A unique biphasic variant of cutaneous fibrous histiocytoma with a storiform pattern and intralesional pigmented melanocytes: “storiform melano-fibrous histiocytoma”

ARTICLE

Auteur(s) : Vidhya Nair¹, Ilan Weinreb¹, Niki Macneil¹, Zoltan Szollosi², Runjan Chetty¹, Danny Ghazarian¹

¹Department of Pathology, University Health Network, Toronto General Hospital, 200 Elizabeth Street, 11th Floor, Eaton Wing, Room 312, Toronto, M5G 2C4, Ontario, Canada
²Department of Pathology, University of Debrecen, Hungary

accepté le 18 Janvier 2008

A 30-year-old female presented with a nodular lesion on the calf of her right leg. It had been present for 10 years and there was no history of prior trauma to the site of the lesion. There had never been pain, tenderness or infection associated with this lesion. She elected to remove the lesion surgically for cosmetic reasons.

Materials and methods

Fluorescence In Situ Hybridization

Regarding the hybridization procedure, deparaffinized tissue sections were treated with a citrate solution (pH. 7.3) in an autoclave for 3 minutes at 121 °C. Pre-treated slides were incubated in a pepsin solution (100 mg/mL, Sigma-Aldrich, UK) for 6 minutes at 37 °C. Then the slides were post-fixed in buffered formalin. Pre-treated tissue sections and probes were denatured at 78 °C for 5 minutes and hybridized overnight at 37 °C in a hot plate (Hybrite chamber, Abbott Molecular-Vysis, Des Plaines, IL). Posthybridization washes were performed in a 50% formamide solution for ten minutes, three times, and afterwards, for ten minutes in a 2 × SSC solution and for five minutes in a 2 × SSC/0.1% NP-40 solution at 45 °C. Tissue sections were counterstained with 7 μL of 4,6-diamino-2-phenilindole (DAPI II counterstain, Abbott Molecular-Vysis, Des Plaines, IL). Results were analyzed in a fluorescent microscope (Olympus, BX51) using the Cytovision software (Applied Imaging, Santa Clara, CA). Appropriate positive controls (known DFSP harboring the translocation) were run in parallel.
Table 1 Immunohistochemical panel of antibodies

Results

Immunohistochemistry: The non-pigmented spindle cells were uniformly and strongly positive for CD34 and focally positive for Factor XIIIa (figures 2A and B). The non-pigmented spindle cells were also strongly positive for vimentin and focally positive for CD10 and were negative for stromelysin, Mac 387, Ham 56, KPI (macrophage markers), CK 5/6, Desmin, p63, Smooth muscle actin (SMA) and Epithelial Membrane Antigen (EMA). The pigmented cells were positive for S-100 (figure 3A), HMB-45, microphthalmia transcription factor and melanoma cocktail (figure 3B) but negative with vimentin, CD 34 and factor XIIIa. The MIB-1 proliferative index in both components was less than 1%.

Cytogenetic analysis: No cytogenetic abnormality could be demonstrated in the chromosomal regions examined; hence the DFSP specific translocation was not detected in this tumor.

On the basis of the well-circumscribed nature of the lesion without extension into subcutaneous fat, lack of mitoses and absence of the t (17;22) translocation, this lesion was categorized as a DF with interspersed melanocytes (A Storiform Melano-Fibrous Histiocytoma). Despite the strong diffuse CD34 immunoreactivity, it was felt that the aforementioned features favored a diagnosis of DF rather than DFSP.

Discussion

It is common to find a variety of epidermal changes overlying DF, including epidermal hyperplasia, basal layer hyperpigmentation, and less commonly basal cell carcinoma, follicular basal cell hyperplasia, squamous cell carcinoma in situ, sebaceous hyperplasia, and seborrheic keratosis-like changes [4-6]. It has been suggested that DF create an environment that induces hyperplasia of the overlying epidermis, perhaps through increased epidermal growth factor receptor expression [7].

DF clinically have a brownish color due to hyperpigmentation of the overlying skin. The number of tyrosinase immunopositive melanocytes in the pigmented epidermis in DF is significantly increased when compared with nonlesional normal epidermis. Reverse transcription polymerase chain reaction analysis of mRNAs encoding stem cell factor and hepatocyte growth factor has shown that there is accentuated expression of both factors which have transcripts in the lesional DF dermis compared with the non-lesional dermis [8].

