Report
An immunohistochemical study of the origin of the solid
strand in syringoma, using carcinoembryonic antigen,
epithelial membrane antigen, and cytokeratin 5
Byung Chul Kim1, MD, Eun Joo Park1, MD, In Ho Kwon1, MD, Hee Jin Cho2, MD,
Hye Rim Park3, MD, Kwang Ho Kim1, MD, and Kwang Joong Kim1, MD
1
Department of Dermatology, College of
Medicine, Hallym University, Hallym
University Sacred Heart Hospital,
Dongan-gu, Anyang-si, Gyeonggi-do,
Korea, 2Department of Dermatology,
College of Medicine, Hallym University,
Chuncheon Sacred Heart Hospital,
Chuncheon-si, Gangwon-do, Korea, and
3
Department of Pathology, College of
Medicine, Hallym University, Hallym
University Sacred Heart Hospital,
Dongan-gu, Anyang-si, Gyeonggi-do,
Korea
Abstract
Background Although much research has been conducted into the origin of syringoma,
the histogenesis and differentiation of it remains controversial. The published studies examined various antibodies, and our study is an additional immunohistochemical work-up.
Objective We attempted to identify the cell that acts as the precise origin of a syringoma,
based on a comparative analysis of carcinoembryonic antigen (CEA), epithelial membrane
antigen (EMA), and cytokeratin (CK) 5 through immunohistochemical staining in the solid
strand of basophilic epithelial cells of syringoma.
Methods A total of 31 patients with biopsy-confirmed syringoma were included in this
study. Each sample was analyzed with antibodies to CEA, EMA, and CK5. These markers
were indicating each part of the normal sweat gland structure: CEA stains the luminal surface of sweat ductal structures; EMA stains the peripheral cells of normal dermal ducts and
Correspondence
Dr Kwang Ho Kim, MD
Department of Dermatology
College of Medicine,
Hallym University
896 Pyeongchon-dong
Dongan-gu
Anyang-si
Gyeonggi-do 431-070
Korea
E-mail: dermakkh@yahoo.co.kr
the intraepidermal duct; CK5 stains the outer cells of the dermal duct and lower intraepidermal duct but does not stain the intraepidermal duct located in the upper epidermis.
Results We were able to confirm that the solid strands stained for EMA and CK5, as did
the outer cells of the ductal structure. However, the solid strands did not stain with CEA.
Conclusions The results indicated that solid strands observed in syringomas originate
from the outer cells of the two layers of cells that compose the lower epidermal duct or the
transitional portion between the intraepidermal duct and dermal duct in the normal eccrine
or apocrine structure. Thus, we surmise that a syringoma is developed by the proliferation
of these cells.
Conflicts of interest: None.
Introduction
Most epithelial tumors react with antibodies against the
epithelial membrane antigen (EMA), including squamous
cell carcinoma, breast carcinoma, and large cell lung carcinoma. EMA will also stain normal sweat and sebaceous
glands, although the epidermis is nonreactive with this
antibody. Carcinoembryonic antigen (CEA) has been
found in normal eccrine and apocrine cells, in benign
sweat gland tumors, and in mammary and extramammary Paget’s disease of the skin.1 Whereas the secretory
portion of the normal sweat gland becomes clearly immunolabeled with both the anti-EMA and anti-CEA antibody, in the ductal portion, EMA is found in two of the
ª 2011 The International Society of Dermatology
three layers of outer cells of the intraepidermal duct and
the upper portion of the intradermal duct, while CEA is
found on the luminal surface of the ductal structure.2,3
Cytokeratin (CK) is the protein of the keratin-containing
intermediate filament found in the intracytoplasmic cytoskeleton of epithelial cells, and at least 20 different subtypes exist.4 Among them, CK5 is found in the basal cells
of the sweat ridge (lower acrosyringium) and the outer
cells of the dermal duct but not in the acrosyringium
located in the upper epidermis.5
Syringoma is an adenoma derived from the sweat duct,
and it displays histopathological traits of multiple cystic
tubules in the dermis with solid strands that consist of
basophilic epithelial cells without tubules, scattered in the
817
International Journal of Dermatology 2012, 51, 817–822
资料来自互联网，仅供科研和教学使用，使用者请于24小时内自行删除
818
Report
An immunohistochemical study of the origin of the solid strand in syringoma, using CEA, EMA, and CK5
fibrous stroma. Although much research has been conducted on the origin of syringoma, the histogenesis and
differentiation of it remains controversial. The published
studies examined various antibodies, and our study is an
additional immunohistochemical work-up.
