Background Angiomatoid fibrous histiocytoma (AFH) is a rare soft-tissue tumour of uncertain differentiation and low metastatic potential. Cytogenetics and/or molecular genetics have revealed that most have a rearrangement of the EWSR1 gene, whereas a FUS gene rearrangement is present in a minority of cases. Although some cases of AFH display striking pleomorphism and mitotic activity, there are no known clinical, morphological or genetic factors that predict metastasis. The authors present clinicopathological features of AFH, including cases showing a pleomorphic histological appearance, and results of fluorescence in situ hybridisation analysis of EWSR1 and FUS rearrangements.
Methods Tumour samples from 10 patients were subjected to clinicopathological and immunohistochemical analysis and dual-colour fluorescence in situ hybridisation for EWSR1 and FUS with split-signal probes.
Results All cases showed clinical features (sites: extremities followed by trunk; age: adolescent to young adult), morphology (multinodular proliferation of spindle cells, lymphoid cuffs and pseudovascular spaces) and immunohistochemical results (more than half were positive for CD68, CD99, desmin and epithelial membrane antigen) typical of AFH. There were two local recurrences in each of two patients. Two patients developed distant metastases and died from the disease; tumours of these two patients showed focal proliferation of large pleomorphic cells with hyperchromatic nuclei and high proliferative activity (>10/10 high-power field and Ki-67 labelling index >10%). There were no clinical, histological or immunohistochemical differences between the nine cases with EWSR1 rearrangement and one case with FUS rearrangement.
Conclusions Wide surgical excision and careful follow-up are necessary for patients with AFH in view of its risk of local recurrence and metastasis leading to a fatal outcome.
- KI 67
- soft tissue tumours
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Angiomatoid fibrous histiocytoma (AFH) is a rare soft-tissue tumour of uncertain differentiation, originally defined as one of the five subtypes of malignant fibrous histiocytoma (MFH) by Enzinger in 1979.1 The term ‘AFH’ has been used rather than ‘angiomatoid MFH’ because the prognosis is favourable.2 It occurs predominantly in the deep dermis and subcutis of extremities of adolescents and young adults.1 3 4 Clinically, patients with AFH can experience up to three systemic symptoms: fever, anaemia and weight loss. Microscopically, AFH is characterised by pseudovascular, blood-filled spaces surrounded by a multinodular proliferation of spindle and/or round cells and lymphoid cuffs. Some cases of AFH display striking pleomorphism and mitotic activity.1 3 5 Immunohistochemically, the tumours are often positive for vimentin, desmin, epithelial membrane antigen (EMA), CD68 and CD99.4 6–8
Genetic studies of fusion genes in AFH have recently been published. To our knowledge, 30 cases of AFH have been reported, 21 containing a EWSR1–CREB1 fusion gene resulting from t(2;22)(q33;q12),9 10 seven containing an EWSR1–ATF1 fusion gene resulting from t(12;22)(q13;q12),9–14 and two containing a FUS–ATF1 fusion gene from t(12;16)(q13;p11).15 16 Some of these fusion genes, such as EWSR1–CREB1 and EWSR1–ATF1, have also been detected in clear-cell sarcoma of soft tissue and gastrointestinal tract,17–19 which is an aggressive soft-tissue sarcoma with melanocytic differentiation, morphologically and clinically distinct from AFH, which shows no evidence of melanocytic differentiation.
Reverse transcription (RT)-PCR can be readily used to detect fusion genes; however, fresh–frozen tumour tissues may not always be available for RT-PCR analysis. Although translocations can be detected in formalin-fixed, paraffin-embedded tissues by RT-PCR, we have used fluorescence in situ hybridisation (FISH) with commercially available DNA probes as an ancillary method for use in routine clinical practice to establish the diagnosis of chromosome translocation-associated sarcomas such as Ewing sarcoma/primitive neuroectodermal tumour, desmoplastic small round-cell tumour, clear-cell sarcoma, alveolar rhabdomyosarcoma and extraskeletal myxoid chondrosarcoma.20–22 As described above, 93% of AFHs have a rearrangement of EWSR1, as manifested by the EWSR1–CREB1 and EWSR1–ATF1 fusion genes, whereas 7% of cases have a rearrangement of FUS harbouring the FUS–ATF1 fusion. The latter fusion gene has been detected by RT-PCR in a large, deep-seated AFH.16 Although AFH with pleomorphic features has been described, with diagnosis confirmed by FISH for identification of EWSR1 rearrangement,5 there are no known clinical, morphological or genetic factors that predict metastasis. Here, we present clinicopathological features of 10 cases of AFH, including cases showing a pleomorphic histological appearance, analysed for EWSR1 and FUS rearrangements by FISH.
