Chromosomal aberrations in malignant gastrointestinal stromal tumors

Cancer Genetics and Cytogenetics 128 (2001) 24–30

Chromosomal aberrations in malignant gastrointestinal stromal tumors: correlation with c-KIT gene mutation1 Maria Debiec-Rychtera,2, Jerzy Lasotaa,*, Maarit Sarlomo-Rikalab, Radzislaw Kordekc, Markku Miettinena,* a

Department of Soft Tissue Pathology, Armed Forces Institute of Pathology, Washington, DC, USA b Department of Pathology, Haartman Institute of the University of Helsinki, Helsinki, Finland c Chair of Oncology, Medical University of Lodz, Lodz, Poland Received 8 September 2000; accepted 13 December 2000

Abstract

Gastrointestinal stromal tumors (GISTs) are distinctive, KIT positive mesenchymal neoplasms. The genetic alterations leading to the malignant behavior of these tumors are not well known. In this study, we looked for recurrent numerical chromosomal changes, which may be associated with malignant GISTs, using interphase fluorescence in situ hybridization (FISH). Fourteen malignant primary tumors and two intra-abdominal recurrences were analyzed. Nine benign tumors were studied for comparison. In all cases, the presence of mutations in exons 9, 11 and 13 of the KIT gene were evaluated. Sixteen centromeric enumeration probes (CEP) for chromosomes 1, 2, 3, 4, 6, 7, 8, 9, 10, 11, 12, 15, 16, 17, 18, and X and three locus specific probes (LSI) for 22q11.2 (BCRlocus), 13q14 (RB1-locus) and 14q32 (IgH-locus) were used. The most common changes seen in malignant GISTs were losses of 14q32 and 22q11. However, these changes were commonly detected in benign tumors and represent early changes related to the pathogenesis of GISTs. Losses of chromosomes 1 and 9 were the only recurrent numerical changes seen exclusively in malignant GISTs. Other recurrent numerical changes seen predominantly in malignant tumors were gain of chromosome 8 and losses of chromosomes 7 and 15. The concurrent loss of chromosome 7 and gain of chromosome 8 (in 4 cases) was never seen together with loss of chromosomes 9 or 15 and only once with loss of chromosome 1. Mutations in KIT were found in the majority of malignant GISTs (64%) confirming a previously shown correlation between presence of such mutations and malignancy. KIT mutations were seen in four of five malignant GISTs with loss of chromosome 9, but only in one of four malignant tumors with loss of chromosome 1. These observations may reflect the different pathways leading to malignant transformation of GISTs. © 2001 Elsevier Science Inc. All rights reserved.

1. Introduction Gastrointestinal stromal tumors (GISTs), previously classified as gastrointestinal smooth muscle tumors (leiomyomas, leiomyoblastomas and leiomyosarcomas), are distinctive KIT positive mesenchymal neoplasms [1]. The

* Corresponding author. Tel.: 202-782-2792 or 2793; fax: 202-7829182. 1 The opinions and assertions contained herein are the expressed views of the authors and are not to be construed as official or reflecting the views of the Departments of the Army or Defense. This study was partially supported by the Grant from the Polish Committee for the Scientific Research (4PO5A07117) and by the American Registry of Pathology. 2 Research fellow (1998–1999) at the Department of Soft Tissue Pathology, AFIP, Washington, DC.

gain-of-function KIT mutations have been described in GISTs [2]. Approximately 50% of GISTs show KIT mutations [3–5]. The presence of such mutations has been shown as an independent adverse prognostic factor [6]. However, the genetic changes leading to the malignant transformation of GISTs are not well understood. Previous cytogenetic data on “malignant gastrointestinal smooth muscle tumors,” probably including GISTs, have shown losses of chromosomes 14, 15, 18, 22 and chromosome arm 1p as the most frequent aberrations [7–10]. Cytogenetic studies of several benign and malignant GISTs have revealed clonal losses of chromosomes 14, 22, 9 and 1p [11–13]. Comparative genomic hybridization (CGH), which allows for global DNA-copy number analysis of archival tissue revealed recurrent losses of 14q in both benign and malignant

0165-4608/01/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved. PII: S0165-4608(01)00 3 9 5 - 8

