Journal Logo


Cellular Storiform Collagenoma

Rudolph, Pierre; Schubert, Christoph; Parwaresch, Reza

Author Information
The American Journal of Dermatopathology: April 2003 - Volume 25 - Issue 2 - p 167-170

To the Editor:

We read with great interest the report by Chen and associates (1) on a case of a skin tumor designated pleomorphic sclerotic fibroma. A solitary lesion from a 72-year old man is described, which is claimed to share features of pleomorphic fibroma (2) and sclerotic fibroma (storiform collagenoma) (3–5).

The lesion presented as a flesh-colored papule that, on histologic sections, appeared as a circumscribed dermal nodule with zonal architecture beneath a flattened epidermis. The upper parts exhibited plump spindle-shaped and stellate cells, many of which were multinucleated, embedded at relatively high density in a haphazardly fibrillar or homogeneous collagenous matrix. Conversely, the lower portion displayed a hypocellular stroma of coarse hyalinized collagen bundles arranged in a whorled pattern and spaced by mucin-containing clefts, which is characteristic of storiform collagenoma. The superficial and deep areas of the tumor were connected by a transitional zone where both traits were present at variable degrees. Cytologically, the tumor cells were depicted as fusiform, stellate, or floret-like, with indistinct borders and mildly hyperchromatic, slightly enlarged nuclei with irregular contours and inconspicuous nucleoli. More than 80% of the fusiform cells, and more than 50% of the multinucleated cells stained positively for FXIIIa in the absence of S-100 or CD34 expression, the reaction pattern being similar in the superficial and deep portion of the tumor. CD68 expression was apparently not assessed. In consonance with two earlier reports (6,7), the authors chose the name of pleomorphic sclerotic fibroma for this case.

In our opinion, the interpretation of the mixed features in this cutaneous mesenchymal tumor is precarious, and the conclusion reached by Chen et al. concerning the classification of this neoplasm might need to be reconsidered. The lower portion of the tumor illustrated in their Figure 4 indeed shows typical features of a storiform collagenoma (sclerotic fibroma), and the overall architecture of the lesion is well in line with this diagnosis. Nevertheless, the superficial areas exhibit features that are not usually encountered in this type of lesion and make this case peculiar. Incidentally, we strongly favor the term “storiform collagenoma” as it emphasizes the characteristic whorled pattern of the profuse collagenous stroma, whereas we feel that “sclerotic fibroma” is a less descriptive and somewhat tautologous designation.

The superficial areas of the lesion described by Chen at al. are claimed to show features of pleomorphic fibroma. However, judging by the photomicrographs, the cytology of the large and occasionally multinucleated cells is not congruent with this diagnosis. Pleomorphic fibroma is a very distinctive tumor (Fig. 1A) characterized by a haphazardly-arranged fibrous stroma containing loosely scattered spindle-shaped and often multinucleated cells, many of which exhibit marked nuclear atypia with grotesquely enlarged multilobated and usually strongly hyperchromatic nuclei (2,8). Atypical mitotic figures are occasionally observed, but the overall mitotic activity is low (2).

FIG. 1.
FIG. 1.:
(B) Cytologic aspect of desmoplastic fibroblastoma; the poorly delimited tumor cells exhibiting plump, slightly pleomorphic nuclei with occasional angular contours and dense chromatin resemble those depicted in Figure 2 of the paper by Chen and coworkers (1). Overt atypia is absent. The abundant stroma is predominantly homogeneous, and haphazardly fibrillar in places. (A) Pleomorphic fibroma typically contains plump spindle-shaped cells with variably enlarged dark-staining nuclei and indistinct borders; in addition, bizarrely shaped hyperchromatic giant nuclei are noted. The stroma consists of hyalinized or wavy fibrillar collagen. (C) Giant cell collagenoma is characterized by distinctive multinucleated cells scattered diffusely through a storiform stroma composed of hyalinized collagen bundles spaced by mucin-containing clefts. The giant cells have sharply demarcated cytoplasmic borders often with angulated contours; the nuclei are monomorphous, round to oval and vesicular, with accentuated membranes and a single small basophilic nucleolus. Nuclear crowding is not infrequent.

We are unable to recognize these features in the tumor described by Chen at al. (1), and our skepticism is further fomented by the immunohistochemical findings in this case. Indeed, we have shown that the cells of pleomorphic fibroma strongly and consistently express CD34 while being thoroughly FXIIIa-negative (8), a finding that was amply corroborated by additional cases we received since the publication of this study. However, in the case reported by Chen et al. (1) the staining reaction is virtually opposite. We believe that this immunophenotype, together with the cytologic aspect of the lesion, is not compatible with the diagnosis of pleomorphic fibroma.

