Gene Ther Mol Biol Vol 10, 179-184,
2006
Title-loss of bcatenin is an independent prognostic factor in ovarian carcinomas: A
multivariate analysis
Research Article
Cristina Faleiro-Rodrigues1,*, Isabel
Macedo-Pinto1, Deolinda Pereira2
1Department of Anatomy and Pathology
2Department of Medical Oncology,
Portuguese Institute of Oncology of Francisco Gentil, Centro Regional do Norte,
Porto, Portugal
__________________________________________________________________________________
*Correspondence: Faleiro-Rodrigues C.,
Instituto Portugus de Oncologia Francisco Gentil, Centro Regional do Norte,
Departamento de Anatomia Patolgica, Rua Dr. Antnio Bernardino de
Almeida4200-072 Porto, Portugal; Telephone: +351-22-5084000 Ext 1002; Fax
+351-22-5084001; e-mail: cristinafaleiro@mail.com
Key words: ovarian cancer, cell
adhesion, epithelial cadherin, b-catenin,
immunohistochemistry
Abbreviations: avidin-biotin peroxidase,
(ABC); E-cadherin catenin unit, (ECCU); Epithelial cadherin, (E-cadherin);
International Federation of Gynaecology and Obstetrics system, (FIGO); Overall
survival, (OS); World Health Organization, (WHO)
Received: 10 March 2006; Revised: 26 April 2006
Accepted: 16 May 2006; electronically published: July 2006
Summary
In ovarian carcinomas, numerous studies
have shown consistent prognostic significance of FIGO stage and residual tumour
as independent prognostic factors. However, these prognostic factors alone
cannot accurately predict disease outcome since a considerable degree of
heterogeneity remains within the various subgroups limiting the predictive
value of these factors. Therefore, the identification of new molecular markers
that may possibly distinguish patients at a higher risk is of great importance.
In two previous studies, the individual loss of E-cadherin and the individual
loss of b-catenin
were important prognostic factors of poorer overall survival in patients with
ovarian carcinomas. Purpose of the present study was to re-analyse the
immunohistochemical expression of E-cadherin and b-catenin in 104 patients with ovarian
carcinomas, and evaluate whether these two proteins continue to be important
independent prognostic factors when assessed together in a multivariate Cox«s
proportional hazard regression analysis. Results In the multivariate analysis,
the most important independent prognostic factors of poorer overall survival
were loss of b-catenin
expression ([HR], 5.79, 95% CI, 2.38 to 14.10; P=0.0001), FIGO stage IV ([HR],
7.19, 95% CI, 1.02 to 50.8; P=0.04) and residual tumour ([HR], 6.78, 95% CI,
1.41 to 32.56; P=0.034). Conclusion The loss of b-catenin
expression is a stronger prognostic factor than E-cadherin. The findings in the
present study and previously reported data suggest that b-catenin is a significant prognostic
indicator in patients with epithelial ovarian cancer, however, these results
should be supported by more and larger studies.
I. Introduction
In ovarian carcinomas, numerous studies have shown
consistent prognostic significance of FIGO tumour stage and size of residual
tumour as independent prognostic factors (Rubin et al, 2003). However, these
prognostic factors alone cannot accurately predict disease outcome since a
considerable degree of heterogeneity remains within the various subgroups
limiting the predictive value of these factors. Therefore, the identification
of new molecular markers that may possibly distinguish patients at a higher
risk is of great importance. Epithelial cadherin (E-cadherin) is a
calcium-dependent cell adhesion molecule which plays a key role in cell-cell
epithelial adhesion and epithelial tissue integrity. The intracellular domain
of E-cadherin is found in a complex linked with the catenins (a-and b-).
The association of catenins to cadherins is a key step in the function of
intact adhesion complexes. The catenins link the cadherin molecules to the
cytoskeleton and mediate signal transduction mechanisms that regulate cell
adhesion, growth and differentiation (Frixen et al, 1991; Tsukita et al, 1992;
Kemler, 1993; Hinck et al, 1994). Detachment of tumour cells from the primary
lesion is considered a main step in the process of invasion and metastases.
