Case report K y RDS xan orna, atypical, differential diagnosis, immunohis- tochemistry, . prognosis A very atypical fibroxanthoma Atypicalfibroxanthoma d-iagnosed as malignant memnoma M. Melato, C. Rizzardi, R. Calacione and G. Trevisan SUMMARY An atypical fibroxanthoma characterized by aneuploidy and local aggressive behavior was misdiag- nosed as malignant melanoma. The reported case contributes to a better understanding of malignant fibro-histiocytic proliferations. A considered evaluation including immunohistochemistry is needed in diagnosing malignant melanoma. Introduction Poorly differentiatecl large celi malignancies of the skin frequently pose a diagnostic challenge for patholo- gists and ultimately many of them are diagnosed as amelanotic malignant melanoma, atypical fibrox- anthoma, pseudosarcomatous squamous celi carci- noma, or undifferentiated leiomyosarcoma and angiosa- rcoma. A predse diagnosis based on morphological features alone is often impossible, and immunohis- tochemistry is therefore mandatory (1). An atypical fibroxanthoma, misdiagnosed as malig- nant melanoma by an experienced pathologist, testi- fies to the difficulties that may be encountered in dif- ferentiating melanoma from the confusing family of fibrohistiocytic tumors. The present case, characterized by an unusually aggressive behavior, contributes to a wider cliscussion on these tumors. Case report In September 2000, a 66-year-old male presentecl with a mass in the right subclavicular region interpreted as the local recurrence of a malignant melanoma (pT4, sentinel lymph node negative), excised four months earlier. Brain CT, chest X-ray, and abdominal ultrasound scan were negative. Histological examination showed a dermal infiltra- tion by a clensely cellular population of epithelioicl and spinclle cells with abundant eosinophilic cytoplasm ar- ranged in a cliffuse ancl fascicular pattern (Fig. 1). The cells were arnelanotic and exhibited .pleornorphism, multinucleation ancl numerous typical and atypical mi- totic figures; a few multinucleated giant cells with ir- regularly distributed overlapping atypical nuclei were also present (Fig. 2) . By immunohistochemist1y, the tumor stained nega- Acta Dermatoven APA Vol 10, 2001, No 3 --------------------- --- ------ ---- 1 OJ A very atypical fibroxanthoma tive for S-100 and HMB-45. It was therefore tested for cytokeratin and vimentin, resulting negative for the former and strongly positive for the latter. A further panel of antibodies demonstrated in neoplastic cells the ex- pression of CD68, lysozyme, alphal-antichymotrypsin and, focally, of muscle-specific actin (HHF-35), as well as the lack of desmin, CD34 and factor VIII. The immu- nophenotypic profile justified the diagnosis of malig- nant fibrohistiocytic proliferation. The first excised material was obtained from a de- partment where it had been described as "Epithelioid and spindle cel! nodular melanoma (6.5 mm in depth; 5 mitoses x mm2) ", without performing immunohisto- chemistry. The histologic features of the lesi on - a promi- nent, ulcerated expansive nodule involving papillary and reticular dermis, containing elastotic material, and showing an epidermal collarette (Fig. 3)- were consis- tent with those previously reported. The staining pat- terns of the two lesions were identical. DNA ploidy analysis by flow cytometry demonstrated aneuploid distribution of nuclear DNA. In conclusion, the original diagnosis was revised and both the first excised material and its recurrence were diagnosed as atypical fibroxanthoma. Discussion Atypical fibroxanthoma is a pleomorphic tumor that usually occurs on the sun-damaged skin of the elderly. It is histologically indistinguishable from the pleomor- phic forms of malignant fibrous histiocytoma. However, from a conceptual point of view, it is classically consid- ered asa superficial form of malignant fibrous histiocy- toma, which, by virtue of its superficial location, almost invariably pursues a benign course, being merely char- acterized by a local aggressive behavior. This justifies its accurate recognition and differentiation, especially from malignant melanoma, undifferentiated squamous cell carcinoma, leiomyosarcoma, and angiosarcoma, in which immunohistochemistry plays a key role. The most common differential diagnosis concerns melanoma, a malignant tumor with a varied histologic appearance, and it can be particularly difficult to differ- entiate atypical fibroxanthoma from balloon cel! mela- noma (2) and spindle cel! melanoma, including des- moplastic and neurotropic types (3-4). Furthermore, although S100 protein stains a majority of these mela- nomas, the staining may be weak or focal, and HMB- 45, a more specific marker of melanoma, is frequently negative in desmoplastic and neurotropic melanoma. Other histiocytic proliferations than atypical fibro- xanthoma may mimic melanocytic tumors: epithelioid histiocytoma, juvenile xanthogranuloma, the adult form Figure 1. A fascicular pattern was focally evident both in the primary tumor and in the recurrence. Figure 2. Most neoplastic cells are pleomorphic and very atypical. Case report 104 ------------------------- - ---------Acta Dermatoven APA Vol 10, 2001, No 3 Case rep ort of the latter and reticulohistiocytoma. Particularly, Busa- m et al. have recently dealt with this problem, describ- ing three cases of xanthogranulomas with incospicuous foam cells and giant cells that were misdiagnosed as malignant melanoma (5). Atypical fibroxanthoma can be misdiagnosed also asa poorly differentiated squamous celi carcinoma but in that case immunohistochemistry (search for cytoche- ratins) easily resolves