OUP user menu

Myofibroblastic Sarcoma vs Nodular Fasciitis
A Comparative Study of Chromosomal Imbalances

Guo-Zhao Meng MD, PhD, Hong-Ying Zhang MD, Zhang Zhang MD, Bing Wei MD, Hong Bu MD, PhD
DOI: http://dx.doi.org/10.1309/AJCPV6H2WSYXLKFB 701-709 First published online: 1 May 2009


We investigated the molecular cytogenetic features in myofibroblastic sarcoma (MS) to gain insight into the nature of the controversial entity. DNA copy number changes were analyzed by comparative genomic hybridization in 29 cases of MS and 5 cases of nodular fasciitis. The characteristic chromosomal imbalances in MS were gains at 1p11 → p36.3 (19/29 [66%]), 12p12.2 → p13.2 (13/29 [45%]), 5p13.2 → p15.3 (9/29 [31%]), and chromosome 22 (8/29 [28%]) and loss at 15q25 → q26.2 (7/29 [24%]). In contrast, only 1 of 5 cases of nodular fasciitis showed genetic aberrations. The average number of aberrations in nodular fasciitis (0.4) was significantly lower than that in MS (5.4). Thus, MS displayed complex DNA copy number changes and shared no range of common chromosomal abnormality with nodular fasciitis, indicating that distinct genetic pathways may be involved in the development of these entities.

Key Words:
  • Myofibroblastic sarcoma
  • Nodular fasciitis
  • Comparative genomic hybridization
  • CGH

Defining neoplastic myofibroblastic sarcoma (MS) as a distinct entity was controversial. With increasing case reports,16 it became clear that MS was a distinct entity in soft tissue sarcomas. Even though only low-grade MS was classified as a distinct entity in the newly published World Health Organization classification of soft tissue tumors,7 intermediate- and high-grade MS cases were documented in the literature.5,6,8 Some cases of MS were easy to misdiagnose as reactive lesions because of the common existence of myofibroblasts in reactive granuloma.9 In addition, the frequent exhibition of bland cytologic features in MS is an important factor in persuading pathologists toward a benign diagnosis.5,6 MS, especially low-grade MS, is easily confused with myofibroblasts composing nodular fasciitis because of their morphologic similarity and the overlapping immunophenotype.

Generally speaking, nearly all human malignant tumors are characterized by genomic changes. Comparative genomic hybridization (CGH), a technique established in 1992,10 enables the analysis of each chromosome in tumor cells for genetic amplification and deletion. CGH analysis could reveal recurrent genetic changes in tumors affecting chromosome sites that harbor genes participating in tumorigenesis. Since the establishment of the CGH method, numerous patterns of chromosomal aberrations, including gains and losses of tumor DNA sequences, were detected in various human malignant neoplasms. The molecular cytogenetic data about MS are presently scarce. To date, documented literature reveals that only 1 case of MS was detected by CGH,11 and just 3 cases were analyzed by karyotyping.12 The cytogenetic data pertaining to nodular fasciitis are also limited.1315 Some karyotype analyses demonstrated that nodular fasciitis was a clonal lesion.16 Whether the cytogenetic profile in MS overlaps with that in nodular fasciitis remains unknown.

In the present study, we analyzed the DNA copy number imbalances in MS using CGH to screen the entire genome for copy number aberrations (gains and/or losses) to investigate the molecular cytogenetic features of the controversial entity. At the same time, considering that it is difficult to distinguish low-grade MS from nodular fasciitis in the routine work of pathologists, we included a cohort of nodular fasciitis cases in the study for comparison. By comparing the DNA imbalance profile in MS with that in nodular fasciitis, we intended to provide some useful evidence for the differential diagnosis at the molecular cytogenetic level between these lesions.

