.

E-mail address

: shipley@icr.ac.uk (J. Shipley).

74

A. Gordon et al. / Cancer Genetics and Cytogenetics 140 (2003) 73–77

79,XY,

�

add(X)(p22),

�

Y,del(1)(p36.1),

�

2,add (2)(p11), add(3)(p21),

�

4,add(4)(p11),add(q35),

�

6,

�

7,add(8)(q24),

�

10,

�

10,i(10)(q10),

�

11,

�

11,

�

13,

�

13,

�

14,add(15)(p11),

�

16,

�

16,add(16)(q22),

�

18,

�

18,

�

19,

�

19,

�

19,

�

22,

�

7mar [6]. Noconsistent structural or numerical abnormalities are apparentfrom these data.

Distinguishing PRMS from other pleomorphic sarcomassuch as malignant fibrous histocytomas (MFH) and pleo-morphic leiomyosarcoma can be difficult [7,8]. Indeed,PRMS may be considered part of the spectrum of MFH[7,8]. A diagnosis of PRMS usually requires positive immun-ostaining with antibodies to desmin, myoglobin, MyoD1, andmore recently myogenin, and the presence of hyperchromaticnuclei with occasional cells showing muscle cross striations[9]. Alveolar RMS is associated with specific translocationsresulting in the fusion genes

PAX3-FOXO1A

and

PAX7-FOXO1A

[10]. Three PRMS cases have been described pre-viously as negative for the presence of these fusion genes byreverse transcriptase polymerase chain reaction (RT-PCR)[11], and two additional cases in a separate study showed noevidence for the

PAX3-FOXO1A

fusion gene [12].In this study we have identified the chromosomal imbal-

ances and determined the

PAX-FOXO1A

fusion gene statusin well-characterized cases of PRMS. The genetic analysishere identifies changes associated with cases classified withthe pleomorphic subtype and also sheds some light on thepossible relationship between rhabdomyosarcoma subtypesand other sarcomas.

2. Materials and methods

Table 1 shows the PRMS samples investigated in thisstudy for which the pathology was reviewed and staining fordesmin, myoglobin, and MyoD1 was demonstrated to con-firm the diagnosis. Four of the cases have been published pre-viously as morphologically typical [9]. Three of the cases(STS 023, 101, and 311) have been published in a study ofPRMS ploidy and its relationship to histology and prognosis(patient numbers 31, 28, and 29, respectively) [13].

To identify the chromosomal imbalances, comparativegenomic hybridization analysis (CGH) was performed es-sentially as described previously, and images were analyzed

using Quips CGH analyzer software (Vysis, Downers Grove,IL, USA) [14,15]. In parallel control self to self-hybridizations,the CGH fluorescence ratios and their standard deviation didnot exceed 0.8–1.2. In the analysis of the tumor DNA, a flu-orescence ratio less than 0.80 was characterized as loss, greaterthan 1.20 as gain, and greater than 1.50 as amplification.

Detection of

PAX3

/

7-FOXO1A

fusion gene transcriptswas undertaken by RT-PCR as described previously [16].Interphase fluorescence in situ hybridization (I-FISH) wasalso performed to deduce the presence or absence of t(2;13)(q35;q14) or t(1;13)(p36;q14) associated with these fusiongene transcripts on all samples except JP9, for which insuf-ficient suitable material was available. Translocations wereidentified as described previously [17], except that in addi-tion to the cosmids Cos1/Cos5 (F. Barr, University of Penn-sylvania, PA, USA) that flank the

FOXO1A

gene, CEPH YAC943 g11 and 875a8 flanking

FOXO1A

were also used. YAC802a9, which lies telomeric to the

PAX3

breakpoint, wasused together with

FOXO1A

YAC 943 g11 and shown to colo-calize in interphase cells when the

PAX3-FOXO1A

fusiongene is present. Some 50–100 interphase images were capturedand scored for each sample. The false-negative rate for 943g11 and 802a9 signals colocalizing on normal control inter-phases was

�

5%. Wherever possible, all FISH and RT-PCR were performed with the investigator blinded to theidentity of the samples.

3. Results

The CGH analysis of the seven PRMS cases is summa-rized in Fig. 1. The smallest overlapping regions of gainwere 1p22

�

p23 (71%), 7p (43%), 18/18p (43%), and 20/20p(43%), and those of loss were 10q23 (71%), 15q21

�

q22(57%), 2q21

�

q35, 3p, 5q32

�

qter, and 13 (all 43%). Sevenamplicons were noted at 1p21

�

p31, 1q21

�

q25, 3p12,3q26

�

qter, 4q28

�

q31, 8q21

�

q23, and 22q, as well asmultiple copies of the whole chromosome arm 8q.