M. Debiec-Rychter et al. / Cancer Genetics and Cytogenetics 128 (2001) 24–30

GISTs, and have confirmed the cytogenetically observed losses in 1p, 15 and 22 as frequent chromosomal changes [14,15]. A combined CGH and loss of heterozygosity (LOH) study revealed loss of chromosome 9 and homozygous deletion of 9p markers flanking the p16INK4A gene in two malignant GISTs [16]. LOH at 1p36 was found in 30% of GISTs and it was suggested that deletion of the gene/genes in this region may account for malignant behavior [17]. More recently, gains at 5p, 8q, 17q and 20q and losses of 9p, 13q, 15q and 19q seen in an extensive CGH study were suggested as new markers of malignant transformation in GISTs [18]. The purpose of this study was to identify recurrent chromosomal changes in GISTs, which may be associated with malignant behavior. We analyzed a series of benign and malignant GISTs by interphase FISH using a comprehensive selection of probes. We have evaluated the FISH findings with the presence of KIT mutations in relation to malignant transformation. The discrepancies between our FISH based study and previously published CGH base studies are discussed. 2. Materials and methods 2.1. Tissue material Twenty-two primary GISTs and two intra-abdominal recurrences of which fresh or ethanol fixed tissue was available were obtained from the files of the Haartman Institute of the University of Helsinki, Helsinki, Finland and Chair of Oncology, Medical University of Lodz, Lodz, Poland. 2.2. Immunohistochemistry Antibodies to the following antigens were used: c-KIT pro-oncogene product (CD117), a polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA, Catalog No sc-168, at 1:400), CD34 (QBEND-10, monoclonal, from Biogenex, diluted at 1:50) -smooth muscle actin (1A4 monoclonal, Sigma, at 1:1600), desmin (D33, monoclonal antibody from Dako, at 1:50), and S100-protein (polyclonal antibody, Dako, at 1:1600). Studies were performed on formalin-fixed and paraffin-embedded tissue by using the avidin biotin peroxidase complex (ABC) detection system with diaminobenzidine as the chromogen. Heat-induced epitope retrieval (20 min in 0.1 mM/L citrate buffer, pH 6.0) was performed for CD117 and CD34.

25

DNA FISH-probes LSI 22q11.2 (BCR-locus) and LSI 13q14 (RB1-locus) (Vysis) were used in this study. In addition, an indirectly labeled (Nick Translation System, GIBCO BRL, Gaithersburg, MD, USA), biotin-conjugated 14q32 (IgH-locus specific) cosmid probe was used. And 10 l of directly labeled CEP and LSI probes was used in hybridization. The probes were assessed individually. The directly labeled probes were denatured with target DNA at 85C for 1 min and hybridized 16 h at 42C (CEP) or at 37C (LSI). After hybridization, two 2 min washing was performed, first in 0.4%  SSC/0.3% Nonidet P-40 at 73C and second in 2  SSC/0.1% Nonidet P-40 at room temperature. For the indirectly labeled IgH-specific probe, DNA was denatured by immersion of slides in 70% deionized formamide/2  SSC (pH 7.0) at 70C for 4 min and dehydrated in ethanol. And 10 l of hybridization mixture containing 50% deionized formamide/2SSC, 10% dextran sulfate, 500 ng/l human Cot-1 DNA and 6 ng/l of indirectly labeled probe DNA was denatured at 75C for 5 min. After brief cooling on ice, the mixture was preannealed at 37C for 1 h and used for hybridization (16 h at 42C). The immunohistochemical detection consisted of three 30 min incubations at 37C, first incubation with 5 g fluorescein avidin DCS/ml (Vector Laboratories, Burlingame, CA, USA) in 4SSC/0.1% Tween 20/1% bovine serum albumin (BSA) was followed by biotinylated anti-avidin DCS/ml (Vector Laboratories; 1: 100 in 4SSC/0.1% Tween 20/1% BSA) and another application of fluorescein avidin. Each incubation step was followed by the washes in 4SSC/0.1% Tween (310 min). Dehydrated and air-dried nuclei were counterstained with 0.4 g/ml propidium iodide or 1 g/ml 4,6-diamidino2-phenylindole dihydrochloride (DAPI) in the antifade p-phenylene-diamine dihydrochloride solution. The slides were evaluated using an Olympus BX60 epifluorescence microscope with a triple-pass filter DAPI/Green/Orange (Vysis). The color images were captured by SPOT Cooled Color Digital Camera (Diagnostic Instruments, Inc., USA). The Image-Pro PLUS software (Media Cybernetics, http:// www.mediacy.com) was used for photographic documentation. In each case, 200 intact, non-overlapping nuclei with well-defined nuclear outlines were chosen for scoring fluorescent signals. The percentage of tumors nuclei containing 0, 1, 2, 3, 4 and 4 signals was captured for each probe. To define the distribution of interphase FISH signals, normal myometrial and normal small intestinal tissues were used for interphase FISH normal value studies.