Materials and Methods


The 29 cases of MS for CGH detection were retrieved from the consultation files and surgical pathology archives of West China Hospital, Sichuan University, Chengdu, China, between 1998 and 2005. The histologic and immunohistochemical features of 15 cases had been previously studied and published.5,6 All sections of the cases included in the study were confirmed independently by at least 2 pathologists (G.M., H.Z., and H.B.) and fulfilled the diagnostic criteria for MS described by Mentzel et al4 and the World Health Organization classification of tumors of soft tissue.7

Histologically, MS showed slender spindle cells arranged in fascicles, vaguely storiform, or tissue culture–like growth patterns. The tumors frequently had various degrees of cellularity accompanied by abundant fibrous collagen, myxoid matrix, or infiltration of inflammatory cells in the stroma. The tumor cells were characterized by eosinophilic cytoplasm with indistinct cell boundaries and fusiform, tapering, and wavy nuclei with small, centrally located, eosinophilic nucleoli. Low-grade MS exhibited mild nuclear atypia and rare mitotic figures Image 1A, Image 1B, and Image 1C. Intermediate-grade MS Image 1D and Image 1E showed moderate nuclear atypia and frequent mitotic figures, and high-grade MS Image 1F was a pleomorphic sarcoma showing prominent nuclear atypia and brisk mitotic figures.

Immunohistochemically, the cases included in the study were positive for smooth muscle actin Image 2A, calponin Image 2B, and fibronectin; rarely positive for desmin; and negative for high-molecular-weight caldesmon, laminin, and type IV collagen.5,6 In addition, all 29 cases included in the study were examined by electron microscopy and demonstrated the ultrastructural features of myofibroblasts Image 3. The clinicopathologic features of 29 patients are summarized in Table 1. The 5 cases of nodular fasciitis were selected from the surgical pathology archives of West China Hospital, Sichuan University in 2005. The clinicopathologic data for 5 cases of nodular fasciitis are summarized in Table 2. None of the patients whose data were included in the study received therapy before operation.

DNA Isolation and Labeling

The tissue specimens were fixed routinely in neutral buffered formalin and embedded in paraffin wax. The paraffin-embedded blocks of tumors were cut in 10-μm sections. Depending on the size of the lesional tissue, 5 to 30 unstained sections were collected in Eppendorf tubes and deparaffinized by washing with xylene and ethanol. Tumor genomic DNA was isolated by using a standard phenol-chloroform extraction procedure after Proteinase K digestion. Reference DNA was obtained from blood samples of healthy male and female donors.

Image 1

Histologic appearance of myofibroblastic sarcoma. A (Case 18), Spindle tumor cells with tapering and elongated nuclei arranged in fascicles, showing mild nuclear atypia (H&E, x400). B (Case 29), Spindle tumor cells with abundant eosinophilic cytoplasm arranged in vaguely fascicle pattern and admixed with sparse lymphocytes and plasma cells (H&E, x200). C (Case 23), Tumor cells with ill-defined eosinophilic cytoplasm and wavy and tapering nuclei infiltrated in the bone trabeculae with abundant fibrous collagen in the stroma (H&E, x200). D (Case 17), Feathery or stellate tumor cells with abundant eosinophilic cytoplasm haphazardly scattered in a myxoid matrix showing a tissue culture–like growth pattern (H&E, x400). E (Case 26), Spindle tumor cells with moderate nuclear atypia arranged in a vaguely storiform pattern (H&E, x200). F (Case 13), High-grade myofibroblastic sarcoma showing pleomorphic tumor cells with abundant amphophilic cytoplasm and prominent atypia in the nuclei (H&E, x400).

A 1-μg sample of each tumor and reference DNA were labeled by nick translation using a kit according to the manufacturer’s instructions (Vysis, Wiesbaden-Delkenheim, Germany) and the literature.17 Tumor DNA was labeled with SpectrumGreen–conjugated deoxyuridine triphosphate and reference DNA with SpectrumRed–conjugated deoxyuridine (Vysis). The DNase concentration in the labeling reaction was adjusted to obtain an average fragment size of 300 to 800 base pairs.


Target metaphase spreads needed for CGH were prepared following standard procedures. For hybridization, 800 ng of each labeled tumor and reference DNA, along with l0 μg of human Cot-1 DNA were precipitated with ethanol, air dried, and dissolved in 10 μL of hybridization buffer (2% dextran sulfate, 2x saline sodium citrate [SSC], and 50% deionized formamide). This mixture was denatured at 73°C for 5 minutes. Normal metaphase spreads were denatured in denature solution (70% formamide/2x SSC, pH 7.5) at 73°C for 5 minutes, dehydrated in a series of ice-cold ethanol (70%, 85%, and 100%) for 2 minutes each, and dried in the dark at 60°C. The denatured CGH probe was pipetted directly onto the slides, enclosed with a coverslip, and sealed with rubber cement. Hybridization was performed in a humidified dark chamber at 37°C for 3 to 4 days.