No disruption of FKHR, via colocalization of the 2q35/13q14YAC, nor

PAX3

/

7-FOXO1A

fusion gene transcripts wereseen in samples 023, 323, 101, 309/311, or 187. Sample JP8,however, was positive for the

PAX3-FOXO1A

fusion genetranscript and the t(2;13)(q35;q14) by RT-PCR and FISH,respectively. In the interphase FISH analysis, colocalizationof 2q35/13q14 YAC was found in

�

50% of all informativelarge nuclei examined.

4. Discussion

The patterns of amplification, gain, and loss found herein PRMS were generally different to those reported previ-ously in embryonal and alveolar RMS [15,18–22]. EmbryonalRMS are characterized by frequent gains of chromosomal ma-terial particularly involving chromosomes 2, 7, 8, 12, and 13,of which only gain/amplification of 8q material is in com-mon with the PRMS findings presented here. Gain of 8q isalso a change found in many sarcomas [22]. Embryonal RMS

Table 1PRMS sample studied

CaseSex/age atdiagnosis Tumor site

I-FISH and RT-PCR for

PAX3/7-FOXO1A

STS 023 M/54 Limb —STS 323 M/45 Head and neck —STS 101 F/84 Limb —STS 309/311 M/68 Limb —STS 187 M/84 Buttock —JP8 M/68 Calf

�(PAX3-FOXO1A)JP9 M/78 Thigh ND

Patient details and PAX3/7-FKHRF translocation status.Abbreviations: I-FISH, interphase fluorescence in situ hybridization; ND,

not determined; RT-PCR, reverse transcriptase polymerase chain reaction.

A. Gordon et al. / Cancer Genetics and Cytogenetics 140 (2003) 73–77 75

have only rarely demonstrated genomic amplification events,with the exception of those associated with anaplasia, whichshowed amplicons on 12 q, 15q, and 18q [21]. The alveolarsubtypes have been associated with amplification events mostfrequently involving regions such as 2p24, 12q13�q15, and13q31 [15,18–22]. Four of the seven cases of the PRMSanalyzed here amplified at least one region with ampliconsat 1p21�p31, 1q21�q25, 3p12, 3q26�qter, 4q28�q31,

8q21�q23, and 22q (Fig. 1). The regions of amplification inthis limited but well-characterized series of cases classifiedas PRMS are not coincident with those regions most fre-quently amplified in alveolar or embryonal RMS associatedwith anaplasia.

Amplicons including the 1q21�q25 and 8q21�q23 re-gions found here are described in other sarcoma types [22].Also, gain/amplification at 1p22�p33 has been reported to

Fig. 1. Summary of genomic imbalances from CGH analysis of PRMS samples. Lines to the left of chromosomes represent losses (CGH fluorescence ratios � 0.80),lines to the right of chromosomes represent gains (ratio �1.20), and thick lines represent amplifications or multiple gains (ratio � 1.50) or total deletions(ratio � 0.50). The numbers above each line denote the case number.

76 A. Gordon et al. / Cancer Genetics and Cytogenetics 140 (2003) 73–77

be the most frequent chromosomal change in malignant fi-brous histiocytomas (MFH) (approximately one-third ofcases) and found to be an indicator of poor prognosis in softtissue MFH [23]. The gain of 7p material and the 3p12 am-plicon, noted in case 311, has also been reported in MFHstudies [22,23] but not noted in other tumor types. Thesmallest region of loss at 10q23 was the most frequentchange seen in PRMS, and a �del(10)(q23)�2 has been re-ported in one of the four PRMS karyotypes [3]. This regionis the site of a tumor-suppressor gene, PTEN, and the regionfrequently shown to be deleted in cancers and sarcomas(e.g., in 69% of leiomyosarcomas of soft tissue) [24]. Over-all, the pattern of amplifications and imbalance seen hereshow similarities to those reported for MFH and other sar-comas, particularly osteosarcomas.

Distinguishing PRMS from other pleomorphic sarcomassuch as MFH or pleomorphic leiomyosarcoma can be diffi-cult [7,8]. Indeed, PRMS can be considered part of the spec-trum of MFH although their nosologic status is verified byimmunohistochemical confirmation of their myogenic po-tential [7,8]. All the cases studied here were histopathologi-cally well characterized, and four of these were published ina study that set diagnostic criteria for PRMS [9]. The simi-larities between the genetic changes found in PRMS andthose reported for MFH and other sarcomas, however, mayreflect the difficulties in classification or represent changescommon to various types of sarcomas. Rare pediatric tumorsresemble PRMS although there is usually another componentin addition to pleomorphic cells and their classification is gen-erally subsumed to either anaplastic or spindle-cell RMS [8].The presence of anaplasia has been described in both the em-bryonal and alveolar subtypes of RMS, but anaplastic RMSis currently not a classification used in adult sarcomas and isa developing concept that has not been fully characterized.