2.3. FISH analysis

2.4. Normal value study and criteria for FISH anomalies

Standard 5-m sections of ethanol-fixed paraffinembedded tissues were prepared for hybridization according to Vysis (Vysis Inc., Downers Grove, IL, USA) protocol developed for FISH of formaldehyde fixed paraffin embedded tissues. Sixteen directly labeled CEP chromosome enumeration DNA FISH-probes 1, 2, 3, 4, 6, 7, 8, 9, 10, 11, 12, 15, 16, 17, 18, X and two directly labeled LSI unique sequence

In control tissues, the mean percentage of nuclei with two signals using all centromeric probes varied from 66.7% to 75.5%, and it ranged from 50.3% to 65.1% using BRClocus specific probe. The ranges for mean percentage of nuclei with zero or one, three, and more than three signals for all CEP probes were 23.1% to 31.4%, 1.3% to 2.7% and 0.1% to 1.2%, respectively, and for locus specific probe was

26

M. Debiec-Rychter et al. / Cancer Genetics and Cytogenetics 128 (2001) 24–30

32.4% to 46.9%, 0.2% to 3.9% and 0.1% to 1.8%, respectively. The mean  3SD percentage of nuclei with zero or one signal was 55.0% and 65.0% for all CEP and locus specific probes, respectively. The mean  3 SD percentage of nuclei with three or more signals was 10.0% for all probes. Based on the normal values, we applied the following conservative criteria for FISH anomalies: (1) abnormal gain required 10% nuclei with three or more signals for all CEP and LSI or IgH-specific probes, and autosomal trisomy was diagnosed if the number of cells with three signals was greater than with four signals; (2) autosomal loss required 55% nuclei with zero or one signal for each CEP probes and 65% nuclei with zero or one signal for LSI or IgH-specific probes; (3) abnormal tetrasomy required the autosomal average of the percentage of cells with three and four signals to be 6%; and (4) abnormal polysomy required the autosomal average of the percentage of cells with five or more signals to be 5%. 2.5. Molecular studies DNA for the PCR amplification was obtained from formalin-fixed and paraffin-embedded tissue and the exons 9, 11 and 13 of c-KIT gene were PCR amplified as previously reported [4]. The PCR products were size fractionated on 2% agarose gels, purified form the gels (Qiagen Inc., Chatsworth, CA, USA) and sequenced directly on a 373 DNAsequencer (Applied Biosystems, Foster City, CA, USA) using forward and reverse primers. Computer analysis of the DNA sequences were performed using Lasergene software (DNASTAR, Madison, WI, USA) in connection with the data of the GeneBank 110/EMBL57 database. 3. Results

