ARTICLE
Basal cell carcinomas (BCC) are locally invasive tumors of slow progression
and are seldom metastatic [1]. Because of these characteristics, doubts
have been raised regarding their malignancy. Occasional cases of highly
invasive [2], sometimes metastatic [3] forms are nevertheless observed.
Some histological subtypes, e.g. infiltrative aggressive-type,
seem to have a higher potential for recurrence. It could be assumed that
the proliferative potential of some tumors is inversely related to their
degree of cellular differentiation.
As a basal keratinocyte differentiates along its journey to the skin
surface, transcription of cytokeratins K5 and K14 is down-regulated whereas
new sets of differentiation-specific genes are up-regulated, including
cytokeratins K1 and K10, profilaggrin, cornified cell envelope components
and new desmosomal proteins.
To study the differentiation stage of tumoral epithelial cells, many
studies demonstrating the usefulness of anti-cytokeratin [4-18] and anti-desmosomal
protein [19] antibodies have been reported. However, no antibody has been
found to specifically label BCC. Two-dimensional gel analysis of tumor
cell proteins [6] confirmed the absence of any specific tumoral cytokeratin
in these carcinomas. Moreover, numerous antibodies that label all BCC
were also found to stain squamous cell carcinomas [9-19].
The aim of this study was to try to differentiate groups of BCC by their
labeling with new monoclonal antibodies (MoAbs) directed against epidermal
differentiation markers. Among the large number of markers described in
the last few years, we chose cytokeratins K1 and K10, and corneodesmosin.
The study was performed on a large collection of 66 BCC biopsies obtained
from patients whose subsequent evolution was recorded. All samples were
reexamined in order to determine their histological type. The histological
and immunohistochemical results are discussed in relation to the clinical
evolution of the tumors.
Materials and methods
Tumors
Sixty-six BCC biopsies obtained from well characterized patients (41
males, 25 females, mean age 69 years, range: 43-91 years, at the time
of diagnosis) before any therapeutic regime was instigated, and stored
in the collection of the Cancer Center of Nancy (France) were used in
this study. Twelve were from clinically aggressive BCC that had relapsed
once or more in spite of extensive surgery. Fifty-four were from BCC that
had been successfully treated by surgery, radiotherapy, electrocoagulation
or local chemotherapy, without any relapse for at least 5 years. Samples
of skin from a fetus of 3 months were used as controls.
Histology
All samples were immersed in Bouin's fixative upon removal, and embedded
in paraffin. Fresh 4-µm sections were prepared and stained with hematoxillin-eosin.
The classification criteria proposed by Kirkham [20] were used to describe
tumoral architecture.
Immunohistochemistry
Antibodies. The three MoAbs used in this study and described
below are part of a series of antibodies directed against epidermal differentiation
antigens, produced and characterized in our laboratory, after immunization
of mice with an homogenate of human plantar stratum corneum [21-27]. Analyzed
by indirect immunofluorescence (IIF), F12-19 (IgG1-kappa) labels all the
human epidermal cell layers as well as sebaceous and sudoral gland cells
and all the cells of the hair follicle outer root sheath. All human cytokeratins
are stained by this MoAb on Western blotting, indicating that it is specific
for an epitope shared by all cytokeratins [21]. EE21-06 (IgG1-kappa) is
specific for human squamous cornified epithelia: epidermis, gingiva and
hard palate. It labels all the suprabasal epidermal cell layers in inter-
and intra-follicular epidermis but not cutaneous appendages. The epitope
recognized is shared by cytokeratins K1, K2, K9, and K10-11, as identified
by Western blotting [22]. G36-19 (IgG1-kappa) specifically labels the
stratum granulosum and the lower stratum corneum of cornified squamous
epithelium and the hair follicle inner root sheath. This antibody stains
corneodesmosin, a 52-56 kDa basic glycoprotein synthesized in the stratum
granulosum, and transported via lamellar bodies to the extracellular
space where it is detected in the desmosome core. Then the protein is
located on the dense plug of corneodesmosomes, the modified desmosomes
specifically observed in the stratum corneum of cornified epithelia [23-25].
