ARTICLE
Dermatofibrosarcoma protuberans (DFSP) is a relatively common, predominantly
dermal neoplasm, which shows a predilection for young or middle-aged adults
and frequently arises on the trunk or thigh [1, 2]. Histological examination
shows a proliferation of closely aggregated fibroblast-like cells arranged
in a typical cartwheel or storiform pattern. The histogenesis of DFSP
has been disputed for many years. Ultrastructurally, the tumor cells have
been regarded as fibroblasts [3], whereas immunohistochemical studies
have suggested fibroblastic [4] or myofibroblastic differentiation [5].
It has been appreciated for some years that a small proportion of DFSP
incorporates a population of melanin-containing cells, which was described
for the first time by Bednar [6]. In recent years, several ultrastructural
and immunohistochemical studies have investigated the origin of fibroblast-like
cells or pigmented dendritic cells and interactions between them [7-12].
We herein report a case of Bednar tumor with interesting immunohistochemical
findings and discuss the relationship between these two types of cells.
Case report
A 47 year-old Japanese woman visited our clinic on September 3, 1996,
with a dark blue tumor on her right shoulder. Clinically, the eruption
was about 15 x 10 mm in diameter and consisted of a blue macule with nodules
and an irregular border, with an erythematous plaque on the upper side
of the macule (Fig. 1).
The lesion was surgically excised 3 cm apart from its margin and no recurrence
has been observed since that operation.
Materials and methods
The excised tissue specimen was divided into 3 pieces and used for histological,
ultrastructural and immunohistochemical studies. The tissue sample for
histology was fixed in 10% formaldehyde neutral buffer solution, embedded
in paraffin wax and 5-µm thick-tissue sections were then cut. These
were stained with hematoxylin and eosin, and by the Masson-Fontana technique
and Berlin blue for melanin and iron stain respectively.
Electron microscopy
The tissue sample for electron microscopy was fixed at 4° C in
2.5% glutaraldehyde buffered with 0.1 M phosphate buffer (pH 7.4) for
2 hrs. After postfixation in 1%, buffered osmium tetroxide for 1 hr, sections
were dehydrated in graded ethanol solutions and propylene oxide, and embedded
in Epon 812. Ultra-thin sections were made with an MT-5000 Ultra Microtome
(Sorvall, Newtown, CT, USA) equipped with a diamond knife (Diatome, Bienne,
Switzerland) and stained with 3.5% uranyl acetate and lead citrate. Sections
were examined with an H-300 electron microscope (Hitachi, Tokyo, Japan)
using a 0.03-mm objective aperture at an accelerating voltage of 75 kV.
The direct magnification range was 2,000-20,000 times.
Immunohistochemistry
Formalin-fixed, paraffin-embedded tissue sections were stained by a
labelled streptavidin-biotin method as described elsewhere [13, 14]. 5-µm
sections were deparaffinized and were incubated in 3% hydrogen peroxide
to reduce the activity of endogenous peroxidase, and then in bovine serum
to block non-specific staining. Primary antibodies against the following
agents were used: factor X IIIa purchased from Celsus Lab. (Cincinnati,
US) vimentin, S-100 protein, neuron-specific enolase (NSE), HMB 45 (human
melanoma), and MAC 387, all of which were obtained from DAKO (Kyoto, Japan),
and CD 34, which was purchased from Novocastra (Newcastle, UK). The details
of these antibodies are listed in Table
I. Negative controls were provided by using preimmune mouse or
rabbit serum. The sections were incubated for 30 min with primary antibodies,
incubated with biotinylated antibodies for 30 min, and were then incubated
with streptavidin-biotin for 10 min. All post-fixation procedures were
performed at room temperature. Aminoethyl-carbazole was used as the chromagen
and Mayer's hematoxylin served as the counterstain. Sections were scored
for staining intensity as follows: , no staining; +, weak, but clear
staining; ++, strong staining.