showed death from tumor in five cases and evidence of recurrent or metastatic disease in six cases. 3.2. Specific chromosome/allelic changes by FISH Chromosomal/allelic changes in GISTs are shown in Table 1, and the most common changes summarized in Table 2. Representative examples of interphase FISH results are illustrated in Fig. 1. All GIST samples showed imbalances, with chromosome/allelic losses occurring more frequently than gains (2.8 losses/case vs. 1.5 gains/case). Losses of chromosome 14q32 (in 20 cases; 87%) and 22q11 (in 12 cases; 48%), followed by in order of decreasing frequency, gain of chromosome 8 (in 8 cases; 35%), losses of chromosomes 7 and 15 (both in 6 cases; 26%), loss of chromosome 9 (26%) and gains of chromosomes 3 and 4 (both in 5 cases; 22%), and loss of chromosome 1 and gain of chromosome 10 (both in four cases; 17%) were the most frequent anomalies. Other changes were detected in less than four specimens. There was no significant difference in the incidence of 14q32 and 22q11 losses between benign and malignant GISTs (8/9 and 12/14 vs. 5/9 and 7/14 tumors, respectively). Similarly, gains of chromosomes 3, 4 and 10 were equally distributed in benign and malignant GIST subgroups. In contrast, numeric aberrations of chromosomes 1, 7, 8, 9 and 15 revealed significant differences of frequency between benign and malignant GIST specimens. Chromosomes 1 and 9 losses were noticed only in malignant tumors. Likewise, five out of six chromosomes 7 and 15 losses and seven out of eight chromosome 8 gains were found in malignant GISTs. Additionally, the concurrent loss of chromosome 7 and gain of chromosome 8 were noticed in four malignant tumors. Loss of RB1-locus was observed only in one malignant GIST.

3.1. Clinicopathological features

3.3. Ploidy analysis by FISH

Demographic and clinical data and histological features of all GISTs analyzed in this study are summarized in Table 1. Twenty-two primary GISTs and two intra-abdominal recurrences were analyzed. There were 14 gastric, 5 small intestinal, 2 rectal and 1 mesenteric primary tumors. In one case, the primary gastric (case 12) and a recurrent lesion (sample 12a) were analyzed. Histologically, 17 GISTs had a predominant spindle cell pattern, and 7 had an epithelioid morphology. CD117immunoreactivity was shown in all cases. The majority of cases (79%) were positive for CD34, and -smooth muscle actin was present in eight cases (29%), usually focally. All cases were negative for desmin and S100-protein. Nine GISTs, were classified as histologically benign based on low if any mitotic activity ( 1 mitoses/10 HPF). Follow-up available in all cases did not show intra-abdominal recurrences (IAR) or metastases. Thirteen GISTs were classified as histologically malignant based on mitotic activity (more than 4 mitoses/10 HPF). Follow-up available in all cases

Of nine benign GISTs, eight were near-diploid (89.0%) and one was polyploid with a mean autosomal polysomy (5 signals) in 11% of the tumor cells. Of 15 malignant GISTs, 10 were near-diploid (67%), 3 were polyploid (with a mean autosomal polysomy of 9.7–13.8%) and 2 were tetraploid (with a mean autosomal tetrasomy of 9.5–10.5%). One of the latter was recurrent tumor; the primary tumor in this case was identified as being diploid. 3.4. C-KIT mutation status The exons 9, 11 and 13 of the c-KIT gene were evaluated for the presence of mutations. Duplication of Ala502 and Tyr503 (exon 9) was found in one case. In 11 GISTs point mutations or several codon deletions were identified in the proximal part of exon 11. No mutations were found in exon 13. In summary, c-KIT mutations were found in 48% of analyzed cases; in 3 of 9 benign and 8 of 14 malignant GISTs. The KIT mutation status of all analyzed cases is shown in the context of the clinical and genetic data in Table 1.

79/F 63/M 64/F 40/M 75/F 61/F 34/M 69/M 52/M 46/M 67/F 47/F 48/F 69/F 63/M 41/M 41/M 71/M 54/M 85/M 63/M 60/M 50/M 71/M

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

Stomach Stomach Stomach Stomach Stomach Stomach Stomach Small bowel Small bowel Stomach Stomach Stomach IAR Stomach Stomach Stomach Stomach Small bowel Small bowel Small bowel Rectum Rectum Mesentery IAR

Location

Spindle Spindle Spindle Epithelioid Spindle Epithelioid Spindle Spindle Spindle Epithelioid Spindle Spindle Spindle Spindle Spindle Epithelioid Spindle Epithelioid Spindle Spindle Spindle Spindle Epithelioid Epithelioid

Cell type 3.5 3.5 3.6 5 4.4 55 6.5 2.52.02.0 776.5 75.54.5 9cm 1065 6 12 19127 30cm 3918 10cm 15136 30305 7 10-12 cm 302513 9 8 30 30 13 15 15 13 10 2