the problem. Similarly, atypical fibroxanthoma can be differentiated from leiomyosar- coma, using HHF-35 and desmin, and from angiosar- coma using CD34 and factor VIII , although cutaneous mesenchymal tumors could show some degree of ex- pression of CD34 and factor VIII (6). The immunoreac- tivity pattern of tumor cells positive for vimentin, lysozyme, alphal-antichymotrypsin and CD 68 will fi- nally justify the diagnosis of fibrohistiocytic prolifera- tion. Assigning our lesion into the family of fibrohistiocytic tumors is a further challenge. In fact , the nosology of these neoplasms has been long debated and particu- larly the categories "fibroxanthomas" and "xanthogranu- lomas" have a long histo1y of confusion surrounding them, which has been the source of continued contro- versy. What is truly needed is an easily understood clas- sification scheme that allows for precise diagnoses to be rendered on sections stained by hematoxylin and eosin; until that is devised, controversy is likely to con- tinue (7). Some authors even question the very existence of this entity, considering atypical fibroxanthoma as a di- agnosis of exclusion after ruling out other neoplasms (8) . They believe that atypical fibroxanthoma (similarly to its deeply located counterpart, malignant fibrous his- tiocytoma) represents a potpourri of histogenetically different, dedifferentiated tumors including sarcomas, carcinomas , melanomas, and lymphomas (9) , suppos- ing that one-da y this enigmatic entity will disappear from the textbooks because of a more sophisticated and con- siderate approach to this lesion (8). Many investigators have proposed that atypical fibroxanthoma may repre- sent a reactive process, while others contend that it is a trne fibrohistiocytic neoplasm, closely related to malig- nant fibrous histiocytoma (10). A review of the literature suggests that atypical fibroxanthoma can be differentiated from malignant fi- brous histiocytoma not only because of its superficial location (dermal) , absence of necrosis, vascular inva- sion, involvement of hypoderm, fascia and muscle, and no distanc metastasis, but also for an absent or just slight expression of LN2 (CD74) (11), and the diploid distri- bution of nuclear DNA, is considered by some authors as significant in understanding the biological behavior of the neoplasm (10). In the reported case, ali histological and clinical fea- A very atypical fibroxanthoma Figure 3. Primary tumor misdiagnosed as malignant melanoma. tures supported the diagnosis of atypical fibroxanthoma, except the DNA content. Indeed, aneuploid distribu- tion of DNA demonstrated by flow cytometry could be considered as indicative of a more aggressive behavior of the tumor that, in fact , recurred locally after surgical excision. In conclusion, we can confirm that fibrohistiocytic tumors of the skin must occasionally be evaluated and treated by a dermatologist and pathologist considering three categories of problems: (I) to distinguish them from other neoplastic processes , particularly from ma- lignant melanoma; (II) to order them cmrectly accord- ing to nosology, and (III) to evaluate appropriately their biologic behavior. We hope that the reported case will contribute to a better recognition of fibrohistiocytic neoplasms and to differentiation from malignant mela- noma. Acta Dermatoven APA Vol 10, 2001, No 3 --------------------------- -------1 OJ A very atypical .fibroxanthoma ll. E F E R E C E S l. Heintz PW, White CRJr. Diagnosis: atypical fibroxanthoma or not? Evaluating spinclle celi malignancies on sun damaged skin: a practical approach. Semin Cutan Med Surg 1999; 18: 78-83. AUTHORS' ADDRESSES 2. Kao GF, Helwig EB, GrahamJH. Balloon cell malignant melanoma ofthe skin. A clinicopathologic study of 34 cases with histochemical, immunohistochemical, and ultrastructural observations. Cancer 1992; 69: 2942-52. 3. Kanik AB, Yaar M, Bhawan J. P75 nerve growth factor receptor staining helps identify desmoplastic and neurotropic melanoma. J Cutan Pathol 1996; 23: 205-10. 4. Metcalf JS. Melanoma: criteria for histological diagnosis and its reporting. Semin Oncol 1996; 23: 688-92. 5. Busam KJ, Rosai J, lversen K, Jungbluth AA. Xanthogranulomas with incospicuous foam celis and giant celis mimicking malignant melanoma. Am J Surg Pathol 2000; 24: 864-9. 6. Altman DA, Nickoloff BJ, Fivenson DP. Differential expression of factor VIII and CD34 in cutaneous mesenchymal tumors. J Cutan Pathol 1993; 20: 154-8. 7. Cockereli CJ, Zafar K. Authors' Reply Am J Dermatopathol 1999; 21: 109-10. 8. Zelger BG, Soyer HP, Zelger B. Giant celi atypical fibroxanthoma : does it really exist? AmJ Dermatopathol 1999; 21: 108-10. 9. Fletcher CDM. Pleomorphic malignant fibrous histiocytoma: fact or fiction? A critical reappraisal based on 159 tumors diagnosed as pleomorphic sarcoma. AmJ Surg Pathol 1992; 16: 213-28. 10. Worreli JT, Ansari MQ, Ansari SJ, Cockereli CJ. Atypical fibroxanthoma: DNA ploidy analysis of 14 cases with possible histogenetic implications. J Cutan Pathol 1993; 20: 211-5. 11. Lazova R, Moynes R, May D, Scott G. LN2 (CD74). A marker to distinguish atypical fibroxanthoma f:rom malignant fibrous histiocytoma. Cancer 1997; 79: 2115-24. Mauro Melato, MD, Istituto diAnatomia patologica, Ospedale Maggiore, via Stuparich 1, I-34125 Trieste, Italy.E-mailmelato@univ.trieste.it Clara Rizzardi, MD, same address Roberta Calacione, MD, Universita di Trieste, Glinica Dermatologica, Ospedale di Cattinara, Strada diFiume 447, 34149 Trieste, Italy Giusto Trevisan, MD, projessor and chairman, same address Case report 106 ----------------------------------Acta Dermatoven APA Vol 10, 2001, No 3