If the features in the superficial part of the lesion are not those of pleomorphic fibroma, what else do they resemble? Getting back to Figure 2 in the paper by Chen et al., we find that the morphology of the cells is strongly reminiscent of the larger stromal cells seen in fibrous papules/angiofibromas (9) or desmoplastic fibroblastomas (10,11). These lesions typically contain plump spindle-shaped or stellate cells with indistinct borders and enlarged hyperchromatic nuclei that may exhibit angulated contours (Fig. 1B); floret-type multinucleated cells may be found. Franc nuclear atypia is not a feature, but it also seems to be lacking in the illustration proffered by Chen and colleagues. Moreover, like the majority of the cells in their case (1), the cells of fibrous papule are known to express FXIIIa (12,13), and the centrofacial location tallies with a fibrous papule rather than with a pleomorphic fibroma (2,8). However, as the authors justly point out, their case did not display dilated vessels or a concentric perifollicular fibrosis as cutaneous adnexa were absent from the lesion (1), and stromal hyalinization in a storiform array is not usually encountered in fibrous papules or desmoplastic fibroblastomas.

The gradual transition from cellular into sclerotic areas suggests a continuum, i.e., an evolution or progression from one pattern into the other. Since the lower portion of the tumor appears to represent a typical storiform collagenoma, the superficial area may correspond to an active zone of tumor growth. This would be in keeping with the regular expression of FXIIIa in the slender interstitial cells of storiform collagenoma (14). Although the characteristic matrix of storiform collagenoma likely results from exuberant collagen synthesis (14,15), it is conceivable that a cellular—and possibly proliferative—stage precedes the formation of the hyalinized stroma characteristic of storiform collagenoma, and that the latter process advances from the base toward the top of the tumor. In this way, Chen et al. may have described an early stage of storiform collagenoma that retains traits reminiscent of fibrous papule. Incidentally, as FXIIIa expression is a hallmark of dermal dendrocytes (16,17), this tumor might be considered a peculiar form of “dermal dendrocytoma” (18). This could be regarded as an argument in favor of Zelger's interpretation of collagenoma being a variety of dermatofibroma (19,20). Remarkably though, even Zelger did not fancy a relationship between pleomorphic fibroma and dermatofibroma (20). Therefore, we estimate that “pleomorphic sclerotic fibroma”, being too plainly evocative of a probably unrelated condition (2,20), is not an appropriate name for the lesion described by Chen and associates. Because it remains conjectural that the case may represent an early stage of storiform collagenoma, terms like “cellular storiform collagenoma”, “mixed storiform collagenoma”, or even “pleomorphic sclerotic dendrocytoma” might be envisaged as more fitting denominations.

As we find in this report (1) no compelling evidence for a morphologic or histogenetic relationship between storiform collagenoma and pleomorphic fibroma, we also expressly object to the authors' conclusion that an entity described by us as giant cell collagenoma (14) might be reclassified under the broader term of pleomorphic sclerotic fibroma. Giant cell collagenoma is entirely distinct from classic pleomorphic fibroma (2) both cytologically (14) and immunophenotypically (8). Also, whilst giant cell collagenoma and the lesion described by Chen et al. have in common a peculiar stromal pattern, there are otherwise no similarities. First, the cases of giant cell collagenoma we had the opportunity to review to date were uniformly composed of a hyalinized storiform matrix and never contained areas of fibrillar or homogeneous stroma. Second, the distinctive giant cells were never clustered in one area of the tumor, but were evenly distributed within the storiform collagenous body of the lesion. Third, their cytology is entirely different as these giant cells may have bizarre shapes yet their nuclei are always round, monomorphous, and evenly sized, with an accentuated membrane, pale-staining chromatin, and small nucleoli (Fig. 1C). Hyperchromatic, pleomorphic or atypical nuclei were never encountered, nor were any mitotic figures observed in these cells. Lastly, none of the giant cells were found to express FXIIIa nor any other marker indicative of dermal dendrocyte lineage (21).

It emerges that storiform collagenoma may assume many faces. In addition to the classic form that is often associated with Cowden's disease (3), several other types have been described. One variety is characterized by the presence of numerous distinctive multinucleated cells devoid of nuclear atypia (14). Another type with marked cytologic atypia and increased proliferative activity was recently described as pleomorphic sclerotic fibroma (7). Because both the overall architecture and the immunohistochemical findings are congruent with storiform collagenoma, pleomorphic storiform collagenoma might be a more suitable name. Although the authors quoted pleomorphic fibroma as a differential diagnosis (7), they wisely refrained from suggesting a relationship between the two entities. The most recent report (1) portrays yet another variety of storiform collagenoma that is clearly distinct from the cases published by Martin-Lopez and colleagues (7). Unless this case merely represents an early stage of the common form, it might be rebaptized cellular storiform collagenoma.