Materials and methods
Archival pathology files from the Department of Surgical Pathology, Sapporo Medical University Hospital, Sapporo, Japan and the Laboratory of Pathology, National Cancer Center Hospital, Tokyo, Japan, and physician consultation files were screened for patients with a diagnosis of AFH. Ten patients with lesions morphologically and immunophenotypically resembling those of AFH after pathological review were selected for further study. The study protocol for collection of tumour samples and clinical information was approved by the ethics committee of Sapporo Medical University School of Medicine.
For light microscopy, all specimens were routinely fixed in 10% formalin and embedded in paraffin, and 4 μm-thick sections were cut and stained with H&E. We studied morphological variant features (round or ovoid cells, nuclear pleomorphism, rhabdoid features, myxoid change, hyalinisation and haemosiderin deposits) as well as typical features (multinodular proliferation, lymphoid cuffs, spindle histiocytic cells and pseudovascular pattern).
Tumour slides of all 10 cases, stained with H&E, were reviewed, and representative sections from the paraffin blocks were examined by the labelled polymer method with the appropriate use of positive and negative controls. Sections were dewaxed, rehydrated and moistened with phosphate-buffered saline (PBS) (pH 7.4) and then pretreated in an autoclave at 121°C for 5 min in 10 mmol/l citrate buffer (pH 6.0), followed by a 30 min incubation with antibodies to the following antigens in an automated immunostaining system (Dako Autostainer; Dako, Carpinteria, California, USA): vimentin (clone V9; prediluted; BioGenex, San Ramon, California, USA), CD68 (clone KP1, 1:50 dilution; Dako, Glostrup, Denmark), CD99 (clone 12E7; 1:50 dilution; Dako), desmin (clone D33; prediluted; BioGenex), α-smooth muscle actin (clone 1A4; 1:50 dilution; Dako), muscle-specific actin (clone HHF35; prediluted; BioGenex), CD34 (clone My10; 1:50 dilution; Becton Dickinson, San Jose, California, USA), cytokeratin (clone AE1/AE3; prediluted; BioGenex), EMA (clone E29; 1:50 dilution; Dako), S-100 protein (clone S100; prediluted; BioGenex) and Ki-67 (clone MIB-1; 1:50 dilution; Dako). The strength of correlation between the Ki-67 labelling index (LI) and the mitotic index was statistically judged by Pearson's correlation coefficient (r).
We performed FISH after identifying the most histologically typical areas (those with less inflammation and fibrosis) on H&E-stained sections. Analysis was carried out on 4 μm-thick, formalin-fixed, paraffin-embedded sections, as described previously.21 22 In brief, after deparaffinisation and dehydration, the sections were immersed in 0.2 M HCl for 20 min and then immersed in pretreatment solution (Abbott Molecular, Abbott Park, Illinois, USA) for 30 min at 80°C. After digestion in protease for 60 min at 37°C, the sections were washed in PBS for 5 min at room temperature, fixed in 10% formaldehyde for 10 min at room temperature, washed in PBS for 5 min at room temperature, and placed in prewarmed denaturation solution (Abbott Molecular) for 5 min at 72°C. Sections were then dehydrated in an ethanol series (70%, 85% and 100%) at room temperature for 1 min each and air-dried.
Commercial EWSR1 (22q12) and FUS (16p11) dual-colour, split-signal probes (Abbott Molecular) were prepared for dual-colour FISH. A 10 μl volume of solution containing each FISH probe was placed on the target areas of each section at 45°C, and the slides were coverslipped, sealed with rubber cement, and incubated for 48 h at 37°C in a humidified chamber. Post-hybridisation washing was performed for 2 min with wash buffer (Abbott Molecular) at 72°C. Nuclei were counterstained with 10 μl 4′,6-diamidino-2-phenylindole (DAPI; Abbott Molecular), and the slides were coverslipped.