1

1

1

1

1

1

1

1

1 11 3 11

Mitosis/10HPF Ex 11 M WT WTa WTa Ex 11 M WTa Ex 11 M WT Ex 11 M WT Ex 11 M Ex 11 Ma Ex 11 M Ex 11 Ma Ex 11 Ma WT Ex 11 M WT WTa Ex 11 Ma Ex 11 Ma WT Ex 11 M Ex9 INS

KIT mutation

3, 7, 8 2, 4, 9

3, 8, 11, 12, 18 7, 8 6, 9, 15

6, 9, 15, 18 1, 6, 15, 16, 17

8

3, 7, X 1, 4, 7, 8, 10

4, 7, 8 1, X 1, 9, 15 9, 10, 11, 11, 15

2, 4, 8, 9, X

( )RB1

( )IgH ( )IgH ( )IgH ( )IgH ( )IgH ( )IgH ( )IgH ( )IgH

( )IgH ( )IgH

( )IgH ( )IgH ( )IgH ( )IgH ( )IgH ( )IgH ( )IgH ( )IgH

11 10, 15, 16

8, 9, 18, X

4, 7

3, 4, 17

14q32 ( )IgH ( )IgH

13q14

( )BCR

( )BCR ( )BCR ( )BCR

( )BCR ( )BCR ( )BCR

( )BCR

( )BCR ( )BCR ( )BCR ( )BCR

22q11.2

Allelic losses/gains by FISH using locus specific probes

2, 6, 10

3, 10

Chromosome copy number by FISH using centromeric probes

Diploid Diploid Diploid Diploid Diploid Diploid Polyploid Diploid Diploid Polyploid Diploid Diploid Tetraploid Diploid Diploid Diploid Diploid Tetraploid Diploid Diploid Polyploid Diploid Polyploid Diploid

Ploidy level by FISH

NED 60 NED 41 NED 21 NED 34 NED 38 NED35 NED 32 NED 14 NED 27 AWD 21 NED 6 AWD 15 AWD 48 NED 27 AWD 27 DOD 1 AWD 16 DOD 5 AWD 8 DOD 16 NED 20 NED 9 DOD 24 DOD 36

Status and follow-up (months)

Note that sample 12a represents intra-abdominal recurrence of case 12. Ex 9 or 11 indicates KIT exons. aIndicates that mutation status of KIT was previously published [4]. Chromosomal losses in more than 70% of cells and chromosomal gains in more than 20% are indicated by bold numbers or characters. Abbreviations: IAR: intra-abdominal recurrence; AWD: alive with disease; NED: no evidence of disease; M: mutation; WT: wild type.

Age/Sex

Case

Tumor size (cm)

Table 1 Summary of demographic, clinical, histological and interphase FISH data in GISTs M. Debiec-Rychter et al. / Cancer Genetics and Cytogenetics 128 (2001) 24–30 27

28

M. Debiec-Rychter et al. / Cancer Genetics and Cytogenetics 128 (2001) 24–30

Table 2 Summary of the most common chromosomal/allelic changes in GISTs. Gastrointestinal stromal tumors (GISTs) Chromosome/allelic changes

All cases (23)

Benign (9)

Malignant (14)

1 7 9 15

3

4

8

10 13q14 (RB1) 14q32 (IgH) 22q11.2 (BCR)

4 (17%) 6 (26%) 5 (22%) 6 (26%) 5 (22%) 5 (22%) 8 (35%) 4 (17%) 1 (4%) 20 (87%) 12 (48%)

— 1 (11%) — 1 (11%) 2 (22%) 2 (22%) 1 (11%) 2 (22%) — 8 (89%) 5 (55%)

4 (29%) 5 (36%) 5 (36%) 5 (36%) 3 (21%) 3 (21%) 7 (50%) 2 (14%) 1 (7%) 12 (86%) 7 (50%)