In conclusion, we are pleased to note that the eremitic family of storiform collagenomas is celebrating the birth of new members. It is to be hoped that molecular biology will some day disclose the true nature of these intriguing neoplasms.

Pierre Rudolph

Christoph Schubert

Reza Parwaresch


1. Chen TM, Purohit SK, Wang AR. Pleomorphic sclerotic fibroma: a case report and literature review. Am J Dermatopathol 2002; 24:54–8.
2. Kamino H, Lee JY, Berke A. Pleomorphic fibroma of the skin: a benign neoplasm with cytologic atypia: A clinicopathologic study of eight cases. Am J Surg Pathol 1989; 13:107–13.
3. Weary PE, Gorlin RJ, Gentry Jr., WC, et al. Multiple hamartoma syndrome (Cowden's disease). Arch Dermatol 1972; 106:682–90.
4. Rapini RP, Golitz LE. Sclerotic fibromas of the skin. J Am Acad Dermatol 1989; 20:266–71.
5. Metcalf JS, Maize JC, LeBoit PE. Circumscribed storiform collagenoma (sclerosing fibroma). Am J Dermatopathol 1991; 13:122–9.
6. Garcia-Doval I, Casas L, Toribio J. Pleomorphic fibroma of the skin, a form of sclerotic fibroma: an immunohistochemical study. Clin Exp Dermatol 1998; 23:22–4.
7. Martin-Lopez R, Feal-Cortizas C, Fraga J. Pleomorphic sclerotic fibroma. Dermatology 1999; 198:69–72.
8. Rudolph P, Schubert C, Zelger BG, et al. Differential expression of CD34 and Ki-M1p in pleomorphic fibroma and dermatofibroma with monster cells. Am J Dermatopathol 1999; 21:414–9.
9. Meigel WN, Ackerman AB. Fibrous papule of the face. Am J Dermatopathol 1979; 1:329–40.
10. Evans HL. Desmoplastic fibroblastoma: a report of seven cases. Am J Surg Pathol 1995; 19:1077–81.
11. Miettinen M, Fetsch JF. Collagenous fibroma (desmoplastic fibroblastoma): a clinicopathologic analysis of 63 cases of a distinctive soft tissue lesion with stellate-shaped fibroblasts. Hum Pathol 1998; 29:676–82.
12. Nemeth AJ, Penneys NS. Fibrous papule: a tumor of fibrohistiocytic cells that contain factor XIIIa. J Am Acad Dermatol 1988; 19:1102–5.
13. Cerio R, Rao BK, Spaull J, et al. An immunohistochemical study of fibrous papule of the nose: 25 cases. J Cutan Pathol 1989; 16:194–8.
14. Rudolph P, Schubert C, Harms D, et al. Giant cell collagenoma: a benign dermal tumor with distinctive multinucleate cells. Am J Surg Pathol 1998; 22:557–63.
15. Shitabata PK, Crouch EC, Fitzgibbon JF, et al. Cutaneous sclerotic fibroma: immunohistochemical evidence of a fibroblastic neoplasm with ongoing type I collagen synthesis. Am J Dermatopathol 1995; 17:339–43.
16. Headington JT. The dermal dendrocyte. Adv Dermatol 1986; 1:159–71.
17. Cerio R, Griffiths CE, Cooper KD, et al. Characterization of factor XIIIa positive dermal dendritic cells in normal and inflamed skin. Br J Dermatol 1989; 121:421–31.
18. Cerio R, Spaull J, Jones EW. Histiocytoma cutis: a tumour of dermal dendrocytes (dermal dendrocytoma). Br J Dermatol 1989; 120:197–206.
19. Zelger BG, Sidoroff A, Zelger B. Combined dermatofibroma: co-existence of two or more variant patterns in a single lesion. Histopathology 2000; 16:529–39.
20. Zelger BG, Zelger B. Skin lesions of fibrocytic and fibrohistiocytic differentiation: a new concept and classification. In: Berry CL, Nesland J, Prat J, eds. Curr Top Pathol 94—Dermatopathology. New York: Springer Verlag, 2001, p. 103–40.
21. Rudolph P, Schubert B, Wacker HH, et al. Immunophenotyping of dermal spindle-cell tumors: diagnostic value of monocyte marker Ki-M1p and histogenetic considerations. Am J Surg Pathol 1997; 21:791–800.
© 2003 Lippincott Williams & Wilkins, Inc.