Hyperkeratotic lesions, which are characterized by an increased number
of corneodesmosomes, also show an increased expression of corneodesmosin
[26]. Moreover, proteolysis of the protein is thought to be a key event
in desquamation [24, 27].
Indirect immunofluorescence. Fresh, 4-µm-thick sections
were prepared for each sample and deparaffinized by immersion for 10 min
in toluene at 37° C. They were then sequentially hydrated in graded
ethanol baths, washed in phosphate-buffered saline (PBS), and incubated
with the appropriate MoAb diluted in PBS, for 30 min at 37° C. Upon
incubation, the slides were washed, first in PBS supplemented with 0.5%
Tween 20, then in PBS. The secondary antibody (fluorescein isothiocyanate-labeled
rabbit antibodies to mouse IgG (H+L), Zymed, San Francisco, CA) was then
added, and the slides were further incubated for 15 min at 37° C.
The same series of washes was performed, and the sections were then examined
under UV light using a Leitz microscope equipped with Ploem epi-illumination.
Controls included incubations with the fluorescent antibody alone, and
staining of 4 normal skin samples.
Results
Histology
Out of the 54 non-relapsing tumors, the classification proposed by Kirkham
[20] allowed us to distinguish 22 BCC of the circumscribed type (nodular
BCC), 11 adenoid BCC, 4 keratotic BCC, 9 superficial BCC, 4 Pinkus fibroepithelial
tumors, and 4 solid infiltrative aggressive-type BCC which appeared as
thin, non-anastomotic trabeculae, deprived of peripheral palissadic cells,
and invading the dermis. The latter aspect, characteristic of "an aggressive
growth pattern" of BCC, as described by Jacobs et al. [28], and
also reported by Sexton et al. [29], was also observed in all the
12 relapsing tumor samples of our series. Among the nodular BCC, large
nodules were always observed. In only 3 cases, were they associated with
micro-nodules.
Immunohistochemistry
To evaluate the differentiation stage of the epithelial cells in the
various histological types of BCC, IIF analysis was then performed with
MoAbs directed against differentiation-related antigens. Results are summarized
in Table I.
F12-19, an MoAb directed against an epitope shared by all the cytokeratins,
homogeneously labeled the center of nodular BCC, whereas there was no
labeling of the palissadic ring (Fig.
1). F12-19 also labeled all the tumor masses of keratotic BCC,
with a reinforcement of the fluorescence intensity on the concentric keratinized
areas, and brightly stained all the trabeculae of the adenoid BCC. Cell-to-cell
heterogeneity was noted in the labeling intensity of infiltrative aggressive-type
BCC (data not shown). In contrast, superficial BCC, and basocellular buds
sprouting from the trabeculae of Pinkus fibroepithelial tumors were not
labeled. Lastly, F12-19 did not react on embryonic primary epithelial
germs (data not shown).
EE21-06, an MoAb specific for the terminal differentiation-related cytokeratins
K1, K2, K9 and K10-11, labeled most nodular, adenoid (Fig.
2) and keratotic BCC, with patterns similar to that reported above
for F12-19. Infiltrative aggressive-type BCC were inconsistently labeled,
with bright heterogeneous reactive patches within a given tumor (Fig.
3). Neither superficial BCC, basocellular buds of Pinkus fibroepithelial
tumors, nor primary epithelial germs, were labeled by this MoAb.
G36-19, an MoAb specific for the cornification-related protein corneodesmosin,
labeled the concentric keratinized areas and/or hair follicle-like structures
of keratotic BCC and some individual cells in the center of nodular BCC
(Fig. 4). Superficial
BCC, basocellular trabeculae of adenoid BCC and of infiltrative aggressive-type
BCC, and primary epithelial germs, were not labeled by G36-19.
Discussion
In this paper, the histological and immunohistological characteristics
of BCC were investigated in a series of 66 samples from patients whose
response to BCC-specific therapy had been recorded for at least 5 years.
