CGH Analysis Shows Genetic Similarities and Differences in Atypical Fibroxanthoma and Undifferentiated High Grade Pleomorphic Sarcoma
Total Page:16
File Type:pdf, Size:1020Kb
ANTICANCER RESEARCH 24: 19-26 (2004) CGH Analysis Shows Genetic Similarities and Differences in Atypical Fibroxanthoma and Undifferentiated High Grade Pleomorphic Sarcoma DANIELA MIHIC-PROBST1*, JIANMING ZHAO1*, PARVIN SAREMASLANI1, ANGELA BAER1, CHRISTIAN OEHLSCHLEGEL2, BRUNO PAREDES3, PAUL KOMMINOTH4 and PHILIPP U. HEITZ1 1Department of Pathology, University Hospital, Zürich; 2Institute of Pathology, Cantonal Hospital, St. Gallen; 3Department of Dermatology, University Hospital Bern; 4Institute of Pathology, Cantonal Hospital, Baden, Switzerland Abstract. Background: Atypical fibroxanthoma (AFX) and hyperchromasia and pleomorphism. Mitotic figures, including undifferentiated high grade pleomorphic sarcoma (UpS) are abnormal forms, are common. In contrast to UpS, AFX is histologically very similar, if not identical. However, they differ located in the dermis, lacks significant subcutaneous infiltration significantly in clinical outcome. Materials and Methods: We and has an excellent prognosis after complete surgical excision. used comparative genomic hybridization (CGH) to screen 24 It should be noted, however, that in exceptionally rare cases a AFX and 12 UpS for genomic alterations. Results: DNA copy lesion qualified as AFX can behave in an extremely aggressive number changes were observed in 20/24 AFX and in all UpS. manner, comparable with that of UpS (1, 2). AFX is a solitary The most frequent alterations occurring with comparable tumor usually occurring on sun-damaged skin of the head and frequency in both tumors were deletions on chromosomes 9p neck of elderly persons. Helwig first described 20 tumors in and 13q. We also detected statistically significant differences of 1963 (3, 4). Solar UV- induced p53 mutations may play an genetic alterations between the two tumors concerning important role in the pathogenesis of AFX (5, 6). deletions on 1q, 3p, 5q, 11p, 11q, gains on 7q, 12q and high The terms UpS and pleomorphic malignant fibrous level gains on 5p and 11q. Conclusion: Despite their very histiocytoma are used synonymously in the literature (7). similar histology, AFX and UpS show clear differences in their However the validity of the latter has been questioned for two genetic alterations. This might contribute to the different main reasons: 1) the introduction of immunohistochemistry biological behavior of the two tumors. On the other hand the has shown a definable line of differentation in most high grade similarities in genetic alterations on chromosomes 9p and 13q pleomorphic sarcomas (8); 2) undifferentiated high grade might suggest a common pathogenetic pathway. pleomorphic sarcomas show no evidence of histiocytic differentiation. The histology of atypical fibroxanthoma (AFX) and Several studies using various approaches, e.g. ploidy undifferentiated high grade pleomorphic sarcoma (UpS) is analysis by flow cytometry (9) and image analysis (10), failed virtually identical: it is characterized by a mixture of large to elucidate the pathogenetic pathway of AFX and UpS (11- atypical giant and spindle cells with considerable nuclear 13). Further attempts to differentiate AFX from UpS on the basis of proliferation (14), apoptosis and regulators thereof, i.e. p53 and bcl2 (15), as well as immunohistologic studies were not successful (16). This has led to the hypothesis that *Contributed equally to this study AFX may represent a superficial form of UpS (17). However, a recent report showed a different expression Correspondence to: Daniela Mihic-Probst, M.D., Department of of the LN2 antigen (CD74) in the two lesions: while ninety Pathology, University Hospital, Schmelzbergstrasse 12, 8091 percent of UpS stained strongly for LN2, this reaction was Zürich, Switzerland. Tel: +41-1-255-2595, Fax: +41-1-255-4440, e- absent in the vast majority of AFX (18). mail: [email protected] To date, a small number of chromosomal analyses of UpS Key Words: Atypical fibroxanthoma, undifferentiated high grade have been carried out (19-22) but, to our knowledge, only pleomorphic sarcoma, pleomorphic malignant fibrous one detailed chromosomal analysis including AFX, focusing histiocytoma, comparative genomic hybridization. on ras-mutations, has been reported (23). 0250-7005/2004 $2.00+.40 19 ANTICANCER RESEARCH 24: 19-26 (2004) In the present study, we used comparative genomic CGH analysis. CGH allows a rapid detection of DNA copy number hybridization (CGH) to screen for genetic alterations in two changes in the entire genome, providing an overview of genomic groups of tumors; i.e. 24 atypical fibroxanthomas (AFX) and imbalances in a given tumor. CGH was performed essentially as previously described (25). 