Recently, there have been an increasing number of
reports on clear cell syringoma in South Korea. Based on
our examination of clear cell syringoma’s structural
slides, we were able to confirm that the staining patterns
of EMA and CEA were different in solid strands. Therefore, we believed that we could identify the cause of the
increase in developments of other subtypes of syringoma
by examining the origin of the solid strand.
Kim et al.
the avidin–biotin complex method, using 3,3¢-diaminobenzidine
for chromogenic development. They were then counterstained
with hematoxylin and mounted in xylene. None of the specific
antibodies used can cross-react with other antigens.
The primary antibodies used in this study are shown in
Table 1.
Results
Light microscopic features of syringoma
A routine light microscopic examination revealed that
syringomas are composed of numerous tubular structures
embedded in a dense collagenous stroma. The walls of
Materials and methods
(a)
Thirty-one indisputable, histologically diagnosed cases of
syringoma were retrieved from the files of the Department of
Dermatology, College of Medicine, Hallym University, Republic
of Korea. All cases presented with histological features of
syringoma.
Hematoxylin and eosin staining
The tissues were embedded in 5-lm-thick paraffin sections,
which were then dewaxed in xylene, rehydrated through grades
of alcohol (100, 90, 80, and 70%) to phosphate-buffered saline,
stained with hematoxylin and eosin, and mounted in resin.
Immunohistochemical staining
For the immunohistochemical studies, formalin-fixed, paraffinembedded tissue blocks were cut into 5-lm-thick sections,
deparaffinized in xylene, and rehydrated through grades of
alcohol to phosphate-buffered saline, then stained using a twostep immunostaining kit (DAB Universal kit; Ventana Medical
Systems, Tucsan, AZ, USA) as follows: Sections were treated
with 3% H2O2 to block endogenous peroxidase, then antigen
retrieval followed the requirements of the individual antigen. For
antibody CEA, EMA, and CK5, sections were heated to 95 C
in a citrate buffer (pH 6.0) for 15 minutes and slowly cooled to
room temperature. After incubation with primary antibodies,
followed by secondary antibodies, the sections were stained by
(b)
Table 1 Antibodies used in this study
Antibody
A0115
E29
EP1601Y
Antigen
recognized
Antibody
source
Carcinoembryonic
antigen
Epithelial membrane
antigen
Cytokeratin 5
Dako, Carpinteria,
CA, USA
NeoMarkers, Fremont,
CA, USA
NeoMarkers, Fremont,
CA, USA
International Journal of Dermatology 2012, 51, 817–822
Dilution
1 : 700
1 : 200
1 : 40
Figure 1 Light microscopic features of syringoma. (a) Multiple cystic structures, ducts, and epithelial strands embedded
in the fibrous stroma. Note the basal hyperpigmentation and
a keratin-filled cyst (hematoxylin and eosin, ·40). (b) The
ducts were lined by two rows of epithelial cells. Small
comma-like configurations were present, giving a tadpole
appearance (hematoxylin and eosin, ·200)
ª 2011 The International Society of Dermatology
资料来自互联网，仅供科研和教学使用，使用者请于24小时内自行删除
Kim et al.
An immunohistochemical study of the origin of the solid strand in syringoma, using CEA, EMA, and CK5
Report
819
the ducts are predominantly lined by two rows of epithelial cells. In most instances, these cells are flat. Some
ducts are lined with an eosinophilic cuticle, whereas others have small, comma-like tails of epithelial cells, giving
them the appearance of tadpoles. In addition, there are
solid strands of basophilic epithelial cells, which are independent of the ducts (Fig. 1).
