Selected Case from the Arkadi M. Rywlin International Pathology Slide Seminar: Desmoplastic Fibroblastoma (Collagenous Fibroma)

Diagnosis:

Desmoplastic fibroblastoma (collagenous fibroma), subcutis, fascia, and muscle of the chest in a patient with an unrelated, synchronous, cutaneous malignant melanoma.

Case ID and Referral Source:

AMR # 31 case 1; FMC 99S02606; STTR T2943. Case originally referred by Dr. Michael Brown, ACT Pathology, Woden, ACT 2606, Australia.

History:

The lesion was an encapsulated mass of firm white tissue 50×50×15 mm, located in the deep subcutis attached to deep fascia, muscle overlying the rib in the anterior lower chest in line with the nipple. A Clark Level-2 melanoma was removed from the right thigh at the same time. Multiple copies of H&E stained slides were submitted to the Club with the additional information that the tumor cells were S-100 positive.

Author's Comments to the Club:

I think this is a desmoplastic fibroblastoma (collagenous fibroma ) as described by Harry Evans and subsequently by Nielsen, O'Connell, Dickersin, and Rosenberg as collagenous fibroma. These papers were followed by Miettinen and Fetsch's large series and a recent cytogenetic study.

Before circulating the case to the Club, I had referred sections to the Australasian Soft Tissue Tumor Registry Panel as an unknown case. Panel diagnoses included hyalinized fasciitis, benign peripheral nerve sheath tumor, fibroma, neurofibroma, myxoid neurofibroma, myofibroblastic tumor of low grade malignant potential, low grade fibromyxoid sarcoma, desmoid tumor, deep fibromatosis, fibromatosis, atypical lipoma, angiomyxoma, low grade sarcoma, and myxoma. Only two of the panel members diagnosed desmoplastic fibroma.

Club Members' Opinions:

The opinions, given in the knowledge that desmoplastic fibroblastoma was the suggested diagnosis, were: desmoplastic fibroblastoma (collagenous fibroma) 20, myxoid neurofibroma 1, maturing nodular fasciitis 1, ? myxoma 1, ? low-grade fibromyxoid sarcoma 1, ? fibromatosis 1, and uncertain 2.

Some of the members' comments were as follows:

1. This is a difficult case. I just had a case of proven desmoplastic melanoma that was in some areas identical to this case. I agree that desmoplastic fibroblastoma is an excellent diagnosis, but I would run microphthalmia transcription factor (Mitg), the most recently described melanoma marker, in order to rule out the diagnosis of desmoplastic melanoma.

2. Whatever one wants to call this, I think it is benign. Myxoma? Desmoplastic fibroma?

3. An interesting array of Australian opinions! The morphology of the cells looks myofibroblastic. Occasional mitotic figures are evident. An extravasation of blood is present. The myxoid and collagenous stroma favors this being a maturing nodular fasciitis.

4. Agree with the diagnosis. If I had not been aware that the S-100 can be positive in collagenous fibroma, it would have been a more difficult diagnosis for me.

5. I believe that desmoplastic fibroblastoma is the best diagnosis, although some might prefer a diagnosis of low-grade malignant myofibroblastic tumor to avoid challenge in case of tumor recurrence.

6. I found it difficult making the differential diagnosis with fibromatosis.

7. Collagenous fibroma, I agree. I recently saw another case in a local slide seminar, and many pathologists offered a spectrum of diagnoses comparable to that of the Australian Panel.

8. This is indeed a pretty example of desmoplastic fibroblastoma (collagenous fibroma). I think that these lesions are probably much more common than the older literature has suggested. This likely reflects that fact that many cases are mislabeled as bunt-out nodular fasciitis, despite the fact that the clinical history in fasciitis is always shorter and, furthermore, fasciitis tends to disappear entirely if left untreated.

9. Instances like this make me so acutely aware of my own fallibility. Personally, I tend to agree with the diagnosis of collagenous fibroma. However, at one or two points on my slide, the process is seen infiltrating skeletal muscle fibers reminiscent of a fibromatosis-type lesion. I would not be very surprised if the lesion recurred regardless of what one calls it.

10. I also favor collagenous fibroma though in several fields the stroma is rather myxoid. Focally the cellular morphology is very similar to proliferative fasciitis. Could collagenous fibroma represent late (collagenous) phase of proliferative fasciitis?

11. Collagenous fibroma. It is interesting how many people were in favor of a nerve sheath tumor, presumably because of the S-100 positivity. On H&E, this lesion does not appear as a nerve sheath tumor to me.

12. I agree with a diagnosis of desmoplastic fibroblastoma or collagenous fibroma. The cytology is good for it, and it has a multinodular appearance. Two of Nielsen's cases showed diffuse but weak cytoplasmic S-100 activity.

13. Agree with desmoplastic fibroblastoma. I would not have even thought of desmoplastic melanoma given the bland cytologic appearance and edematous-myxoid stroma and absence of neurotropism. The somewhat stellate and dendritic nature of the spindle cells is also inconsistent with a schwannoma (despite S-100 protein positivity). The issue of whether the “name” of desmoplastic/collagenous fibroma/fibroblastoma is the best term for this lesion is, in my mind, subjective and of secondary importance. Reactive fibro/myofibroblastic proliferations appear to show a broad spectrum of histologic features with frequent overlaps being encountered. Whether one wishes to interpret some of these lesions as a “burnt-out” or late stage of nodular fasciitis, or one prefers to make up a more distinctive and catchy name such as “collagenous fibroblastoma” is, for all practical purposes, irrelevant. The story of reactive fibro/myofibroblastic proliferations still remains to be unravelled, and a more unifying concept perhaps would help shed some sanity on this topic. In fact, Cyril Toker published cases of nodular fasciitis showing clear histologic overlap with fibromatosis in an unfortunately little-known paper (¹).