A retrospective study of their histological features showed that all clinically
aggressive tumors relapsing after surgical treatment, were of the infiltrative
aggressive-type, as previously shown in other series [5, 28-32]. All these
BCC displayed similar characteristics: thin, non-anastomotic trabeculae
without palissadic cells, without peripheral palleal dermis and directly
infiltrating the reticularis dermis. This homogeneity could be due to
the selection criteria we used, requiring both examination of the initial
biopsy and relapse after extensive surgery.
One of the major concerns with immunohistochemical analysis performed
with MoAbs is whether a negative result reflects masking of the epitope
or a real absence of the antigen. In particular, masking of cytokeratin
epitopes has already been described. This may explain why F12-19, an MoAb
directed to all cytokeratins including K5 and K14, did not label either
superficial BCC, some parts of the tumor masses or embryonic epithelial
germs. This could be related to an immature state of the tumor cells,
characterized by a specific organization of intermediate filaments masking
the F12-19 epitope. Indeed, superficial BCC as well as embryonic epithelial
germs were labeled by N40-13, an MoAb likely directed against a conformational
epitope of cytokeratins K5/K14 (unpublished observation of the authors).
Our immunohistological analysis, as well as data already published [28],
supports the conclusion of previous morphological studies in discriminating
three major groups of BCC: (1) superficial, (2) nodular, including keratotic
and adenoid variants, and (3) infiltrative aggressive-type BCC. Previous
studies of the differentiation stage of BCC showed that epithelial BCC
cells are labeled by antibodies directed to epidermal basal cell cytokeratins
[11-13, 15, 17, 18] and suggested that most BCC are undifferentiated,
except keratotic BCC [9, 15, 16]. In our study, the use of EE21-06 and
G36-19 (MoAbs specific for the differentiation-related cytokeratins K1,
K2, K9 and K10-11, and for the cornification-related antigen corneodesmosin,
respectively) allowed various degrees of BCC cell differentiation to be
more precisely determined. Superficial, undifferentiated BCC appear somewhat
similar in their antigenic expression to embryonic primary epithelial
germs, as previously reported by Van Cauwenberge et al. [33]. EE21-06
strongly labeled the keratinized areas and/or the hair follicle-like structures
of keratotic BCC, the center of the tumor masses in nodular BCC and trabeculae
in Pinkus fibroepithelial tumors. In a recent study however [34], expression
of cytokeratin K10 has not been observed in nodular (primary or recurrent)
BCC. Therefore, these BCC could rather only express cytokeratin K1 or
K9. More likely, this could be due to differences in specificity or sensitivity
of the MoAbs used or to tumor heterogeneity. The keratinized areas and/or
the hair follicle-like structures of keratotic BCC, a few patches of cells
within nodular BCC, and cells in the center of trabeculae in Pinkus fibroepithelial
tumors were labeled by G36-19, displaying characteristics of granular
keratinocytes. The maturation of nodular and keratotic BCC thus appears
centripetic, from peripheral undifferentiated cells to central granular
cells. Use of other MoAbs would be necessary to test whether these cells
cornify or not. Conversely, adenoid BCC cells appeared to be homogeneously
differentiated, since all trabeculae were brightly labeled by EE21-06
but were not stained by G36-19 and did not show signs of granular maturation.
Infiltrative aggressive-type BCC are characterized by the heterogeneous
cell-to-cell labeling produced by F12-19 anf EE 21-06. A wide range of
monospecific antibodies should be used to precisely determine the pattern
of cytokeratin expression in these BCC. Such IIF labeling heterogeneity
has previously been reported [18, 34, 35], and is also displayed by squamous
cell carcinomas [15, 17]. This similarity with squamous cell carcinomas
could bear some relationship to the greater aggressiveness of these BCC.
CONCLUSION
This retrospective histological and immunohistological study performed
with a new series of MoAbs, confirms and completes previous reports on
BCC classification [28] and suggests the discrimination of three major
types of BCC: superficial, nodular (including adenoid and keratotic) and
infiltrative aggressive-type. In particular, some nodular BCC cells are
differentiated and even show characteristics of granular keratinocytes.