Immunohistochemical staining of syringoma
Carcinoembryonic and epithelial membrane antigens
In all cases of lesional skin, CEA is expressed in the inner
ductal cells as well as in secretions within the lumina, but
negative staining was noted with the antibodies against
solid strands (Fig. 2). CEA positive findings were
observed only in two cases (6.5%) of the outmost ductal
cells. EMA was detected with variable frequency in the
peripheral cells of both ductal and solid structures
(Fig. 3). Among the inner ductal cells, only three cases
(9.7%) were EMA positive, while in 29 cases (93.5%)
the solid strands were EMA positive.
Figure 3 Distribution of epithelial membrane antigen (EMA)
in syringomas. Positive staining for EMA is visible in
peripheral cells and epithelial strands. The luminal cells of
the duct did not stain positive for EMA (·400)
Cytokeratin 5
Uniform staining for CK5 was observed in all cell types
present in the syringomas studied. Strong positive staining
for CK5 was visible in the outmost ductal cells and solid
strands (Fig. 4).
Table 2 shows immunohistochemical staining patterns
of syringomas sought in each case. The results are summarized in Table 3.
Figure 4 Distribution of cytokeratin (CK)5 in syringomas.
Antibody against CK5: the outer cells and epithelial strands
are all strong positive. The luminal cells of the duct did not
stain positive for CK5 (·400)
Discussion
Figure 2 Distribution of carcinoembryonic antigen (CEA) in
syringomas. Strong positive staining for CEA is visible in the
inner cells as well as in secretions within the lumina. The
outer cells of the ducts and epithelial strands did not stain
positive for CEA (·400)
ª 2011 The International Society of Dermatology
The eccrine sweat glands are major cutaneous appendages. Their principal function is thermoregulation during
exposure to a hot environment or during physical exercise.6 The eccrine sweat gland can basically be divided
into two main tissue segments: a proximal, secretory
globule in the lower reticular dermis, and a distal, excretory duct extending through the papillary dermis and
opening at the surface of the epidermis. The duct itself
International Journal of Dermatology 2012, 51, 817–822
资料来自互联网，仅供科研和教学使用，使用者请于24小时内自行删除
820
Report
An immunohistochemical study of the origin of the solid strand in syringoma, using CEA, EMA, and CK5
Kim et al.
Table 2 Immunohistochemical staining patterns of the cases studied
Inner ductal cells
Outmost ductal cells
Epithelial strands
Case
CEA
EMA
CK5
CEA
EMA
CK5
CEA
EMA
CK5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
++
++
+
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
)
)
)
)
(+)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
++
)
+
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
±
)
)
)
)
+
)
+
+
+
+
++
+
++
+
+
++
+
+
++
+
+
+
+
(+)
+
±
+
±
–
±
++
+
+
+
+
+
+
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
+
+
–
++
++
+
++
++
(+)
+
+
(+)
++
++
+
+
++
(+)
±
+
+
±
–
±
++
+
++
++
±
+
+
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
CEA, carcinoembryonic antigen; CK5, cytokeratin 5; EMA, epithelial membrane antigen.
Staining intensity: –, negative; ±, slight; +, moderate; ++, strong. ( ), focal staining.
Table 3 Immunohistochemical staining in syringomas
Inner ductal cells
Outmost ductal cells
Epithelial strands
Carcino
embryonic
antigen
n (%)
Epithelial
membrane
antigen
n (%)
Cytokeratin 5
n (%)
31 (100)
2 (6.5)
0 (0)
3 (9.7)
30 (96.8)
29 (93.5)
0 (0)
31 (100)
31 (100)
can be divided into four segments: proximal intraglandular duct; straight intradermal duct; lower acrosyringium
(sweat duct ridge); and upper acrosyringium. Apocrine
glands, such as eccrine glands, are composed of two segments (i.e., secretory portion and the excretory duct).