4. Discussion Gastrointestinal stromal tumors are distinctive KIT positive mesenchymal neoplasms previously classified as smooth muscle tumors [1]. The karyotypes of gastrointestinal leiomyosarcomas (many of them probably representing GISTs), use to be analyzed and reported together with “true” leiomyosarcomas (LMSs) [9,10]. This makes it difficult or impossible to specifically identify the cytogenetic profiles of GISTs and “true” LMSs. Current cytogenetic data on tumors diagnosed as GISTs are scant [11–13]. Recently, additional cytogenetic observations have been made by CGH, which allows for a global survey of DNA copy number changes in archival tissue [14,15]. The genetic changes seen, gains at 5p, 8q, 17q and 20q and losses of 9p, 13q, 15q and 19q, suggested these as markers of malignant transformation in GISTs. Gain-of-function mutations in exon 11 of c-KIT [2] have been found in approximately 50% of GISTs, mostly in malignant tumors [3–5]. The presence of such mutations has been shown as an independent prognostic marker [6]. Recently, mutations in exons 9 and 13 of the c-KIT gene have been reported in GISTs [20] and correlation with malignant behavior was suggested [19]. In this study, we investigated recurrent chromosomal changes in GISTs, in relation to malignant behavior. We analyzed a series of benign and malignant GISTs by interphase FISH using a comprehensive selection of probes. The FISH data were compared with the presence of KIT mutations. The most frequent chromosome aberrations among the 24 GISTs studied were loss of 14q32 and 22q11.2 seen with nearly equal frequency in both benign and malignant GISTs suggesting these as early cytogenetic changes. This observation, combined with the previously reported cytogenetic and CGH-studies, suggest common losses of chromosomes 14 and 22 in GISTs [7–12,14,15]. However, we could not confirm recently published CGH based observation that losses

of 1p may be an early event frequently seen (approximately 30% of cases) in benign tumors [18]. Rather, the loss of chromosome 1 was seen in our group exclusively in malignant tumors. A similar percentage of malignant tumors showed the loss of chromosome 1 in our FISH study and loss of entire 1p in a recently published CGH study [18]. Also, gain of chromosomes 3 and 4 was seen in both benign and malignant tumors and may therefore represent an early cytogenetic change in GIST tumorigenesis. Because the number of cases showing these changes was relatively small, further studies are required to confirm these observations. Another numerical chromosomal change seen in the current study was a gain of chromosome 8 mainly in malignant primary tumors and their intra-abdominal recurrences. Similar findings were obtained in the CGH studies [14,18]. Trisomy 8 is a well-recognized secondary abnormality in a number of other mesenchymal neoplasms, such as Ewing sarcoma, dermatofibrosarcoma protuberans and clear-cell sarcoma [9] and usually coincident with tumor progression. The loss of chromosome 7 (5/15 samples) was also mainly seen in malignant GISTs. In one case studied, loss of chromosome 7 was detected in an intra-abdominal recurrence but not in the primary tumor. This observation supports the idea that loss of chromosome 7 is a marker for tumor progression. However, this observation differs from that of a previously published CGH study, which suggested gain rather than loss of genetic material from chromosome 7 [17]. In four cases, loss of chromosome 7 was associated with gain of chromosome 8. The concurrent loss of chromosome 7 and gain of chromosome 8 was not seen together with loss of chromosome 9 or 15 and only once with loss of chromosome 1. Losses of chromosomes 1, 9 and 15 were almost exclusively seen in malignant GISTs. These changes were occasionally reported in GIST-karyotypes [9,11,21], and losses of only chromosomes 1 and 15 but not 9 were initially reported by CGH [14]. Recently, a combined CGH and LOH study revealed loss of chromosome 9 and homozygous deletion of 9p LOH markers flanking p16INK4A gene in two malignant GISTs [16]. More recently, losses in chromosome arm 9p were shown exclusively in malignant GISTs, both primary and metastatic tumors but not in benign GISTs [18]. In a recent CGH study, loss of 13q were reported in more than 40% of malignant GISTs and was listed among genetic alterations correlating with malignant transformation in GISTs [17]. In previously published cytogenetic and molecular genetic studies, loss of 13q and deletion of the RB-1 tumor suppressor gene were commonly detected in “true” leiomyosarcomas [22,23]. In our study, loss of 13q14 (RB1locus) was seen only in 1 malignant tumor (less than 10% of malignant GISTs). There are several possible explanations of this discrepancy. Since the minimal common overlapping region of deletion was defined by CGH between 13q21–qter it is possible that our probe, located centromeric to this re-