Infiltrative aggressive-type BCC present cell-to-cell heterogeneity, similar
to that observed in squamous cell carcinomas. This feature, that is not
observed in other types of BCC, correlates with greater malignancy and
relapse risk. We suggest that an immunohistochemical analysis with antibodies
specific for different stages of keratinocyte differentiation may help
to indicate surgical treatment of infiltrative aggressive-type BCC.
Acknowledgements
The research was supported by grants from the "Association pour la Recherche
contre le Cancer" and from the "Fédération Nationale des
Centres de Lutte contre le Cancer".
REFERENCES
1. Ten Seldan R, Helwig R. Types histologiques des tumeurs cutanées.
In: Ten Seldan R, Helwig R, eds. Classification histologique internationale
des tumeurs. Genève: WHO Inc., 1975: 49-50.
2. Ko CB, Walton S, Keczes K. Extensive and fatal basal cell carcinoma:
a report of three cases. Br J Dermatol 1992; 127: 164-7.
3. Mikhail GR, Nims LP, Kelly AP, Ditmars DM, Eyler WR. Metastatic basal
cell carcinoma. Arch Dermatol 1977; 113: 1261-9.
4. Moll R, Franke WW, Schiller DL, Geiger B, Krepler R. The catalog
of human cytokeratins: patterns of expression in normal epithelia, tumors
and cultured cells. Cell 1982; 31: 11-24.
5. Miller SJ. Biology of basal cell carcinoma (Part 1). J Am Acad
Dermatol 1991; 24: 1-13.
6. Moll R, Moll I, Franke WW. Differences of expression of cytokeratin
polypeptides in various epithelial skin tumors. Arch Dermatol Res
1984; 276: 349-63.
7. Kubilius J, Baden HP, MacGilvray N. Filamentous protein of basal
cell epithelioma: characteristics in vivo and in vitro.
J Natl Cancer Inst 1980; 65: 869-71.
8. Shimizu N, Ito M, Tazawa T, Katsuumi K, Sato Y. Anti-keratin monoclonal
antibody against basal cell epithelioma keratin: BKN-1. J Dermatol
1987; 14: 359-63.
9. Loning T, Staquet MJ, Thivolet J, Seifert G. Keratin polypeptides
distribution in normal and diseased human epidermis and oral mucosa. Virchows
Arch Path Anat and Histol 1980; 388: 273-88.
10. Viac J, Réano A, Thivolet J. Cytokeratins in human basal
and squamous cell carcinomas: biochemical, immunohistological findings
and comparisons with normal epithelia. J Cutan Pathol 1982; 9:
377-90.
11. Eichner P, Bonitz P, Sun TT. Classification of epidermal keratins
according to their immunoreactivity, isoelectric point, and mode of expression.
J Cell Biol 1984; 98: 1388-96.
12. Sun TT, Eichner P, Nelson WG, et al. Keratin classes: molecular
markers for different types of epithelial differentiation. J Invest
Dermatol 1983; 81: 109s-11s.
13. Hosokawa M, Masu SI. Immunofluorescence studies on epidermal keratinization
in some skin disorders with keratin antisera. J Exp Med 1983; 141:
163-73.
14. Debus E, Moll R, Franke WW, Weber K, Osborn M. Immunohistochemical
distinction of human carcinomas by cytokeratin typing with monoclonal
antibodies. Am J Pathol 1984; 114: 121-30.
15. Thomas P, Said JW, Nash G, Banks-Schlegel S. Profiles of keratin
proteins in basal and squamous cell carcinomas of the skin: an immunohistochemical
study. Lab Invest 1984; 50: 36-41.
16. Viac J, Staquet MJ, Chomel G, Thivolet J. Alteration in the expression
of two epidermal differentiation antigens in human epidermal disorders.
Path Res Pract 1985; 180: 577-83.
17. Robinson JK. Expression of keratin proteins in deeply invasive basal
and squamous cell carcinoma: an immunohistochemical study. J Dermatol
Surg Oncol 1987; 13: 283-94.
18. Stoler A, Kopan R, Duvic M, Fuchs E. Use of monospecific antisera
and cRNA probes to localize the major changes in keratin expression during
normal and abnormal epidermal differentiation. J Cell Biol 1988;
107: 427-46.