12 undifferentiated high grade pleomorphic sarcomas One microgram of tumor DNA was labeled with Spectrum (UpS). Phenotypic differentiation of UpS was excluded. Green-dUTPs (Vysis, Downers Grove, IL, USA) by nick The goals of this study were: 1) to screen for frequently translation (BioNick kit, Life Technologies, Basel, Switzerland). altered chromosomal regions in AFX and UpS, 2) to Spectrum Red-labeled normal reference DNA (Vysis) was used elucidate molecular similarities and/or differences of the for co-hybridization. The hybridization mixture consisted of 200 two tumors, and 3) to provide a molecular basis for the ng Spectrum Green-labeled tumor DNA, 200 ng of Spectrum similarity in morphology of the two tumors on the one hand Red-labeled normal reference DNA and 10-20 Ìg of human Cot- 1 DNA (Life Technologies) dissolved in 10 Ìl of hybridization and of their different biological behavior on the other. buffer (50% formamide, 2X SSC, pH 7.0). Hybridization took place over 3 days at 37ÆC to sex-matched normal human Materials and Methods metaphase spreads (Vysis). Slides were washed at 45ÆC three times for 10 min in wash solution (50% formamide, 2X SSC, pH Patients and tumor specimens. Specimens of thirty-six patients (28 7.0) followed by twice for 5 min in 2X SSC. The chromosomes men and 8 women, age 39 to 97 years) were selected for this study. were counterstained with 4,6-diamidino-2-phenylindole (DAPI) Twenty-four patients suffered from AFX, twelve from UpS for identification. (Clinical data of the patients: Table I). None of the patients had Digital images were collected from six to seven metaphases received chemotherapy or radiotherapy prior to surgery. Histology using a cooled charge-coupled device (CCD) camera of all tumors was reviewed by one pathologist (D.M.). Histological (Microimager 1400; Xillix Technologies, Vancouver, Canada) evaluation and classification of the tumors were based on attached to a Zeiss Axioskop microscope and a computer. The established criteria (17, 7). software program QUIPS (Vysis) was used to calculate average All AFX were located in the dermis without infiltration of green-to-red ratio profiles for each chromosome. At least six subcutaneous tissue or necrosis. Only tumors with abundant observations per autosome and three observations per sex atypical pleomorphic cells and atypical mitoses were observed. chromosome were included in each analysis. Chromosomal gains Immunohistochemically they sometimes showed focal positivity for and losses were defined as fluorescence ratio values above 1.2 antitrypsin, antichymotrypsin and CD 68. Other types of neoplasms or below 0.8, respectively. Amplification was assumed at were excluded by lack of immunostaining for cytokeratins, chromosomal regions where the green to red ratio exceeded 1.5. epithelial membrane antigen, S-100 protein, desmin, ·-smooth Since some false-positive results were found in normal tissues at muscle actin and CD 34 (Fig. 1A, B, C). 1p32-pter, 16p, 19 and 22, gains at these G-C-rich regions were Seven UpS were localized in the subcutis while five were found excluded from all analyses. within skeletal muscle. Two of the subcutaneous tumors showed involvement of the adjacent dermis (Table I ). The study comprised Statistics. For scoring of genetic alterations, whole chromosome exclusively UpS not showing any definable line of differentiation. changes were scored as one event. All other changes were scored The tumors were extensively sampled and phenotyping resulted in by chromosomal arm. A loss and a gain on one arm were scored as a lack of immunoreactivity for desmin, ·-smooth muscle actin, two changes, whereas two separate losses (or gains) on the same myogenin, cytokeratins, CD34 and S-100 protein in all tumors arm were scored as one change. (Figure 1D, E, F). Contingency table analysis was used to compare frequencies of genomic alterations in different groups of tumors. All statistical DNA extraction. All examined tumor specimens were formalin-fixed tests were two-sided and were considered to be statistically and paraffin-embedded. Representative tumor samples were significant at p<0.05. microdissected and scraped from five to ten 10-Ìm-thick unstained tissue sections under a stereomicroscope (Stemi DV4, Zeiss, Results Germany) at 10 to 40x magnification. Genomic DNA was isolated DNA copy number changes were observed in 20/24 AFX as previously described (24). Briefly, representative tumor material and in all 12 UpS (Table I ). UpS showed a statistically was collected in a 1.5 ml Eppendorf tube containing 100% ethanol and centrifuged to the bottom of the tube (5 min at 14000 rpm). significant (p value < 0.0009) larger number of genomic After the ethanol had been removed, the tissue material was alterations than AFX. The average number of alterations treated three times with 1 ml of fresh xylol for 15 min at 55ÆC in a per tumor was 3.3±0.7 in AFX vs. 8.9±1.1 in UpS (Table thermomixer. Subsequently, the specimens were treated with 100% II). High-level gains (amplifications) were observed in 9/12 ethanol and centrifuged for 10 min at 14000 rpm. After a second UpS. By contrast, only 1/24 AFX showed amplifications. treatment with 70% ethanol and subsequent centrifugation, the The chromosomal regions with DNA copy number tissue material was dried in a speed-vac for 10 min at room alterations (losses and gains) identified by CGH are shown temperature. The tissue was digested overnight with 0.5 mg/ml Proteinase K (Sigma, St. Louis, MO, USA) in 1ml digestion buffer in Figure 2, Figure 3 and Table III.