Because apocrine glands originate from the hair germ, or
International Journal of Dermatology 2012, 51, 817–822
primary epithelial germ, the duct of an apocrine gland
usually leads to pilosebaceous follicle, entering it in the
infundibulum, above the entrance of the sebaceous duct.
An occasional apocrine duct, however, opens directly on
to the skin surface close to a pilosebaceous follicle. The
duct portion of the apocrine gland has the same histologic
appearance as the eccrine duct, showing a double layer of
basophilic cells and a periluminal eosinophilic cuticle.1
A syringoma is a common adnexal tumor derived from
sweat glands. It was considered to be a tumor derived
from acrosyringeal eccrine ducts.7 Recently, some authors
have suggested that it might represent differentiation in
the lower acrosyringium or in the transitional portion
between the acrosyringium and dermal duct.5,8 However,
syringoma virtually never develops at sites replete with
eccrine elements, such as the palms and soles. Acral
syringomas are such a rarity that the observation of one
ª 2011 The International Society of Dermatology
资料来自互联网，仅供科研和教学使用，使用者请于24小时内自行删除
Kim et al.
An immunohistochemical study of the origin of the solid strand in syringoma, using CEA, EMA, and CK5
forms the basis of a case report. Instead, syringomas are
found almost exclusively on the face and genitalia, sites
at which apocrine elements are occasionally identified.
This distribution lends no support to the conclusion that
syringoma is of eccrine lineage.9 Furthermore, Yamamoto
et al.10 examined the intradermal ductal, transitional, and
secretory portions of the eccrine and apocrine glands, and
obtained almost similar results in both glands. We
decided to conduct this study to compare the developmental behaviors of various solid strands observed in
syringoma through immunohistochemical staining in
order to examine their origins and help explain the mechanism behind the development of syringoma subtypes.
Carcinoembryonic antigen, a glycoprotein containing
approximately 50% carbohydrate with an approximate
molecular weight of 200 kDa, was originally found in
fetal tissues and in certain gastrointestinal tumors as an
oncofetal antigen.11–13 Later, it was found in various normal adult tissues, including sweat glands.14 Li et al.13
found that the CEA antibody was unreactive, except in
the sweat glands, which made it a specific immunological
marker for them. CEA is present on the secretory lumen
of secretory cells and on the luminal surface of ductal
structures. A close similarity has been reported for the
inner ductal cells of syringomas and the luminal cells of
dermal eccrine ducts, both of which are characterized by
the presence of CEA.2,3,15–17 In this study, CEA displayed
positive findings in the luminal cells of the ductal structure of syringomas and in the amorphous contents of the
lumen. These were identical to the results of previous
studies. However, as CEA did not stain solid strands at
all, we were not able to confirm the association with the
luminal cells of the eccrine duct.
Antisera raised against defatted human milk fat globule
membranes have been shown to react with various normal epithelial tissues and carcinomas.18 The glycoproteins, with which these antisera react, have been termed
EMA. Ceriani et al.19 reported that an anti-EMA antibody reacted with cell membranes of normal breast and
breast carcinoma cells; however, that antibody did not
react with other carcinoma cells, lymphoma cells, melanoma cells, or mesenchymal cells. They found that the
EMA antibody reacted in hair follicles, sebaceous glands,
and apocrine sweat glands, besides the eccrine sweat
glands of the skin, which was different from the reactivity
of CEA antibody in the skin. There have been a number
of papers written on the expression of EMA in the eccrine
sweat gland, with varying degrees of EMA expression
reported. Although several papers report EMA monoclonal antibody staining the luminal membrane of the
eccrine sweat glands, there has been no clear indication
as to whether staining occurred in the secretory portion,
the ductal portion, or both. Ohnishi and Watanabe5
ª 2011 The International Society of Dermatology
Report
821
reported specifically that EMA was clearly found in the
acrosyringium of normal eccrine glands, the luminal cells
of the dermal duct, and the luminal surface of conventional syringoma and clear cell syringoma. On the other
hand, some papers claimed that the granular membraneassociated reaction of EMA was detected on the outer
cells of both the acrosyringium and upper portion of
intradermal eccrine ducts, as well as on the luminal surfaces and intercellular canaliculi of eccrine glands.2,3,20
Although we did not perform experiments regarding
EMA expression in the normal sweat glands for this
study, from a total of 31 cases, 30 (96.8%) were stained
in the outer cells of the ductal structure of the syringomas, and 29 (93.5%) in the solid strands, whereas only
two (6.5%) were stained in the inner cells of the ductal
structure. This leads us to believe that the solid strands
originate from the intraepidermal duct and from the outer
cells of the upper portion of the intradermal sweat duct.