Stained sections were observed under a fluorescence microscope equipped with appropriate filters, and the resultant images were captured with a charge-coupled device camera. A total of 50–60 nuclei exhibiting both green and orange signals were counted by two different investigators. A split signal was considered positive for EWSR1 or FUS rearrangement if the distance between the flanking signals was 2–3 times the estimated signal diameter.
A summary of the clinical findings for the 10 patients with AFH is shown in table 1. Patients included five men and five women, with an age range of 4–58 years (mean 25.8; median, 21.5). Nine tumours were located in the subcutis, and one was located in the skeletal muscle. Six were in the extremities (forearm, upper arm, thigh and knee), and four in the trunk (inguinal area, back and hip). All tumours were cystic and filled with haemorrhagic fluid or blood; they ranged in size from 1.7 cm to 23.0 cm (mean 6.4; median 4.3). The follow-up period ranged from 3 years to 25 years 3 months (mean 7 years 10 months; median 3 years). Marginal excision was performed in seven patients, and wide excision in three; two patients who had wide excision also had chemotherapy or chemotherapy combined with radiotherapy. Seven patients had neither local recurrence nor metastasis at the latest follow-up. Two local recurrences occurred in each of two patients (patients 2 and 4); however, one (patient 4) was alive without any evidence of the disease 17 years after surgery. Two patients (patients 2 and 5) developed lung metastases and died from the disease 17 years 6 months and 1 year 3 months after surgery, respectively.
Histological features of all 10 cases are shown in table 2. Typical features (multinodular proliferation, lymphoid cuffs, and pseudovascular, blood-filled spaces) were seen in all cases (figure 1A). Spindle cells with bland, vesicular nuclei, which were arranged in short fascicles and an occasional storiform pattern, were present in all cases (figure 1B). In seven cases, the tumours displayed a sheet of round or ovoid cells with pale eosinophilic cytoplasm. Two of these (patients 2 and 5) contained focal areas of larger pleomorphic cells with hyperchromatic nuclei (figure 2). Stromal myxoid change was seen in four cases. Mitotic figures ranged from 1 to 30/10 high-power fields (HPF) (mean, 8.7/10 HPF; median, 1.5/10 HPF). There was no necrosis in these cases.
Immunohistochemical results for the 10 cases of AFH are summarised in table 3. All of the tumours were positive for vimentin; reactivity was present in more than 50% of the tumour cells. Immunoreactivity for CD68, CD99, desmin, α-smooth muscle actin and EMA was found in eight (80%), seven (70%), six (60%), four (40%) and six (60%) of the tumours, respectively. The tumours were negative for muscle-specific actin, CD34, cytokeratin and S-100 protein immunoreactivity.
Tumours were positive for Ki-67 in all cases, and the Ki-67 LI ranged from 2% to 50% (mean 14.2%; median 10%). Six cases showed a high Ki-67 LI of 10% or more, whereas four showed an LI less than 10%. The Ki-67 LI correlated strongly with the mitotic index (r=0.86).
The FISH findings are summarised in table 1. Hybridisation with EWSR1 and FUS probes produced positive results in all 10 cases. A split-signal pattern of green and orange demonstrating rearrangement of the FUS gene was detected in an average of 50% of cells in one tumour (patient 1) (figure 3A). Rearrangement of the FUS gene was not detected in nine tumours (patients 2–10). In contrast, a split-signal pattern demonstrating rearrangement of the EWSR1 gene was detected in 50–90% of cells in nine tumours (patients 2–10) (figure 3B) with one case (patient 1) showing a fusion-signal pattern.