M. Debiec-Rychter et al. / Cancer Genetics and Cytogenetics 128 (2001) 24–30

29

Fig. 1. Representative images on FISH analysis of malignant GISTs using directly labeled spectrum green or spectrum orange centromeric (A, B, D, F) and locus specific (E) probes (Vysis) and indirectly labeled IgH-specific probe (C). One signal indicates one copy of chromosome 1 (A), one copy of chromosome 7 (B), loss of 14q32 (C), one copy of chromosome 15 (D), loss of 22q11.2 (E). Three signals indicate three copies of chromosome 8 (F).

gion, failed to detect specific loses, although 12 of 15 malignant GISTs reported by CGH showed losses of the entire 13q [18]. Possible retention of RB1 locus in GISTs should also be considered. Extensive FISH and LOH studies are required to confirmed CGH results regarding loss of 13q and malignant transformation in GISTs. In this study, KIT mutations were found in 30% of benign and in 60% of malignant GISTs, confirming a previ-

ously shown correlation between presence of such mutations and malignancy [3,4,6,19]. There was clear association with malignancy, KIT mutations and loss of chromosome 9. KIT mutations were seen in 4 of 5 malignant GISTs with loss of chromosome 9. Malignant tumors with loss of chromosome 1 predominantly showed lack of KIT mutations. These observations may reflect different progression pathways in GISTs. A recent study showed KIT

30

M. Debiec-Rychter et al. / Cancer Genetics and Cytogenetics 128 (2001) 24–30

mutations in only 25% (3 of 12) of GISTs with LOH at 1p36 [17]. Also, LOH at 1p36 has been shown to be independent prognostic marker, which correlates with poor prognosis in GISTs [17]. In summary, our study based on interphase FISH confirmed previous CGH studies showing that losses of chromosomes 9 and 15 and gain of chromosome 8 are mainly seen in malignant GISTs and may represent changes related to the malignant transformation. However, discrepancies remain regarding the correlation between losses of chromosomes 1, 7 and 13 and malignant behavior of GISTs. The differences probably reflect the complex genetic alterations that occur during malignant transformation and metastatic progression of GISTs. Studies involving a spectrum of different techniques are required to address this problem.

[8]

[9] [10]

[11]

[12]

[13]

Acknowledgments The authors wish to thank Dr. I. Wlodarska from the Centrum for Human Genetics, University of Leuven, Belgium, for a generous gift of IgH specific cosmid probe. This study was partially supported by the Grant from the Polish Committee for the Scientific Research (4PO5A07117).

[14]

[15]

[16]

References [1] Miettinen M, Sarlomo-Rikala M, Lasota J. Gastrointestinal stromal tumors - recent advances in understanding of their biology. Hum Pathol 1999;30:1213–30. [2] Hirota S, Isozaki K, Moriyama Y, Hashimoto K, Nishida T, Ishiguro S, Kawano K, Hanada M, Kurata A, Takeda M, Tunio GM, Matsuzawa Y, Kanakura Y, Shinomura Y, Kitamura Y. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 1998;279:577–80. [3] Ernst SI, Hubbs AE, Przygodzki RM, Emory TS, Sobin LH, O’Leary TJ. KIT mutation portends poor prognosis in gastrointestinal stromal/ smooth muscle tumors. Lab Invest 1998;78:1633–6. [4] Lasota J, Jasinski M, Sarlomo-Rikala M, Miettinen M. Mutations in exon 11 of c-kit occur preferentially in malignant versus benign gastrointestinal stromal tumors and do not occur in leiomyomas or leiomyosarcomas. Am J Pathol 1999;154:53–60. [5] Moskaluk CA, Tian Q, Marshall CR, Rumpel CA, Franquemont DW, Frierson Jr HF. Mutations of c-kit JM domain are found in a minority of human gastrointestinal stromal tumors. Oncogene 1999;18:1897– 902. [6] Taniguchi M, Nishida T, Hirota S, Isozaki K, Ito T, Nomura T, Matsuda H, Kitamura Y. Effect of c-kit mutation on prognosis of gastrointestinal stromal tumors. Cancer Res 1999;59:4297–300. [7] Bardi G, Johansson B, Pandis N, Heim S, Mandahl N, Bak-Jensen E, Frederiksen H, Andren-Sandberg A, Mitelman F. Recurrent chromo-