The major structural protein of the epidermis is the
CK, a family of water-insoluble polypeptides, which form
the major intermediate filament network of epithelial
cells.13 They comprise a family of 20 different polypeptides distributed in a more or less tissue-specific fashion.4
A number of investigators have performed immunohistochemical analysis on various differentiation markers,
including the CK polypeptides present in normal adnexal
structures of the skin and their neoplasms. They found
that CKs were expressed in normal eccrine glands,
although the CKs showed different distributions between
the secretory and ductal portions.21 A recent study conducted on immunohistochemical differentiation, using
CK7, CD34, CK6, CK10, smooth muscle action, and
CD10, to better understand the derivation and differentiation of benign eccrine tumors, showed that, in normal
eccrine structures, CK6 is present in the inner ductal cells
while CK10 is present in the middle ductal cells. In
syringomas CK6 is specifically expressed in the inner
tumor cells and CK10 in the middle tumor cells, as is the
case with normal eccrine ducts.2,22,23 The outer cells of the
dermal eccrine ducts were found to be particularly characterized by strong staining for CK5/6.2,8 Among these cells,
CK5 was observed in the basal cells of the sweat duct
ridge (lower acrosyringium) and in the outer cells of the
dermal eccrine duct but not in the acrosyringium located
in the upper epidermis.5 Furthermore, because of positive
staining with RCK102 (against CK5, 8) and negative staining with 35bH11 (against CK8) in the outer cells of syringoma, CK5 was thought to be expressed in the outer layer
of the nests and strands in syringoma.5 Likewise, in all 31
cases of this study, CK5 was clearly stained in the outer
cells and the epithelial strands of the ductal structure of
syringoma but not in the luminal cells. Therefore, we
speculate that the syringomas differentiated toward the
International Journal of Dermatology 2012, 51, 817–822
资料来自互联网，仅供科研和教学使用，使用者请于24小时内自行删除
822
Report
An immunohistochemical study of the origin of the solid strand in syringoma, using CEA, EMA, and CK5
transitional portion between the intraepidermal duct and
the dermal duct, rather than in the intraepidermal duct in
the upper epidermis.
In conclusion, the solid strands of syringoma tested
positive for EMA and CK5 but negative for CEA. The
results of this study, therefore, suggest that the solid
strands in syringomas originate from the outer cells of the
two layers of cells that compose the lower intraepidermal
duct or the transitional portion. However, it has been
reported that the histologic appearance of the eccrine
duct is identical to that of the apocrine duct. Therefore,
we believe that a lineage-tracing experiment using transgenic mice is necessary in the future to prove that a syringoma develops in a specific cell that differentiates the
eccrine duct from the apocrine duct.
References
1 Elentisas R, Nousari CH, Ayli E, Seykora JT. Laboratory
methods. In: Elder DE, Elenitsas R, Johnson BL Jr, et al.,
eds. Lever’s Histopathology of the Skin, 10th edn.
Philadelphia: Lippincott Williams & Wilkins, 2009:
67–81.
2 Demirkesen C, Hoede N, Moll R. Epithelial markers and
differentiation in adnexal neoplasms of the skin: an
immunohistochemical study including individual
cytokeratins. J Cutan Pathol 1995; 22: 518–535.
3 Park ES, Jung JB, Kwon KY. A case of eccrine poroma
and distribution of epithelial membrane antigen. Ann
Dermatol 1995; 7: 192–196.
4 Moll R, Franke WW, Schiller DL, et al. The catalog of
human cytokeratins: patterns of expression in normal
epithelia, tumors and cultured cells. Cell 1982; 31:
11–24.