Image 2

Immunohistochemically, myofibroblastic sarcomas showed positive staining for smooth muscle actin (A, case 18) and calponin (B, case 13) (A and B, x400).

Image 3

Ultrastructural features of myofibroblastic sarcoma. The spindle tumor cells showed abundant rough endoplasmic reticulum (RER) in the cytoplasm and fibronexus junctions (arrow) at the cell surface (A, case 18, x12,000; B, case 13, x15,000)

Slides were washed in 0.4x SSC/0.3% NP-40 (Roche, Kulmbach, Germany) at 74°C for 2 minutes followed by 2x SSC/0.1% NP-40 at room temperature for 2 minutes. Slides were allowed to air dry in the dark for 30 minutes and then counterstained with 10 μL of 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI; 40 μg/mL) containing an antifading agent.

Image Acquisition and Analysis

Digital gray level images of DAPI, fluorescein isothiocyanate, and tetramethylrhodamine isothiocyanate fluorescence were captured using a CCD camera integrated with an Olympus fluorescence microscope (Olympus, Tokyo, Japan). Karyotyping and image analysis were carried out with the ISIS digital image analysis system (MetaSystems, Altlussheim, Germany). For defining the chromosomal region of DNA sequence losses and gains, the 50% thresholds (upper, 1.25; lower, 0.75) were used. Chromosomal regions were interpreted as overrepresented when the green/ red ratio was higher than 1.25 and underrepresented when the green/red ratio was lower than 0.75. These limits were based on control experiments with 2 differentially labeled normal DNAs. Gains exceeding the limit of 2.0 were regarded as high-level amplifications.18 A positive control sample (a tumor with known DNA sequence copy number aberrations) and negative control samples (2 differentially labeled normal DNAs) were included in each hybridization. For each tumor, an average of 10 metaphases were analyzed for each chromosome. Because of false-positive results owing to G-C–rich regions, chromosomes X and Y were excluded from the analysis.19 Telomeric and pericentromeric regions were also excluded from evaluation. Gains and losses were considered recurrent if they were detected in 2 or more cases.

View this table:
Table 1

Statistical Analysis

The χ2 test was used to compare the frequency of genetic aberrations with clinicopathologic parameters. P values less than .05 were considered statistically significant.


DNA copy number changes were detected in 22 (76%) of 29 cases of MS studied by CGH. The mean number of aberrations was 5.4 per sample (range, 0–16). Of the tumors, 9 showed only gains or only losses. Gains were more frequent than losses (gain/loss = 3:2). High-level amplification was not detected. The minimal common regions for the most frequent gains were as follows: 1p11 →p36.3 (19/29 [66%]), 12p12.2 → p13.2 (13/29 [45%]), 5p13.2 → p15.3 (9/29 [31%]), 22 (8/29 [28%]), 2p11.2 → p25.1 (6/29 [21%]), and 2q34 → q37.3, 3q24.3 → q25, and 5q31.3 →q35.3 (5/29 [17%] each). The minimal common regions for the most frequent losses were as follows: 15q25 → q26.2 (7/29 [24%]) and 3p14.3 → p25, 6q27, 17p13, and 18p11.3 (4/29 [14%] each). The CGH data for 29 cases of MS are given in Image 4 and Table 3.

Correlation of the total number of genetic aberrations and the most frequent aberrations in 29 cases of MS with clinicopathologic parameters revealed that the number of aberrations per case was associated with higher tumor grade (P < .0001). The gain of 1p was associated with higher tumor grade and larger tumor size (P < .05). The gain of 12p was associated with higher tumor grade (P < .0001). When comparing the number of aberrations and the specific patterns of aberrations with the established prognostic parameters, no statistically significant difference was observed Table 4.

View this table:
Table 2

In 5 cases of nodular fasciitis examined by CGH, only 1 case showed 2 loci of DNA copy number changes (gains at 10p14 → p15 and 20q12 → q13.3). The other 4 cases showed no genetic alteration Table 5. When compared with the CGH data for MS, the average number of aberrations per case in nodular fasciitis (0.4) was significantly lower than that in MS (5.4; P < .0001), and no identical or overlapping recurrent gain or loss was identified between MS and nodular fasciitis.