Increasing evidence points to a role for E-cadherin and the catenins in cancer
progression since the loss or reduced expression of E-cadherin and b-catenin correlates with invasive behaviour, increased
lymph node metastasis and poor outcome in patients with malignant melanoma and
gastric carcinomas (Jawhari et al, 1997; Ramesh et al, 1999; Kageshita et al,
2001; Tanaka et al, 2002). In two previous individual studies, a significant
correlation between poor overall survival and the loss of E-cadherin and the
loss of b-catenin was observed in patients with ovarian
carcinomas. The loss of E-cadherin and b-catenin immunoexpression was also shown to be independent predictors
of poorer survival in a multivariate analysis (Faleiro-Rodrigues et al 2004a,
b). This study represents a re-analyse of previously published data with the
purpose of determining whether these two proteins continue to be important
independent prognostic factors when assessed together in a multivariate Coxās
proportional hazard regression analysis.
II. Material and Methods
Routinely formalin-fixed and
paraffin-embedded tissue samples from 104 cases of primary ovarian carcinomas were
retrieved from the Department of Pathology at the Portuguese Institute of
Oncology of Francisco Gentil, Porto, from January 1995 to December 1999. The
mean age at the time of diagnosis was 56 years (range, 21 to 89 years). None of
these patients had undergone neoadjuvant chemotherapy prior to surgery. All
tissue specimens were reviewed and re-evaluated by an experienced
gynaecological pathologist. Histological classification was performed according
to the World Health Organization (WHO) standards. The grading and staging of
the tumours were assigned according to the International Federation of
Gynaecology and Obstetrics system (FIGO). The mean overall survival duration of
the patients was 35 months. At the end of the follow-up period, 65 (62%)
patients were without evidence of disease, 35 (34%) patients had died of
disease, and 4 (4%) patients were lost for follow up.
A. Tissue sections
All the tissue sections
(stained by haematoxylin and eosin) from each case were observed. Areas of
necrosis or deterioration of tissue morphology were avoided. The pathologist
selected the best tumour section representing well preserved tissue
architecture and cell morphology with approximately 2.0 x 1.0 cm.
B. Immunohistochemical staining
Archival tissue was fixed in
10% formalin and 3 µm sections were used for both histological and
immunohistochemical studies. Immunohistochemistry was performed in all cases
using the avidin-biotin peroxidase (ABC) complex with an additional step for
microwave antigen retrieval as described (Faleiro-Rodrigues et al, 2004b). The
following monoclonal antibodies were used: E-cadherin (C20820), and b-catenin (C19220)
(Transduction Laboratories, Lexington, UK). To ensure accurate and reproducible
staining, normal skin epithelium was used as a positive control.
Staining of E-cadherin and b-catenin was localized on the
cell membrane of epithelial cells, particularly at areas of cell-to-cell
contact. Normal skin epithelium without the primary antibody was used as a
negative control.
C. Evaluation of E-cadherin and b-catenin immunostaining
Membranous immunoreactivity
of the catenins was assessed by light microscopy by two independent observers,
without previous knowledge of the patients clinicopathological details.
E-cadherin and b-catenin immunoexpression in the tumours was
scored semi-quantitatively on a scale of 0 to 3 (0=complete absence of
expression, 1=10%, 2 >10 and = 50%, 3 >50%). For all the association
analyses, the subdivision into negative (0 scale) and positive expression (1-3
scale) was used.
D. Statistical analysis
The statistical software used
was the Statistical Package for the Social Sciences (SPSS version 8.0, SPSS,
Chicago). Clinical data was obtained from the Cancer Registry Records of the
patients and evaluated by a Medical Oncologist. Overall survival (OS) was
defined as the time from diagnosis to death or last clinical control date, and
used as a measure of prognosis. Univariate survival curves were estimated using
the Kaplan-Meier method and compared using the Log-rank or the Breslow test.
Multivariate analysis was performed using the Cox«s proportional hazards
regression model with overall survival as the outcome measure. Forward stepwise
procedure was used to select the independent variables in the multivariate
analysis. Forward selection, allows variables to be considered
one at a time for entry into the model. After a variable is added to the model,
all variables already in the model are examined for removal. The algorithm
stops when no more variables meet entry or removal criteria. A value of P<0.05 was
regarded as significant.