Results
Light microscopy
The tumor was located predominantly in the dermis and the overlying
epidermis was not involved. The tumor was not encapsulated but was relatively
well circumscribed, and a slight infiltration of the fatty tissue was
present. Neoplastic cells were predominantly plump and spindle-shaped,
and were arranged in a typical storiform pattern, whereas tumor cells
close to the pigmented cells showed a rather round or polygonal form rather
than a spindle form. Nuclear atypia was not conspicuous and mitotic figures
were not frequent. Pigmented bipolar or multipolar dendritic cells were
also arranged in small nests (Fig.
2). Masson-Fontana and Berlin blue staining of the pigmented dendritic
cells were strongly positive and negative, respectively.
Electron microscopy
Some tumor cells were spindle-shaped with rather folded or oval nuclei,
and frequently showed long, slender cytoplasmic projections, and others
contained lysosomes which represented phagocytotic melanosomes (Fig.
3A). The cytoplasm of these tumor cells contained relatively developed
rough-surfaced endoplasmic reticulum and mitochondria. No basal lamina
or pinocytotic vesicles were identified.
Pigmented dendritic cells were arranged in small nests among tumor cells
and closely apposed pigmented cells were connected by tight junctions.
They contained premelanosomes and mature non-membrane-bound type melanosomes,
with a predominance of the latter (Fig.
3B).
Immunohistochemistry
The results are summarized in Table
I. The majority of tumor cells exhibited positive staining for
vimentin and CD 34 (Fig. 4A),
although the intensity of CD 34 staining was decreased in tumor cells
around the dendritic pigmented cells. In contrast, antibody against factor
X IIIa showed a positive reaction for tumor cells only around melanin-containing
cells (Fig. 4B). Antibodies
against S-100, NSE, HMB-45 and MAC 387 elicited no staining. Melanin-containing
cells yielded a positive stain with antibodies against S-100 protein and
vimentin.
Discussion
Bednar tumor was described for the first time as a specific variant
of DFSP in 1957 [6], and since then, a small number of isolated cases
and short series have been reported [7-11]. Bednar tumor and DFSP share
many clinical and pathological features, and microscopically, the two
tumors are similar, except for the presence of melanin-containing dendritic
cells in the Bednar tumor. The biological behavior of both types of tumor
is considered to be intermediate malignancy, although local recurrences
have been less commonly observed in Bednar tumor [11]. In our case, the
local recurrence may be related to incomplete surgical resection as previously
described [7, 10].
In this study, histological and ultrastructural characteristics were
typically identified with those of Bednar tumor. There have been some
reports describing immunohistochemical localization of various antigens
in this tumor [7-12]. In general, spindle cells exhibit a positive reaction
for antibodies against vimentin and do not react with antibodies against
S-100, HMB-45 or NSE [8-12]. In dendritic melanin-containing cells, antibodies
against vimentin and HMB-45 usually show positive and negative reactions
respectively [8, 10, 12], and while some investigators have reported positive
reactions for antibodies against S-100 and NSE, another has reported a
negative reaction [12]. From these aspects, our immunohistochemical results,
except for factor X IIIa and CD 34, seem to be consistent with those previous
studies.
There have been a few reports describing the distribution of CD 34 or
factor X IIIa in this tumor [15, 16]. Akasaka et al. [15] noted
that tumor cells exhibited a positive and negative reaction against CD
34 and factor X IIIa respectively. Our immunohistochemical results do
not always agree with those above-mentioned reports. Interestingly, Mochizuki
et al. [17] demonstrated that the primary tumor in Bednar tumor
did not stain with CD 34, but that the recurrent lesion did. Therefore,
all of tumor cells may not express CD 34 because of the phenotypic variety
of tumor cells. We suggest that the interesting changes of staining patterns
of tumor cells around pigmented cells may be due to the relationship of
melanin-containing cells. The reason for this is not yet completely understood
and further investigation, especially as to the relation or interaction
between dendritic pigmented cells, is needed.