Author's Additional Comments:

In addition to the series referred to above (^2–5), there have been a few other recent papers on this entity (^5–12). Based on the large series of 63 cases from the Armed Forces Institute of Pathology (⁴), 80% of patients are men. The median age is 50 years (range 16 to 81 years), the history is usually of a slowly growing, painless mass ranging in size from 1 to 20 cm (median 3 cm). Most lesions are subcutaneous but approximately 25% involve skeletal muscle. Tumors arise in the trunk, limbs, and girdles with roughly equal frequency. Grossly, they are pearl-gray, firm and homogeneous, and appear well circumscribed but microscopically, they may infiltrate the surrounding fat or muscle. Histologically (Figures 1 and 2), they are sparsely cellular and consist of stellate and spindle-shaped fibroblasts and myofibroblasts separated by a collagenous or myxocollagenous matrix. Mitoses are minimal or absent. Tumor cells are focally positive for muscle-specific and alpha smooth muscle actins, and rarely for keratins. Desmin and CD34 are not expressed. After the Club Members' opinions had been received, I performed CD34 and bcl-2 stains. Both were negative. I was unable to obtain the antibody for microphthalmia transcription factor.

Fig. 1

Fig. 2

As none of the reported tumors have recurred, simple conservative excision is the treatment of choice.

The broad range of diagnoses suggested by the Australasian Soft Tissue Tumor Registry Panel probably indicates lack of awareness of this rare, recently described entity rather than informed disagreement with the diagnosis. Comments by the Club members, made in the knowledge that desmoplastic fibroblastoma was the suggested diagnosis, indicate a majority acceptance of the existence of this “new” entity. However, a minority expressed reservations and suggested alternate diagnoses, including neurofibroma, metastatic malignant melanoma, aggressive fibromatosis, fasciitis, and myxoma.

The positive S-100 stain raised the possibility of a neurofibroma, particularly amongst members of the Australasian Registry Panel. While most reported desmoplastic fibroblastomas have been S-100 negative (⁴), two cases have been positive (³) and this, coupled with the absence of sinuously curved tumor nuclei and undulating collagen bundles, are points against neurofibroma. The patient's history of malignant melanoma and the positive S-100 invoke metastatic melanoma. A solitary soft tissue metastasis excised at the same time as the primary in the absence of nodal metastases would be unusual. Moreover, the histology is bland and is not particularly suggestive of melanoma. I have no information about the staining of desmoplastic fibroblastoma for microphthalmia transcription factor, but I doubt that enough cases have been studied to apply the results to this case.

“Fibromatosis” is a generic term that includes the Dupuytren's types (palmar and plantar fibromatosis and knuckle pads), as well as the more aggressive and generally deeply situated desmoid fibromatoses (abdominal wall desmoids, extra-abdominal desmoids or aggressive fibromatosis, intracavity desmoids, desmoids of bone, etc). While desmoids occasionally exhibit hypocellular, collagenized nodules, they are mainly composed of moderately cellular fibrovascular bundles in which the nuclei are aligned more or less in one direction. By contrast, the nuclei in desmoplastic fibroblastomas exhibit no particular alignment or polarity and the tumor is hypocellular overall (Figures 1 and 2). However desmoplastic fibroblastoma may occasionally infiltrate muscle as occurred in this case.

The fact that the reported cases have not recurred is pointer against a form of fibromatosis and the lack of metastases rule out a sarcoma.

It is difficult to absolutely exclude the possibility of a nonspecific scar or a regressed nodular fasciitis, but fasciitis disappears completely if it is incompletely removed and the slow growth and comparatively large size of some desmoplastic fibroblastomas make a relationship to fasciitis difficult to accept.

Desmoplastic fibroblastoma is much less myxoid and more fibrous than intramuscular myxoma and is less vascular than superficial angiomyxoma.

The recent cytogenetic findings of abnormalities on the long arm of chromosome 11 in two cases raise the possibility of a relationship of desmoplastic fibroblastoma to fibroma of tendon sheath, which also shows rearrangements in the same region (⁵).

As to the “correct” name, Harry Evans, the person who first recognized the tumor, proposed desmoplastic fibroblastoma (²). Those who benefited by his discovery (³) took exception to both “desmoplastic” and “fibroblastoma,” ignored the convention that allows the “Father” to name his offspring, and proposed “collagenous fibroma.” While it is tempting to assume the role of a judge, past experience suggests that the choice of name will eventually be determined by the tyranny of common usage.

Editor's Summary:

* Desmoplastic fibroblastoma (collagenous fibroma) is a rare and recently described benign soft tissue tumor.

* It is more common in men, median age is 50 years, the history is usually long, and the usual size is around 3 cm.

* Most lesions are subcutaneous but approximately 25% involve skeletal muscle.

* Histologically, they are sparsely cellular with stellate and spindle-shaped fibroblasts—myofibroblasts that are separated by a collagenous or myxocollagenous matrix. Mitoses are minimal or absent.

* Tumor cells are focally positive for muscle-specific and alpha smooth muscle actins, and rarely, for keratins. The S-100 is usually negative but a very few cases have been positive.

* It can be confused with one of the fibromatoses, myxomas of various kinds and neural tumors.

* It neither recurs nor metastasizes.