III. Results
The present series consisted of 104 carcinomas that
were classified into the following histological types, 56 serous carcinomas, 22
mucinous carcinomas, 16 clear cell carcinomas, 8 endometrioid and 2
transitional cell carcinomas. These carcinomas were graded into 26
well-differentiated, 27 moderately differentiated and 51 poorly differentiated
tumours. In this series, 31 cases were diagnosed with FIGO stage I tumours, 7
in FIGO stage II, 47 in FIGO stage III and 19 in FIGO stage IV. The
clinicopathological parameters studied were FIGO staging, histological type,
tumour differentiation, peritoneal metastasis, and residual tumour after
surgery, the appearance of the ovarian capsule, peritoneal cytology and
lymphatic/vascular invasion (previously described in Faleiro-Rodrigues et al
2004a).
A. Immunoreactivity of E-cadherin and b-catenin
in carcinoma tissue
Negative E-cadherin expression was observed in 7 (7%)
malignant tumours, and positive in 97 (93%). Negative b-catenin expression was observed in 15 (14%) malignant
tumours, and positive in 89 (86%).
B. Relationship between E-cadherin and b-catenin expression in carcinoma tissue
In
the 15 carcinomas demonstrating negative expression for b-catenin, 3 carcinomas showed negative expression for E-cadherin, and 12
carcinomas showed positive expression for E-cadherin (Table 1).
C. Relationship between the expression of
E-cadherin and b-catenin
and patient overall survival
In the univariate
survival analysis, patients whose carcinoma tissue demonstrated negative
E-cadherin expression had a statistically significant decreased 5-year overall
survival rate compared with patients showing positive expression (29% versus
66%, P=0.006), (Faleiro-Rodrigues et al, 2004a). Patients whose carcinoma
tissue demonstrated negative b-catenin expression had a statistically
significant decreased 5-year overall survival rate compared with patients
showing positive expression (44% versus 66%, P=0.022), (Faleiro-Rodrigues et
al, 2004b).
The parameters that had
a significant impact on overall survival as E-cadherin (P=0.006), b-catenin (P=0.022), FIGO
stage (P²0.0001), peritoneal metastasis (P²0.0001), and post-operative residual
tumour (P²0.0001), peritoneal cytology (P²0.0001) and lymphatic/vascular
invasion (P=0.008), were then reviewed by a multivariate analysis (Cox«s proportional
hazards regression model, Table 2).
Negative expression of b-catenin (P=0.0001); (Figure 1), FIGO stage IV (P=0.04) and residual postoperative
tumour (P=0.01) were shown to associate significantly with poor patient
prognosis.
IV. Discussion
The cause of epithelial ovarian carcinoma is unknown
and diagnosis is retarded by the lack of symptoms in early stage disease.
Consequently, the poor overall survival and morbidity associated with
epithelial ovarian cancer deaths results from the detection of the disease in
advanced tumour stages with widespread metastatic disease at the time of
diagnosis (Ozols et al, 2000). To date, the molecular mechanisms that allow
ovarian cancer cells to detach from the primary tumour and consequently
interact with the mesothelium are not fully characterized.
Cell adhesion molecules may play an important role in
epithelial ovarian carcinogenesis, since cell-to-cell adhesion plays a critical
role in a wide variety of biological processes including embryogenesis,
maintenance of cell polarity, cell growth, and cell differentiation (Skubitz,
2002). The loss of cell adhesion molecules may lead to changes in cellular
adhesion and to increased motility, processes that contribute to the invasive
and/or metastatic potential of cells (Vleiminckx et al, 1991; Birchmeier and
Behrens, 1993; Mareel et al, 1994; Van Aken et al, 2001).
E-cadherin has been identified as an important
transmembrane molecule involved in the adhesion of epithelial cells at adherens
junctions. Adherens junctions are organized around transmembrane proteins of
the cadherin family. While the extracellular domain of the E-cadherin molecule
interacts with that of an opposing E-cadherin on a neighbouring cell, the
intracellular cytoplasmic domain of E-cadherin associates with §-catenin, which
in turn complexes with a-catenin mediating the connection of E-cadherin to the
cytoskeleton (Tsukita et al, 1992). The complex of E-cadherin and the
catenins
is a functional unit, which is termed here as the E-cadherin-catenin unit
(ECCU). Binding to catenins is important for E-cadherin function, rendering the
catenins regulatory molecules of E-cadherin. Thus, alterations in E-cadherin or
the catenins may lead to loss of cell-cell adhesion, resulting in tumour
aggressiveness and invasiveness in neoplastic disease (Ozawa et al, 1990;
Frixen et al, 1991; Mareel et al, 1994).
