Share this article on:

Pigmented Atypical Fibroxanthoma, a Dermatofibroma Variant?: Pigmented Atypical Fibroxanthoma: Not a Variant of Dermatofibroma: Concept that Atypical Fibroxanthoma Is Squamous Cell Carcinoma Is a Misconstruction: Reply

Diaz-Cascajo, Carlos MD

The American Journal of Dermatopathology: February 2004 - Volume 26 - Issue 1 - p 86-87
Letters to the Editor

First, I would like to thank Dr. B. Zelger for praising the quality of figures of our paper that led him to the conclusion that the figures show a pigmented dermatofibroma (DF) rather than a pigmented atypical fibroxanthoma (AFX) and, by extension, that pigmented AFX is a fiction, representing either a so-called aneurysmal or angiomatoid dermatofibroma or a squamous-cell carcinoma. Unfortunately, contrary to Dr. Zelger's opinion, it is now my impression that photomicrographs were not good enough to allow him to recognize pigmented AFX, mistaking it for DF.

The reasons argued by B. Zelger to make this point are, in my opinion, poorly substantiated and highly subjective:

Back to Top | Article Outline


  1. Clinically, pigmented atypical fibroxanthoma (AFX) arises on the face, including the ears, of older persons, whereas other locations are rare. On the contrary, aneurysmal DF often appears on the limbs of middle-aged adults. 1 In large series of aneurysmal DF reported in the literature, including 40 cases analyzed by Calonje and Fletcher, 2 no single case of aneurysmal DF arising on the face was found. 1,2
  2. Microscopically, pigmented AFX appears as a relatively well-circumscribed dermal nodule with infiltrative borders located in sun-damaged skin. Neither acanthotic epidermis, fibrosclerotic collagen within the tumors nor variable amounts of lymphocytes within and around the lesion, as wrongly pointed out by Dr. Zelger, were present in our series. Furthermore, the antibody KiM1p, which, according to Dr. Zelger, shows “variably reactivity” in our series, has not been applied in our cases.
  3. In regard to cytologic findings of pigmented AFX, as mentioned and illustrated in our paper, neoplastic cells show prominent cellular atypia including cellular pleomorphism, nuclear hyperchromatism, prominent nucleoli, irregular nuclear membrane, and numerous typical and atypical mitoses (see Fig. 2 and Fig. 4). Neoplastic mononuclear cells and multinucleated giant cells with bizarre nuclei often show erythrocytes, hemosiderin, and even hyaline bodies in the cytoplasm (see Fig. 3 and Fig. 4).
  4. Microscopically, aneurysmal DF is often covered by hyperplastic epidermis. Neoplastic cells show bland appearance ranging from short spindle-shaped cells to histiocyte-like cells intermixed with hemosiderin-containing giant cells and macrophages. Cytologic atypia is usually absent or mild, 2 which contrasts with the prominent cellular atypia of pigmented AFX. Contrary to AFX, atypical mitotic figures in aneurysmal DF have never been described.
  5. Immunohistochemically, according to Calonje and Fletcher, 2 neoplastic cells of aneurysmal DF are factor XIIIa and CD 68 negative, which contrasts with our immunohistochemical results in pigmented AFX, where a number of neoplastic cells expressed factor XIIIa and were weakly positive for CD68.

Because of the clear-cut clinicopathological difference between pigmented AFX and aneurysmal DF, I would like to remind Dr. Zelger that aneurysmal DF is not the only tumor showing angiomatoid or “aneurysmal” spaces and, therefore, angiomatoid spaces can not be used as a major argument in the differential diagnosis AFX versus DF. In fact, similar angiomatoid spaces have been described in a series of tumors, often of fibrohistiocytic derivation, such as angiomatoid malignant fibrous histiocytoma, a low-grade sarcoma in children and young adults, some pleomorphic malignant fibrous histiocytomas (MFH), AFX, and even in dermatofibrosarcoma protuberans. 3,4

In sum, for the above-mentioned reasons, pigmented AFX and pigmented DF are different entities clinically, histopathologically, and apparently immunohistochemically also.