To test for a possible link between the alveolar subtypeof RMS and PRMS, the PRMS samples were examined forthe presence of the PAX3/FOXO1A and PAX7/FOXO1A fu-sion genes products and translocations associated with thesethat are frequently found in alveolar RMS. One of the sixsamples analyzed (JP8) by two independent techniquesshowed clear evidence of the presence of a PAX3/FOXO1Afusion gene. Two previous studies involving five PRMS caseshave not found evidence for the PAX3/FOXO1A fusion gene[11,12]. Alveolar RMS has been very rarely reported in the el-derly although, for example, cytogenetic analysis of a tumorfrom a 68-year-old was shown to have a t(2;13)(q35;q14) [25].JP8, however, clearly showed the histologic features ofPRMS and demonstrated a pattern of amplifications, gains,and losses different from the pediatric forms of RMS.

The histopathologic differences between PRMS and thejuvenile forms of RMS have been taken to indicate distinctdisease entities [8]. The CGH evidence presented here, al-beit from a small sample set, also suggests that PRMS maybe a separate disease genetically with a pattern of amplifica-tion, gain, and loss common to other sarcomas, particularlyMFH or osteosarcoma. Evidence for the PAX3-FOXO1A fu-

sion gene in one case suggests a specific similarity between tu-mors classified as alveolar and pleomorphic RMS. However,further identification of the defining molecular features ofPRMS and other sarcomas is required.

References

[1] Hawkins WG, Hoos A, Antonescu CR, Urist MJ, Leung DH, GoldJS, Woodruff JM, Lewis JJ, Brennan MF. Clinicopathologic analysisof patients with adult rhabdomyosarcoma. Cancer 2001;91:794–803.

[2] Enzinger F, Weiss S. Soft tissue tumours. St. Louis: Mosby, 1994.[3] Douglass EC, Valentine M, Etcubanas E, Parham D, Webber BL,

Houghton PJ, Houghton JA, Green AA. A specific chromosomal abnor-mality in rhabdomyosarcoma. Cytogenet Cell Genet 1987;45:148–55.

[4] Kullendorff CM, Donner M, Mertens F, Mandahl N. Chromosomalaberrations in a consecutive series of childhood rhabdomyosarcoma.Med Pediatr Oncol 1998;30:156–9.

[5] Mertens F, Fletcher CD, Dal Cin P, De Wever I, Mandahl N, Mitel-man F, Rosai J, Rydholm A, Sciot R, Tallini G, Van den Berghe H,Vanni R, Willen H. Cytogenetic analysis of 46 pleomorphic soft tis-sue sarcomas and correlation with morphologic and clinical features:a report of the CHAMP Study Group. Chromosomes and morphol-ogy. Genes Chromosomes Cancer 1998;22:16–25.

[6] Sonobe H, Takeuchi T, Taguchi T, Shimizu K, Furihata M, OhtsukiY. A new human pleomorphic rhabdomyosarcoma cell-line, HS-RMS-1, exhibiting MyoD1 and myogenin. Int J Oncol 2000;17:119–25.

[7] Hollowood K, Fletcher CD. Rhabdomyosarcoma in adults. SeminDiagn Pathol 1994;11:47–57.

[8] Parham D. Pathological classification of rhabdomyosarcomas andcorrelations with molecular studies. Mod Pathol 2001;14:506–14.

[9] Gaffney EF, Dervan PA, Fletcher CD. Pleomorphic rhabdomyosar-coma in adulthood. Analysis of 11 cases with definition of diagnosticcriteria. Am J Surg Pathol 1993;17:601–9.

[10] Anderson J, Gordon A, Pritchard-Jones K, Shipley J. Genes, chromo-somes, and rhabdomyosarcoma. Genes Chromosomes Cancer 1999;26:275–85.

[11] Downing JR, Khandekar A, Shurtleff SA, Head DR, Parham DM,Webber BL, Pappo AS, Hulshof MG, Conn WP, Shapiro DN.Multiplex RT-PCR assay for the differential diagnosis of alveolarrhabdomyosarcoma and Ewing’s sarcoma. Am J Pathol 1995;146:626–34.

[12] Chen BF, Chen ML, Liang DC, Huang YW, Liu HC, Chen SH. Detec-tion of PAX3-FKHR and PAX7-FKHR fusion transcripts in rhabdomyo-sarcoma by reverse transcriptase-polymerase chain reaction using paraf-fin-embedded tissue. Zhonghua Yi Xue Za Zhi (Taipei) 1999;62:86–91.