5 Ohnishi T, Watanabe S. Immunohistochemical analysis of
keratin expression in clear cell syringoma. J Cutan Pathol
1997; 24: 370–376.
6 Shibasaki M, Wilson TE, Crandall CG. Neural control
and mechanisms of eccrine sweating during heat stress
and exercise. J Appl Physiol 2006; 100: 1692–1701.
7 Hashimoto K, Gross BG, Lever WF. Syringoma:
histochemical and electron microscopic studies. J Invest
Dermatol 1966; 46: 150.
8 Eckert F, Nilles M, Schmid U, Altmannsberger M.
Distribution of cytokeratin polypeptides in syringomas.
An immunohistochemical study on paraffin-embedded
material. Am J Dermatopathol 1992; 14: 115–121.
9 McCalmont TH. A call for logic in the classification of
adnexal neoplasms. Am J Dermatopathol 1996; 18:
103–109.
International Journal of Dermatology 2012, 51, 817–822
Kim et al.
10 Yamamoto O, Doi Y, Hamada T, et al. An
immunohistochemical and ultrastructural study of
syringocystadenoma papilliferum. Br J Dermatol 2002;
147: 936–945.
11 Saga K. Histochemical and immunohistochemical
markers for human eccrine and apocrine sweat glands: an
aid for histopathologic differentiation of sweat gland
tumors. J Investing Dermatol Symp Proc 2001; 6: 49–53.
12 Saga K. Structure and function of human sweat glands
studied with histochemistry and cytochemistry. Prog
Histochem Cytochem 2002; 37: 323–386.
13 Li HH, Zhou G, Fu XB, Zhang L. Antigen expression of
human eccrine sweat glands. J Cutan Pathol 2009; 36:
318–324.
14 Metze D, Luger TA. Ultrastructural localization of
carcinoembryonic antigen (CEA) glycoproteins and
epithelial membrane antigen (EMA) in normal and
neoplastic sweat glands. J Cutan Pathol 1996; 23:
518–529.
15 Penneys NS, Nadji M, Morales A. Carcinoembryonic
antigen in benign sweat gland tumors. Arch Dermatol
1982; 118: 225–227.
16 Penneys NS. Immunohistochemistry of adnexal
neoplasms. J Cutan Pathol 1984; 276: 357–364.
17 Nap M, Mollgard K, Burtin P, Fleuren GJ.
Immunohistochemistry of carcino-embryonic antigen in
the embryo, fetus and adult. Tumour Biol 1988; 9:
145–153.
18 Heyderman E, Steele K, Ormerod MG. A new antigen on
the epithelial membrane: its immunoperoxidase
localisation in normal and neoplastic tissue. J Clin Pathol
1979; 32: 35–39.
19 Ceriani RL, Thompson K, Peterson JA, Abraham S.
Surface differentiation antigens of human mammary
epithelial cells carried on the human milk fat globule.
Proc Natl Acad Sci U S A 1977; 74: 582–586.
20 Takanashi M, Urabe A, Nakayama J, Hori Y.
Distribution of epithelial membrane antigen in eccrine
poroma. Dermatologica 1991; 183: 187–190.
21 Watanabe S, Ichikawa E, Takanashi S, Takahashi H.
Immunohistochemical localization of cytokeratins in
normal eccrine glands, with monoclonal antibodies in
routinely processed, formalin-fixed, paraffin-embedded
sections. J Am Acad Dermatol 1993; 28: 203–212.
22 Missall TA, Burkemper NM, Jensen SL, Hurley MY.
Immunohistochemical differentiation of four benign
eccrine tumors. J Cutan Pathol 2009; 36: 190–196.
23 Obaidat NA, Alsaad KO, Ghazarian D. Skin adnexal
neoplasms – part 2: an approach to tumours of
cutaneous sweat glands. J Clin Pathol 2007; 60: 145–
159.
ª 2011 The International Society of Dermatology
资料来自互联网，仅供科研和教学使用，使用者请于24小时内自行删除