In a study conducted by King et al., 14 cases of melanocytic lesions arising in association with DF occurring over a period of 4 years were retrieved [9]^. In all cases there was accompanying epidermal hyperplasia with variable basal layer hyperpigmentation. The melanocytic lesions included a lentiginous junctional nevus (2), dermal nevus (3), compound nevus (8) and superficial spreading type malignant melanoma in situ (1) [9]. However, they did not encounter the presence of melanocytes interspersed amongst the typical spindle cells within the DF cases they examined.

There has been a case report of a malignant melanoma occurring in association with a DF. It has been postulated that DF could be due to a reactive process secondary to the malignant melanoma or it could occur as part of a collision tumor [7]. In our case, benign melanocytes were interspersed within the fibrohistiocytic lesion and there was no junctional or intraepidermal component associated with it. There was no cytologic atypia of either component and the MIB-1 index was very low (< 1%) making an atypical melanocytic lesion most unlikely.

There are several lesions that have to be considered morphologically in the differential diagnosis (table 2). Desmoplastic nevus can be mistaken for dermatofibroma; features that are shared include epidermal hyperplasia, hyperpigmentation of the overlying epidermis, and the presence of a desmoplastic stroma, keloidal collagen and multinucleated cells. Desmoplastic nevus does not show adnexal induction, a characteristic feature of histiocytic lesions. The presence of nested junctional melanocytes, an increase in non-nested junctional melanocytes, and dermal melanocytic nests differentiate desmoplastic nevus from dermatofibroma. Ki-67 positive cells are more frequently observed in DF compared to desmoplastic nevi [10]. In our case the lesion was predominantly fibro-histiocytic as evidenced by the plump spindle cells being diffusely positive for CD34 and focally for factor XIIIa. The melanocytes admixed with these spindle cells were positive for S-100, HMB-45, microphthalmia factor and the melanoma cocktail.

The common blue nevus ranges from an inconspicuous to an obvious diffuse spindle cell population within the dermis. The melanocytes are bipolar, dendritic, finely pigmented and associated with variable numbers of melanin-laden macrophages containing large aggregates of melanin [11]. Cellular blue nevus is composed of oval, spindled, fusiform and epithelioid melanocytes occupying the deep dermis and extending along the adnexae and into the subcutaneous tissue. The cells have eosinophilic or clear cytoplasm with small nucleoli. Amelanotic cellular blue nevus has been reported where the nevic cells have minimal pigmentation and stain focally for S-100 protein and HMB-45. However, these are negative for CD 34 and do not exhibit a fairly well circumscribed storiform pattern as seen within this lesion [12]. However, in contrast to the usual focal positivity in DF, the case in this paper showed diffuse strong CD 34 positivity. This is an unusual finding in typical DF but must be interpreted in the context of the other histologic findings and the absence of the translocation.

Dermatofibrosarcoma protuberans (DFSP) histologically show a range of architectural and cytological patterns, the most characteristic of which is a poorly circumscribed, uniform population of spindle cells that are arranged in a monomorphous storiform pattern, which extends to the subcutis, often with infiltration in a lace-like or linear fashion into the adipose tissue. The epidermis over the lesion is usually normal or atrophic. Cutaneous adnexae are often entrapped within the tumor, in contrast to fibrous histiocytoma [11]. Bednar tumors are rare and comprises about 1-5% of the DFSPs. They resemble classical DFSPs clinicopathologically except for the presence of melanocytes, which stain positively for HMB-45, and S-100 [13]. There has been a case report of Bednar tumor with associated “dermal melanocytosis” pointing towards a neuromesenchymal origin of this tumor [14].