Image 4

Comparative genomic hybridization ideogram summarizing gains and losses of DNA sequences in 29 cases of myofibroblastic sarcoma. The vertical red bars on the left indicate losses, and the green bars on the right indicate gains of genetic material in the tumor genomic DNA. Each bar illustrates the region of the chromosome affected in 1 tumor.


MS has been a controversial neoplasm for a long time because myofibroblasts may exist in various reactive conditions, including reparative granulation, granulomas, and inflammatory pseudotumor, and because various benign and malignant soft tissue tumors may contain myofibroblastic cells in varying degrees. Even though numerous studies concerning the clinicopathologic features of MS have been published,18 the molecular cytogenetic data for the disease are limited.11,12 We therefore aimed to investigate the chromosomal imbalances in MS using the CGH method to gain insight into the nature of the controversial entity.

The present study revealed that 76% of MS cases exhibited multiple and complex DNA copy number changes. The mean number of aberrations per sample was 5.4, and gains were more frequent than losses. It was noticed that higher grade tumors exhibited a larger number of aberrations, and the 7 cytogenetically normal cases in the study were all low-grade MS, indicating that the number of aberrations was associated with the progress of MS. Even though some cases in the present study showed no genetic aberrations, such as cases 5, 12, 16, 18, 21, 25, and 29, it has to be kept in mind that CGH cannot detect chromosomally balanced translocation or mutation, which may occur in some low-grade MS cases. Furthermore, some gene aberrations occurring in some tumors are too minimal to be detected by CGH. On the other hand, CGH had implications for the clonal nature of the sample detected because normal tissue contamination will mask the changes in DNA copy number and many cells may not acquire the same alteration independently.20 The molecular cytogenetic data presented herein show that MS exhibited recurrent genetic alterations, supporting the concept that at least a proportion of the cases may be clonal neoplasms.

When the CGH data from our series of MS were compared with data in the literature, good concordance was generally seen. MS was first studied cytogenetically by Fletcher et al,12 who detected 3 cases of MS by using a karyotype analysis method and found complex karyotypes and characteristic chromosomal rearrangements of 1p, 9p, and 10p. The other study was reported by Morawietz et al,11 who detected a case of intermediate-grade MS by using CGH and found chromosomal gains at 1p34.1 → p36.2, 11q12, 16p12 → p13, 17p, 12p1 → p3, 17q, 12q2 → q5, 19p+q, 20q12 → q13, and 22q12 → q13 and loss at 13q13 → q34. Thus, gains of 1p, 12p, and 22q were the common aberrations for MS detected in our study and reported in the literature. Apart from these previously recognized DNA copy number changes, we found numerous novel recurrent chromosomal imbalances in MS, in particular frequent gains of 5p13.2 → p15.3 (31%), 2p11.2 →p25.1 (21%), 2q34 → q37.3 (17%), 3q24.3 → q25 (17%), and 5q31.3 → q35.3 (17%) and loss of 15q (24%).

The present study showed that the gain of 1p was the most frequent aberration in MS and was associated with higher tumor grade and larger tumors. So, DNA gain of 1p may be important for MS development, because it was seen preferentially in larger and higher grade tumors. Besides in MS, gain of 1p can also be detected in other soft tissue sarcomas such as malignant fibrous histiocytoma (MFH),18 leiomyosarcoma,21 liposarcoma,22 and rhabdomyosarcoma.23 The fact that MS showed frequent gains of 1p indicated that potential oncogenes in the region may be involved in tumorigenesis. In the chromosomal region, the candidate oncogenes include AF1P, PAX7, and BLYM and transcription factors such as MYCL1, TAL1, E2F2, and JUN.19,24

The present study revealed that gain of 12p occurred in 45% of MS cases and was associated with higher tumor grade. The gain of 12p was also detected in MFH25 and germinoma.26 The chromosomal region was known to include the ras oncogene, the mutation of which was associated with malignant transformation of the tumor.27 Other candidate oncogenes in the region included JAW1, SOX5, and CCND2.28,29 Gain of 5p was detected in 31% of MS cases in our study; it was also detected in MFH18 and leiomyosarcoma.30 In the chromosomal region, the known candidate oncogene was SKP2.31 Gain of chromosome 22 occurred in 28% of MS cases in the present study. The same genetic aberration was also detected in other soft tissue sarcomas, such as leiomyosarcoma21 and dermatofibrosarcoma protuberans.32 The specific oncogene located in the chromosomal band is not yet known. Because high frequent gains of 12p, 5p, and 22 were detected in MS, the aforementioned candidate oncogenes may be involved in the tumorigenesis of MS.