In two previous
individual studies, the significance of E-cadherin and the catenins a, b- and g-, as predictors of
poorer survival in patients with ovarian carcinomas was assessed. The first
study showed that negative E-cadherin immunoexpression significantly predicted
a poorer overall survival, and was an independent prognostic factor in the
multivariate analyses (Faleiro-Rodrigues et al 2004a). In the second study, although negative
immunoexpression of a-catenin and g-catenin was observed, only negative b-catenin expression was associated with patient poorer
overall survival in the univariate analyses. In the multivariate analysis, b-catenin immunoexpression and residual tumour were
shown to be independent prognostic factors for survival (Faleiro-Rodrigues et
al 2004b).
In the present study, when E-cadherin and b-catenin were assessed together in a Cox«s
multivariate regression analysis to determine whether the immunoexpression of
these two proteins continued to be independent prognostic factors, only b-catenin continued to be an independent prognostic
factor of poor survival. The loss of b-catenin expression, FIGO stage IV and residual tumour, when considered
with parameters that had a significant impact on overall survival as peritoneal
metastasis, peritoneal cytology, lymphatic/vascular invasion and E-cadherin
expression, were shown to be the strongest independent predictors of poor
survival. The results of this study suggest that when b-catenin and E-cadherin are assessed in a multivariate
analysis, the loss of b-catenin proves to be a
more important prognostic marker than the loss of E-cadherin in patients with
ovarian carcinomas.
A study by the group of Akimoto et al, showed that the
expression of E-cadherin in murine adenocarcinomas correlated well with the
expression of b-catenin. They also
showed that reduced expression of b-catenin in these tumours correlated with enhanced metastasis formation
(Akimoto et al, 1999). Whether b-catenin
alone could have affected the propensity of these tumour cells to metastasise
is unclear. However, some recent studies show an independent role of the
catenins in tumour invasion and metastases (Kawanishi et al, 1995; Vermeulen et
al, 1995).
Studies on the molecular organization of the ECCU
using recombinant proteins have demonstrated b-catenin to play a central role in the formation of
the E-cadherin complex (Oyama et al, 1994; Kawanishi et al, 1995; Vermeulen et
al, 1995; Harington and Syrigos, 2000). In general, adhesion between normal
epithelial cells is strong and stable. For tumour cells to dissociate, invade
and metastasize, cell-to-cell associations must be disrupted. In our series of
ovarian carcinomas, despite the small number of tumours showing loss of expression
for b-catenin the observation that
a) 15 carcinomas demonstrated
negative expression for b-catenin,
of which 3 were negative and 12 were positive for E-cadherin, respectively, and b) b-catenin
expression was shown to be an independent prognostic factor in a previous
(Faleiro-Rodrigues et al 2004b) and in the present study, reinforces the
viewpoint that b-catenin is a crucial and
indispensable component in the formation of the ECCU, and that loss of b-catenin alone may be sufficient to disassemble the adherent
junction, leading to loss of intercellular adhesion. Thus, loss of b-catenin expression may be an important step in the
development of a malignant tumour, by this approach, enable the dissociation of
cells from the primary tumour, and thus possibly contribute to tumour cell
invasion and tumour peritoneal implantation in ovarian cancer patients.
Although the prognostic value of b-catenin needs to be supported by more studies and a
larger number of patients, this retrospective study, suggests that the
immunohistochemical assessment of b-catenin into negative versus positive expression on primary ovarian
carcinomas may prove to be a useful marker for selecting a small group of
patients with a high risk of suffering an unfavourable clinical outcome. Whether
this information can be used to stratify patients for therapeutic strategies
also needs to be explored in future clinical studies.