19. Moll R, Cowin P, Kapprell HP, Franke WW. Desmosomal proteins: new
markers for identification and classification of tumors. Lab Invest
1986; 54: 4-25.
20. Kirkham N. Tumors and cysts of the epidermis. In: Lever's Histopathology
of the skin, D. Elder et al. eds. Eighth edition. Philadelphia:
Lippincott-Raven Publishers, 1997: 685-746.
21. Serre G, Bettinger S, Kadouche J, Breitburd F, Soleilhavoup JP.
Monoclonal antibodies to keratins and other structural components of human
skin (abstr). Fourth International Conference on "Differentiation of normal
and neoplastic cells", Heidelberg, April 1985.
22. Serre G, Conter A, Ouhayoun JP, et al. EE 21-06. A mouse
monoclonal antibody to a sequential epitope common to the keratin polypeptides
specifically expressed in the human keratinized squamous epithelia (abstr).
In: Rousset BAF, ed. Structure and Functions of the Cytoskeleton. Biological
and Physiopathological Aspects. Colloque INSERM, Paris: John Libbey
Eurotext Ltd, 1988: 524.
23. Serre G, Mils V, Haftek M, et al. Identification of late
differentiation antigens of human cornified epithelia, expressed in re-organized
desmosomes and bound to cross-linked envelope. J Invest Dermatol
1991; 97: 1061-72.
24. Simon M, Montézin M, Guerrin M, Durieux J-J, Serre G. Characterization
and purification of human corneodesmosin, an epidermal basic glycoprotein
associated with corneocyte-specific modified desmosomes. J Biol Chem
1997; 272: 31770-6.
25. Haftek M, Serre G, Mils V, Thivolet J. Immunocytochemical evidence
for a possible role of cross-linked keratinocyte envelopes in stratum
corneum cohesion. J Histochem Cytochem 1991; 39: 1531-8.
26. Haftek M, Simon M, Kanitakis J, Marechal S, Claudy A, Serre G, Schmitt
D. Expression of corneodesmosin in the granular layer and stratum corneum
of normal and diseased epidermis. Br J Dermatol 1997; 137: 846-73.
27. Lundström A, Serre G, Haftek M, Egelrud T. Evidence for a role
of corneodesmosin, a protein which may serve to modify desmosomes during
cornification, in stratum corneum cell cohesion and desquamation. Arch
Dermatol Res 1994; 286: 369-75.
28. Jacobs GH, Rippey JJ, Altini M, Dent M. Prediction of aggressive
behavior in basal cell carcinoma. Cancer 1982; 49: 533-7.
29. Sexton M, Jones DB, Maloney ME. Histologic pattern analysis of basal
cell carcinoma. J Am Acad Dermatol 1990; 23: 1118-26.
30. Sloane JP. The value of typing basal cell carcinomas in predicting
recurrence after surgical excision. Br J Dermatol 1977; 96: 127-32.
31. Lang PG, Maize JC. Histologic evaluation of recurrent basal cell
carcinoma and treatment implications. J Am Acad Dermatol 1986;
14: 186-96.
32. Dixon AY, Lee SH, MacGregor DH. Factors predictive of recurrence
of basal cell carcinoma. Am J Dermatopathol 1989; 11: 222-32.
33. Van Cauwenberge D, Piérard GE, Foidart JM, Lapière
CM. Immunohistochemical localization of laminin, type IV and type V collagen
in basal cell carcinoma. Br J Dermatol 1983; 108: 163-70.
34. Krekels GAM, Verhaegh MEJM, Wagenaar SS, Ramaekers FCS, Neumann
HAM. Keratins (K8 and K19) as potential markers of recurrent basal cell
carcinoma. Eur J Dermatol 1997; 7: 158-60.
35. Habets JMW, Tank B, Vuzevski VD, Brevé J, Stolz E, Van Joost
Th. Absence of cytokeratin 8 and inconsistent expression of cytokeratin
7 and 19 in human basal cell carcinoma. Anticancer Res 1988; 8:
611-6.
|