[17]

[18]

[19]

[20]

[21] [22]

[23]

some aberrations in abdominal smooth muscle tumors. Cancer Genet Cytogenet 1992;62:43–6. Dal Cin P, Aly MS, De Wever I, van Damme B, van den Berghe H. Does chromosome investigation discriminate between benign and malignant gastrointestinal leiomyomatous tumors. Diagn Oncol 1989;2:55–9. Sandberg AA, Bridge JA. The cytogenetics of bone and soft tissue tumors. Austin, TX: RG Landes, 1994. p. 35–6. Saunders AL, Meloni Am, Chen Z, Sandberg AA, Lauwers GY. Two cases of low-grade gastric leiomyosarcoma with monosomy 14 as the only change. Cancer Genet Cytogenet 1996;90:184–5. Bergmann I, Gunawan B, Hermanns B, Hoer J, Schumpelick V, Fuzesi L. Cytogenetic and morphologic characteristics of gastrointestinal stromal tumors. Recurrent rearrangement of chromosome 1 and losses of chromosomes 14 and 22 as common anomalies. Verh Dtsch Ges Pathol 1998;82:275–8. Breiner JA, Meis-Kindblom J, Kindblom LG, Nelson M, Bridge JA. Loss of 14q and 22q in gastrointestinal stromal cell tumors (Pacemaker cell tumors). Cancer Genet Cytogenet 2000;120:111–6. Marci V, Casorzo L, Sarotto I, Dogliani N, Milazzo MG, Risio M. Gastrointestinal stromal tumor, uncommitted type, with monosomies 14 and 22 as the only chromosomal abnormalities. Cancer Genet Cytogenet 1998;102:135–8. El-Rifai W, Sarlomo-Rikala M, Miettinen M, Knuutila S, Andersson LC. DNA copy number losses in chromosome 14: an early change in gastrointestinal stromal tumors. Cancer Res 1996;56:3230–3. Sarlomo-Rikala M, El-Rifai W, Andersson L, Miettinen M, Knuutila S. Different patterns of DNA copy number changes in gastrointestinal stromal tumors, leiomyomas and schwannomas. Hum Pathol 1998; 29:476–81. Kim N-G, Kim JJ, Ahn J-Y, Seong C-M, Noh SH, Kim CB, Min JS, Kim H. Putative chromosomal deletions on 9p, 9q and 22q occur preferentially in malignant gastrointestinal stromal tumors. Int J Cancer 2000;85:633–8. O’Leary T, Ernst S, Przygodzki R, Emory T, Sobin L. Loss of heterozygosity at 1p36 predicts poor prognosis in gastrointestinal/ smooth muscle tumors. Lab Invest 1999;79:1461–7. El-Rifai W, Sarlomo-Rikala M, Andersson LC, Knuutila S, Miettinen M. DNA sequence copy number changes in gastrointestinal stromal tumors: tumor progression and prognostic significance. Cancer Res 2000;60:3899–903. Lasota J, Wozniak A, Sarlomo-Rikala M, Rys J, Kordek R, Nassar A, Sobin LH, Miettinen M. Mutations in exon 9 and 13 of KIT gene are rare events in gastrointestinal stromal tumors. A study of two hundred cases. Am J Pathol 2000;157:1091–5. Lux ML, Rubin BP, Biase TL, Chen C-J, Maclure T, Demetri G, Xiao S, Singer S, Fletcher CDM, Fletcher JA. KIT Extracellular and kinase domain mutations in gastrointestinal stromal tumors. Am J Pathol 2000;156:791–5. Heim S, Mitelman F. Cancer cytogenetics. New York: Wiley-Liss, 1995. Dei Tos AP, Maestro R, Doglioni C, Piccinin S, Libera DD, Boiocchi M, Fletcher CD. Tumor suppressor genes and related molecules in leiomyosarcoma. Am J Pathol 1996;148:1037–45. Otano-Joss M, Mechtersheimer G, Ohl S, Lehnert T, Willeke F, Moller P, Otto HF, Lichter P, Joss S. Analysis of chromosome copy number changes in leiomyosarcoma through molecular cytogenetic methods. Verh Dtsch Ges Pathol 1998;82:207–9.