The cases in the present series showed the clinical features (sites: extremities followed by trunk, age: adolescent to young adult) and morphology (multinodular proliferation of spindle cells, lymphoid cuffs and pseudovascular spaces) of AFH.1 3 4 In addition, more than half were positive for CD68, CD99, desmin and EMA. Four cases exhibited immunoreactivity for α-smooth muscle actin, which is described as a ‘myoid’ marker.6 These immunohistochemical findings correspond to those of previous reports.4 6 8
Most AFHs follow a less aggressive clinical course, with a regional recurrence rate of 12% and a metastatic rate of ∼1%.3 Rare deaths from late distant metastases have been reported.1 23 Therefore AFHs are classed as tumours of intermediate biological behaviour by WHO.2 Morphologically, AFHs may show pleomorphism and mitotic activity.1 3 5 However, these features do not correlate with clinical behaviour3 and have not proved to be of prognostic significance, owing to a short follow-up period.5 In the present study, two patients (20%) developed distant metastases and died from the disease. Early metastasis, involving a 23 cm tumour in the skeletal muscle, occurred in one of these patients. In a larger series, deep intramuscular location was reported to constitute a risk factor for metastasis.3 In addition, these two tumours showed focal proliferation of large pleomorphic cells with hyperchromatic nuclei and high proliferative activity (>10/10 HPF and Ki-67 LI >10%). As judged from patients with conventional AFH who develop metastasis,1 23 all primary AFH tumours require wide surgical excision and careful follow-up.
Recent FISH analysis of a pleomorphic AFH showed EWSR1 rearrangement but no FUS rearrangement.5 In the present study, two cases with pleomorphic features showed translocations involving the EWSR1 gene. Interestingly, one tumour showing FUS rearrangement reached a large size, in accordance with a previous report.16 Overall, 90% of our cases showed EWSR1 rearrangement resulting from either EWSR1–CREB1 or EWSR1–ATF1 fusion; the former represents the predominant event in AFH.9–14 There was no difference clinically, histologically or immunohistochemically between most cases with EWSR1 and one case with FUS rearrangement in the present study. Further study is necessary to clarify these points in a large number of cases with these rearrangements.
Pleomorphic MFH and myxofibrosarcoma, formerly known as myxoid MFH, arising in the subcutis or skeletal muscle of extremities and trunk, may be listed in the differential diagnosis of AFH showing pleomorphism and mitotic activity. Pleomorphic MFH is characterised by pleomorphic spindle cells arranged in a storiform or small fascicular pattern, with a varying amount of inflammatory infiltrate and myxoid and collagenous stroma. Myxofibrosarcoma is a malignant fibroblastic tumour with a variably myxoid stroma and pleomorphism. Immunohistochemically, neoplastic cells of both of the above tumours are diffusely positive for vimentin and variably positive for CD68, and CD34, α-smooth muscle actin, muscle-specific actin and desmin are expressed in a minority of cases.24 In difficult cases, FISH analysis may be helpful in distinguishing AFH from pleomorphic MFH and myxofibrosarcoma because the latter tumours lack specific chromosomal and genetic abnormalities.2 Desmin-positive AFH may often be confused with rhabdomyosarcoma, particularly in pleomorphic cases. Careful histological evaluation of AFH reveals cytologically bland spindle cells, unlike the clearly malignant round or spindle cells with deeply eosinophilic cytoplasm—that is, rhabdomyoblasts of rhabdomyosarcoma. In addition, AFHs do not express muscle-specific actin and myogenin (data not shown), which are expressed in nearly all rhabdomyosarcomas.
In summary, the present results of FISH analysis show that EWSR1 gene rearrangement is the most common genetic change in AFH. Wide surgical excision and careful follow-up are necessary for patients with AFH in view of its risk of local recurrence and metastasis leading to a fatal outcome.
Angiomatoid fibrous histiocytomas (AFHs) from patients who developed distant metastases and died from the disease showed focal proliferation of large pleomorphic cells with hyperchromatic nuclei and high proliferative activity (>10/10 high-power field and Ki-67 labelling index >10%).
There were no clinical, histological or immunohistochemical differences between the nine cases with EWSR1 rearrangement and one case with FUS rearrangement.
Wide surgical excision and careful follow-up are necessary for patients with AFH in view of its risk of local recurrence and metastasis leading to a fatal outcome.
Competing interests None.
Ethics approval This study was conducted with the approval of the Sapporo Medical University School of Medicine.
Provenance and peer review Not commissioned; externally peer reviewed.