The lesion described herein has the following features that favor a diagnosis of DF: long-standing clinical history with no associated change in size or characteristics of the lesion, small size, superficial location circumscription without subcutaneous fat extension, absence of mitoses and the lack of the t (17; 22) translocation. Features pointing to a DFSP are: a lesion composed of spindle cells with few secondary elements arranged in a storiform pattern, no associated epidermal hyperplasia and diffuse immunopositivity with CD34, and focal positivity for CD10 and Factor XIIIa [15, 16]. This lesion raises considerable difficulty and obvious concern in terms of diagnostic classification. Does one use morphology as the gold standard for diagnosis in concert with the cytogenetic findings, thereby making this lesion a DF with melanocytes, or is this a DFSP in view of the diffuse CD34 positivity and absence of epidermal hyperplasia? We feel that the weight of evidence favors the former diagnosis. This case highlights the difficulty of reliance on CD34 as a distinguishing factor between DF and DFSP. It is felt that despite the strong diffuse immunoexpression of CD34, the long clinical history, the histological features and absence of the DFSP translocation make this lesion a DF. An additional unusual feature is the presence of benign melanocytes within the lesion, leading to the descriptive term of “storiform melano-fibrous histiocytoma”.
Table 2 Histologic differential diagnosis

Acknowledgements

Conflict of interest: none. Financial support: none.

References

2 Wagamon K, Somach SC, Bass J, Sigel JE, Xue W, Schluchter M, Jaworsky C. Lipidized dermatofibromas and their relationship to serum lipids. J Am Acad Dermatol 2006; 54: 494-8.

3 Kahn HJ, Fekete E, From L. Tenascin differentiates Dermatofibroma from dermatofibrosarcoma Protuberans: Comparison with CD 34 and Factor XIIIa. Hum Pathol 2001; 32: 50-6.

4 Herman KL, Kantor GR, Katz SM. Squamous cell carcinoma in situ overlying dermatofibroma. J Cutan Pathol 1990; 17: 385-7.

5 Goette DK, Helwig EB. Basal cell carcinomas and basal cell carcinoma-like changes overlying dermatofibromas. Arch Dermatol 1975; 111: 589-92.

6 Cheng L, Amini SB, Tarif Zaim M. Follicular basal cell hyperplasia overlying dermatofibroma. Am J Surg Pathol 1997; 21: 711-8.

7 Kovach BT, Boyd AS. Melanoma associated with a dermatofibroma. J Cutan Pathol 2007; 34: 420-2.

8 Shishido E, Kadono S, Manaka I. The mechanism of epidermal hyperpigmentation in dermatofibroma is associated with stem cell factor and hepatocyte growth factor expression. J Invest Dermatol 2001; 117: 627-33.

9 King R, Googe PB, Page RN, Mihm MC. Melanocytic lesions associated with dermatofibromas: a spectrum of lesions ranging from junctional nevus to malignant melanoma in situ. Mod Pathol 2005; 18: 1043-7.

10 Harris GR, Shea CR, Horenstein MG, Reed JA, Burchette JL, Prieto VG. Desmoplastic (sclerotic) nevus: an underrecognized entity that resembles dermatofibroma and desmoplastic melanoma. Am J Surg Pathol 1999; 23: 786-94.

11 Santa Cruz DJ. Tumours of the skin. In: Fletcher CDM, ed. Diagnostic Histopathology of tumors. Volume 2. Elsevier Limited, 2007: 1423-526.

12 Carr S, See J, Wilkinson B, Kossard S. Hypopigmented common blue nevus. J Cutan Pathol 1997; 24: 494-8.

13 Kaburagi Y, Hatta N, Kawara S, Takehara K. Pigmented dermatofibrosarcoma protuberans (Bednar tumor) occurring in a Japanese infant. Dermatol 1998; 197: 48-51.

14 Goncharuk V, Mulvaney M, Carlson JA. Bednar tumor associated with dermal melanocytosis: melanocytic colonization or neuroectodermal multidirectional differentiation? J Cutan Pathol 2003; 30: 147-51.

15 Enzinger FM, Weiss SW. Fibrohistiocytic tumors of Intermediate malignancy. In: Enzinger and Weiss, ed. Soft tissue tumors. 4th ed. St Louis: MO: Mosby, 2001: 491-534.

16 Lee KJ, Yang JKM, Lee ES, Lee DY, Jang KT. CD 10 is expressed in dermatofibromas. Br J Dermatol 2006; 155: 632-3.