[13] Kilpatrick SE, Teot LA, Geisinger KR, Martin PL, Shumate DK, Zbi-eranski N, Russell GB, Fletcher CD. Relationship of DNA ploidy tohistology and prognosis in rhabdomyosarcoma. Comparison of flowcytometry and image analysis. Cancer 1994;74:3227–33.

[14] Kallioniemi OP, Kallioniemi A, Piper J, Isola J, Waldman FM, GrayJW, Pinkel D. Optimizing comparative genomic hybridization foranalysis of DNA sequence copy number changes in solid tumors.Genes Chromosomes Cancer 1994;10:231–43.

[15] Weber-Hall S, Anderson J, McManus A, Abe S, Nojima T, PinkertonR, Pritchard-Jones K, Shipley J. Gains, losses, and amplification ofgenomic material in rhabdomyosarcoma analyzed by comparative ge-nomic hybridization. Cancer Res 1996;56:3220–4.

[16] Anderson J, Renshaw J, McManus A, Carter R, Mitchell C, Adams S,Pritchard-Jones K. Amplification of the t(2;13) and t(1;13) transloca-tions of alveolar rhabdomyosarcoma in small formalin-fixed biopsiesusing a modified reverse transcriptase polymerase chain reaction. AmJ Pathol 1997;150:477–82.

[17] McManus AP, O’Reilly MA, Jones KP, Gusterson BA, Mitchell CD,Pinkerton CR, Shipley JM. Interphase fluorescence in situ hybridization

A. Gordon et al. / Cancer Genetics and Cytogenetics 140 (2003) 73–77 77

detection of t(2;13)(q35;q14) in alveolar rhabdomyosarcoma—a diag-nostic tool in minimally invasive biopsies. J Pathol 1996;178:410–4.

[18] Pandita A, Zielenska M, Thorner P, Bayani J, Godbout R, GreenbergM, Squire JA. Application of comparative genomic hybridization,spectral karyotyping, and microarray analysis in the identification ofsubtype-specific patterns of genomic changes in rhabdomyosarcoma.Neoplasia 1999;1:262–75.

[19] Bridge JA, Liu J, Weibolt V, Baker KS, Perry D, Kruger R, QualmanS, Barr F, Sorensen P, Triche T, Suijkerbuijk R. Novel genomic im-balances in embryonal rhabdomyosarcoma revealed by comparativegenomic hybridization and fluorescence in situ hybridization: an in-tergroup rhabdomyosarcoma study. Genes Chromosomes Cancer2000;27:337–44.

[20] Bridge JA, Liu J, Qualman S, Suijkerbuijk R, Wenger G, Zhang J,Wan X, Baker K, Sorenson P, Barr F. Genomic gains and losses aresimilar in genetic histologic subsets of rhabdomyosarcomas and am-plification predominates in embryonal with anaplasia and alveolarsubtypes. Genes Chromosomes Cancer 2002;33:310–21.

[21] Gordon AT, Brinkschmidt C, Anderson J, Coleman N, Dockhorn-

Dworniczak B, Pritchard-Jones K, Shipley J. A novel and consistentamplicon at 13q31 associated with alveolar rhabdomyosarcoma.Genes Chromosomes Cancer 2000;28:220–6.

[22] Knuutila S, Bjorkqvist AM, Autio K, Tarkkanen M, Wolf M, MonniO, Szymanska J, Larramendy ML, Tapper J, Pere H, El-Rifai W,Hemmer S, Wasenius VM, Vidgren V, Zhu Y. DNA copy numberamplifications in human neoplasms: review of comparative genomichybridization studies. Am J Pathol 1998;152:1107–23.

[23] Larramendy ML, Tarkkanen M, Blomqvist C, Virolainen M, Wik-lund T, Asko-Seljavaara S, Elomaa I, Knuutila S. Comparative ge-nomic hybridization of malignant fibrous histiocytoma reveals anovel prognostic marker. Am J Pathol 1997;151:1153–61.

[24] El-Rifai W, Sarlomo-Rikala M, Knuutila S, Miettinen M. DNA copynumber changes in development and progression in leiomyosarcomasof soft tissues. Am J Pathol 1998;153:985–90.

[25] Stindl R, Fiegl M, Regele H, Gisslinger H, Breitenseher MJ, FonatschC. Alveolar rhabdomyosarcoma in a 68-year-old patient identified bycytogenetic analysis of bone marrow. Cancer Genet Cytogenet 1998;107:43–7.