REFERENCES
1. Fletcher CDM, Evans BJ, Macartney JC, Smith N, Wilson JE, McKee PH.
Dermatofibrosarcoma protuberans: a clinicopathological and immunohistochemical
study with a review of the literature. Histopathol 1985; 9: 921-38.
2. Heenan PJ. Tumors of the fibrous tissue involving the skin. In: Elder
D, Elenitsas R, Jaworsky C, Johnson B Jr., eds. Lever's histopathology
of the skin. 8th ed. Philadelphia: Lippincott-Raven Publications,
1997: 847-87.
3. Alguacil-Garcia A, Unni KK, Goellner JR. Histogenesis of dermatofibrosarcoma
protuberans: an ultrastructural study. Am J Clin Pathol 1978; 69:
427-34.
4. Lautier R, Wolff HH, Jones RE. An immunohistochemical study of dermatofibrosarcoma
protuberans supports its fibroblastic character and contradicts neuroectodermal
or histiocytic components. Am J Dermatopathol 1990; 12: 25-30.
5. Ma CK, Sorb RJ, Gown AM. Immunohistochemical characterization of
atypical fibroxanthoma and dermatofibrosarcoma protuberans. Am J Clin
Pathol 1992; 97: 478-83.
6. Bednar B. Storiform neurofibromas of the skin, pigmented or non-pigmented.
Cancer 1957; 10: 368-76.
7. Dupree WB, Langloss JM, Weiss SW. Pigmented dermatofibrosarcoma protuberans
(Bednar tumor): a pathologic, ultrastructural, and immunohistochemical
study. Am J Surg Pathol 1985; 9: 630-9.
8. López JI, Elizalde JM, Fernandez-Larrinoa A. Pigment dermatofibrosarcoma
protuberans (Bednar tumour). Dermatol 1992; 184: 281-2.
9. Nakamura T, Ogata H, Katsuyama T. Pigmented dermatofibrosarcoma protuberans:
report of two cases as a variant of dermatofibrosarcoma protuberans with
partial neural differentiation. Am J Dermatopathol 1987; 9: 18-25.
10. Fletcher CDM, Theaker JM, Flanagan A, Krausz T. Pigmented dermatofibrosarcoma
protuberans (Bednar tumor): melanocytic colonization or neuroectodermal
differentiation? A clinicopathological and immunohistochemical study.
Histopathol 1988; 13: 631-43.
11. Ding J, Hashimoto H, Sugimoto T, Tsuneyoshi M, Enjoji M. Bednar
tumor (pigmented dermatofibrosarcoma protuberans). An analysis of six
cases. Acta Pathol Jpn 1990; 40: 744-54.
12. Onoda N, Tsutsumi Y, Kakudo K, Ozawa A, Niizuma K, Ohkid M, Osamura
RY. Pigmented dermatofibrosarcoma protuberans (Bednar tumor): an autopsy
case with systemic metastasis. Acta Pathol Jpn 1990; 40: 935-40.
13. Shikata N, Kurokawa I, Andachi H, Tsubura A. Expression of androgen
receptors in skin appendage tumors: an immunohistochemical study. J
Cutan Pathol 1995; 22: 149-53.
14. Hameed A, Hruban RH, Gage W, Pettis G, Fox WM. Immunohisto-chemical
expression of CD 68 antigen in human peripheral bood T cells. Hum Pathol
1994; 25: 872-6.
15. Akasaka T, Ohyama N, Kon S. A case of pigmented dermatofibrosarcoma
protuberans (Bednar tumor). J Dermatol 1997; 24: 390-4.
16. Kobayashi T, Hasegawa Y, Konohana A, Nakamura N. A case of Bednar
tumor. Immunohistochemical positivity for CD 34. Dermatol 1997;
195: 57-9.
17. Mochizuki Y, Narisawa Y, Kohda H. A case of Bednar tumor recurring
after 23 years. J Dermatol 1996; 23: 614-8.
|
Version imprimable