In several carcinomas, loss of b-catenin expression by immunohistochemistry has been
associated with malignant transformation as increased invasiveness, disease
progression, and poor prognosis (Takayama et al, 1996; Jawhari et al, 1997;
Muzio et al, 1999; Ramesh et al, 1999; Garcia del Muro et al, 2000; Kageshita
et al, 2001; Tanaka et al, 2002). However, the molecular mechanisms that bring
about the loss of b-catenin in these tumours
have not been characterized and yet to be investigated. The causal mechanism
for the loss of b-catenin protein
expression in our series of ovarian carcinomas is not clear. Several mechanisms
may impair b-catenin from being expressed, such as
hypermethylation of the b-catenin gene (CTNNB1)
promoter, CTNNB1 mutations and deletions (Ebert et al, 2003; Ueda et al, 2001).
It is now of interest to evaluate further the molecular mechanisms that underlie
the observed loss of b-catenin observed in the
present study. Future investigations on the regulation of the expression of b-catenin may elucidate possible mechanisms resulting
in the loss of this protein. Nevertheless, irrespective of the mechanism that
impairs the expression of b-catenin,
based on the above results, it seems that tumour cells may become increasingly
invasive and show an aggressive cellular phenotype upon the loss of b-catenin, which may be an important step in the
progression of ovarian carcinomas.
In conclusion, these findings suggest that b-catenin immunoexpression may assist in the
identification of a group of patients who run a higher risk of an unfavourable
disease outcome, and may be a useful prognostic marker for the clinical assessment
of epithelial ovarian cancer complementary to other established prognostic
factors as FIGO tumour stage and residual tumour. It should be noted that these
results need be supported by more studies and a larger number of patients.
Acknowledgements
This project was supported by a Ph.D. grant PRAXIS
XXI/BD/9615/96 from the Foundation of Science and Technology (FCT).
References
Akimoto
T, Kawabe S, Grothey A, Milas L (1999) Low
E-cadherin and b-catenin expression correlates with increased
spontaneous and artificial lung metastases of murine carcinomas. Clin Exp Metastasis 17, 171-176.
Birchmeier
W, Behrens J (1993) Cadherin
expression in carcinomas: role in the formation of cell junctions and the
prevention of invasiveness. Biochim
Biophys Acta 1198, 11-26.
Ebert
MP, Yu J, Hoffmann J, Rocco A, Rocken C, Kahmann S, Muller O, Korc M, Sung JJ,
Malfertheiner P (2003) Loss of b-catenin expression in
metastatic gastric cancer. J Clin Oncol
21, 1708-1714.
Faleiro-Rodrigues
C, Macedo-Pinto I, Pereira D, Lopes CS (2004a)
Prognostic value of E-cadherin immunoexpression in patients with primary
ovarian carcinomas. Ann Oncol 15,
1532-1542.
Faleiro-Rodrigues
C, Macedo-Pinto I, Pereira D, Lopes CS (2004b)
Loss of b-catenin, is associated with poor survival in
ovarian carcinomas. Int J Gynecol Pathol
23, 337-346.
Frixen
UH, Behrens J, Sachs M, Eberle G, Voss B, Warda A, Lochner D, Birchmeier W (1991) Ecadherin mediated cell-cell
adhesion prevents invasiveness of human carcinoma cells. J Cell Biol 113, 173-185.
Garcia
del Muro X, Torregrosa A, Munoz J, Castellsague X, Condom E, Vigues F, Arance
A, Fabra A, Germa JR (2000) Prognostic
value of expression of E-cadherin and b-catenin in bladder cancer. Eur J Cancer 36, 357-362.
Harington
KJ, Syrigos KN (2000) The role of E-cadherin-catenin
complex: more than an intercellular glue? Ann
Surg Oncol 7, 783-788.
Hinck
L, Nthke IS, Papkoff J, Nelson WJ (1994)
Dynamics of cadherin/catenin complex formation: novel protein interactions
and pathways of complex assembly. J Cell
Biol 125, 13271340.
Jawhari
A, Jordan S, Poole S, Browne P, Pignatelli M, Farthing MJ (1997) Abnormal immunoreactivity of the E-cadherin-catenin complex
in gastric carcinoma: relationship with patient survival. Gastroenterology 112, 46-54.