View this table:
Table 3
View this table:
Table 4
View this table:
Table 5

In the present study, losses of DNA copy numbers in MS were less numerous as compared with the gains of DNA copy numbers, with the loss of chromosomal material of 15q25 → q26.2 the most frequent aberration. The finding was in accordance with the results of Fletcher et al,12 who also detected frequent loss of chromosome 15q in MS. Loss of 15q was demonstrated to have high frequency in colon polyp adenoma33 and colon carcinoma.34 A putative tumor suppressor gene residing within the chromosomal band was TPM1.35

MFH, the most frequent diagnosis in soft tissue sarcomas in the past decades, is now considered to be an exclusive diagnosis. After excluding pleomorphic liposarcoma, pleomorphic leiomyosarcoma, and pleomorphic rhabdomyosarcoma, pleomorphic sarcoma that shows no distinct line of differentiation is designated as MFH, and any MFH showing myofibroblastic differentiation should allow a diagnosis of pleomorphic myofibrosarcoma.8,36 It is difficult to distinguish MFH from high-grade MS without electron microscopy. So far, numerous CGH studies on MFH have been reported in the literature.18,3739 When comparing the distribution and frequency of specific chromosomal imbalances between MS and MFH, significant differences could be observed. Instead of gains at 1p, 12p, and 5p as frequently detected in MS, MFH typically showed gains at chromosomes 1q21 → q22, 17q23 → qter, 17p, Xp, and 1q21.18,3739 Furthermore, losses of 15q and 3p, both relatively frequent in MS, were not detected in MFH, and losses of 9p21 → pter, 13q21 → 22q, 13q21, 13q22, 13q12 → q14, and 13q21, which were recurrent genetic aberrations in MFH,18,3739 were not detected or were less frequent in MS. In addition, some frequent genetic aberrations in MS, such as gains of 1p and 5p, have also been detected in MFH, with identical, overlapping, or slightly different minimal common regions.18,37,39 Thus, the comparison between CGH findings in MS and MFH supports the assumption that these 2 lesions are somewhat cytogenetically related but distinct tumor entities.

Nodular fasciitis, a lesion consisting of myofibroblasts and fibroblasts, is histologically and immunophenotypically similar to low-grade MS and is the most important disease in the differential diagnosis for MS. Nodular fasciitis is usually recognized as a reactive lesion. However, cytogenetic findings indicated that the lesion had clonal DNA aberrations, including 15q25 breakpoint,13 chromosome rearrangement of 3q21,15 and chromosome rearrangements of 15p11.2 and 16p13.3.16 Thus, nodular fasciitis is considered a benign tumor rather than a reactive lesion. In the present study, most cases of nodular fasciitis displayed no genetic aberrations, and the average number of aberrations in nodular fasciitis was significantly lower than in MS. By comparison, no recurrent overlap or identical genetic aberration was found between nodular fasciitis and MS, further supporting the concept that nodular fasciitis and MS are cytogenetically different individual disease entities, in accordance with the clinical observation that MS and nodular fasciitis develop de novo and show different clinical behavior.

The present study systematically investigated the molecular cytogenetic patterns of MS in a large series, demonstrating that MS showed characteristic cytogenetic aberrations and should be regarded as a clonal disease. In addition, MS showed different cytogenetic aberration profiles compared with the profiles of nodular fasciitis. This study demonstrates that despite the morphologic and phenotypic similarity, MS and nodular fasciitis should be regarded as different cytogenetic entities, and distinct genetic pathways may be involved in the development of these diseases.


We are grateful to Feng Li and Chun-Xia Liu, Department of Pathology, Shihezi University, Shihezi, Xinjiang, China, and Cai-Pu Chun, The Fourth Affiliated Hospital of Shihezi University, Aksu, Xinjiang, for help in CGH image analysis. We thank Pei-Dan Huo, Genetic Laboratory, Sichuan Reproductive Health Institute, Chengdu, for excellent technical assistance in preparing the metaphase spreads.


  • * Dr Meng is now a postdoctor in the Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
View Abstract