Back to Top | Article Outline


In my view, AFX is unlikely to represent a variant of squamous-cell carcinoma (SCC) for the following reasons:

  1. Atypical fibroxanthoma lacks squamous differentiation (keratinization of solitary or groups of cells within the tumor, formation of horn pearls, intercellular bridges, etc.).
  2. There is an absence of carcinoma in situ in adjacent structures covered by squamous epithelium within or in the vicinity of the neoplasm. In some cases, however, the overlying epidermis of AFX shows actinic keratosis, or even carcinoma in situ. This may lead to the wrong conclusion that the neoplasms in the epidermis and in the underlying dermis represent the same lesion. In those cases, however, vimentin and cytokeratins help to distinguish them: actinic keratosis is cytokeratin positive and vimentin negative, whereas AFX reacts with vimentin and lacks immunoexpression for any available cytokeratin. In this regard, Dr. Zelger's argument that, in cases of his series, primary melanomas and leiomyosarcomas were negative for melanocytic and smooth muscle markers, respectively, becoming later positive in the recurrences and/or metastases can not be extrapolated to SCC. In fact, SCC has been reported to react invariably with cytokeratins of low and high molecular weight, both in the skin and outside of the skin. For the same reasons, negative immunostaining of certain “dedifferentiated” carcinomas of the thyroid gland, stomach, etc., for cytokeratin can not be used as argument, as Dr. Zelger does, to justify the fact that, contrary to SCC, AFX is invariably cytokeratin negative and, for this and other above-mentioned reasons, cannot be called SCC. Furthermore, AFX is unlikely to represent a dedifferentiated neoplasm, as demonstrated by histopathological overlap of these tumors at different stages of their evolution. For this reason, Zelger's considerations about negative cytokeratin stain in dedifferentiated extracutaneous carcinomas can not be applied to AFX.
  3. Squamous-cell carcinoma of the skin is surrounded by a variably dense, chronic inflammatory infiltrate with numerous plasma cells, a feature not seen in AFX.

Finally, at the end of his letter, Dr. Zelger pointed out that, in his paper “Pleomorphic MHF: Fact or Fiction”, C.D.M. Fletcher “was able to demonstrate that pleomorphic MFH. . . . is a potpourri of different entities”. 5 Reading Fletcher's paper I arrived at the conclusion that a series of sarcomas and even carcinomas are able to undergo dedifferentiation along the disease, resulting in highly pleomorphic tumors reminiscent of pleomorphic MFH. However, this does not exclude “per se” pleomorphic MFH as a distinctive entity.

Carlos Diaz-Cascajo, MD

Back to Top | Article Outline


1. Santa Cruz DJ. Aneurysmal (“angiomatoid”) fibrous histiocytoma of the skin. Cancer. 1981; 47:2053–2061.
2. Calonje E, Fletcher CDM. Aneurysmal benign fibrous histiocytoma.: clinicopathological analysis of 40 cases of a tumor frequently misdiagnosed as a vascular neoplasm. Histopathology. 1995; 26:323–331.
3. Enzinger FM. Angiomatoid malignant fibrous histiocytoma. A distinct fibrohistiocytic tumor of children and young adults simulating a vascular neoplasm. Cancer. 1979; 44:2147–2157.
4. Costa MJ, Mcglotsen L, Pierce M, et al. Angiomatoid features in fibrohistiocytic sarcomas. Immunohistochemical, ultrastructural, and clinical distinction from vascular neoplasms. Arch Pathol Lab Med. 1995; 119:1065–1071.
5. Fletcher CDM. Pleomorphic malignant fibrous histiocytoma: Fact or Fiction? A critical reappraisal based on 159 tumors diagnosed as pleomorphic sarcoma. Am J Surg Pathol. 1992; 16:213–228.
© 2004 Lippincott Williams & Wilkins, Inc.