Kageshita
T, Hamby CV, Ishihara T, Matsumoto K, Saida T, Ono T (2001) Loss of b-catenin expression
associated with disease progression in malignant melanoma. Br J Dermatol 145, 210216.
Kawanishi
J, Kato J, Sasaki K, FujiisS, Watanabe N, Niitsu Y (1995) Loss of E-cadherin dependent cell-cell adhesion due to
mutation of the b-catenin gene in a human cancer cell line
HSC-39. Molecular Cell Biol 15,
1175-1181.
Kemler
R (1993) From cadherins to catenins:
cytoplasmatic protein interactions and regulation of cell adhesion. Trends Genetics 9, 317-321.
Lo
Muzio LL, Staibano S, Pannone G, Grieco M, Mignogna MD, Cerrato A, Testa NF, De
Rosa G (1999) b and g-catenin expression in oral
squamous cell carcinoma. Anticancer Res
19, 3817-3826.
Mareel
M, Bracke M, Van Roy F (1994) Invasion
promoter versus invasion suppressor molecules: the paradigm of E-cadherin. Mol Biol Rep 19, 45-67.
Oyama
T, Kanai Y, Ochiai A, Akimoto S, Oda T, Yanigihara K, Nagafuchi A, Tsukita S,
Shibamoto S, Ito F et al (1994) A
truncated b-catenin disrupts the interaction between
E-Cadherin and a-catenin: A cause of loss of intercellular
adhesiveness in human cancer cell lines. Cancer
Res 54, 6282-6287.
Ozawa
M, Ringwald M, Kemler R (1990) Uvomorulin-catenin
complex formation is regulated by aspecific domain in the cytoplasmic region of
the cell adhesion molecule. Proc Natl Acad Sci USA 87, 4246-4250.
Ozols
RF, Rubin SC, Thomas GM, Robboy SJ (2000)
Epithelial ovarian cancer. In Hoskins W J, Perez CA, Young RC (eds.):
Principles and Practice of Gynecologic Oncology, 3rd edition. Philadelphia: PA:
Lippincott Williams & Wilkins; 981-1058.
Ramesh
S, Nash J, McCulloch PG (1999) Reduction
in membranous expression of b-catenin and increased
cytoplasmic E-cadherin expression predict poor survival in gastric cancer. Br J Cancer 81, 1392-1397.
Rubin
SC, Sabbatini P, Randall ME (2002) Ovarian
Cancer. In: Pazdur R, Coia LR, Hoskins W J,Wagman LD (eds): Cancer Management:
A Multidisciplinary Approach, 3rd edition. 287-307.
Skubitz APN (2002)
Adhesion molecules. Cancer Treat Res 107, 305-329.
Takayama
T, Shiozaki H, Shibamoto S, Oka H, Kimura Y, Tamura S, Inoue M, Monden T, Ito
F, Monden M (1996) b-catenin expression in human
cancers. Am J Pathol 148, 39-46.
Tanaka
M, Kitajima Y, Edakuni G, Sato S, Miyazaki K (2002) Abnormal expression of E-cadherin and b-catenin may be a molecular
marker of submucosal invasion and lymph node metastasis in early gastric
cancer. Br J Surgery 89, 236-244.
Tsukita
S, Tsukita S, Nagafuchi A, Yonemura S (1992)
Molecular linkage between cadherins and actin filaments in cell-cell
adherens junctions. Curr Opin Cell Biol 4,
834-839.
Ueda
M, Gemmill RM, West J, Winn R, Sugita M, Tanaka N, Ueki M, Drabkin HA (2001) Mutations of the b- and g-catenin genes are uncommon
in human lung, breast, kidney, cervical and ovarian carcinomas. Br J Cancer 85, 64-68.
Van Aken E, De Wever O, Correia da Rocha AS,
Mareel M (2001) Defective
E-cadherin/catenincomplexes in human cancer. Virchows Archn 439, 725-751.
Vermeulen
SJ, Bruyneel EA, Bracke ME, De Bruyne GK, Vennekens KM, Vleminckx KL, Berx GJ,
van Roy FM, Mareel MM (1995) Transition
from the noninvasive to the invasive phenotype and loss of b-catenin in human colon
cancer cells. Cancer Res 55,
4722-4728.
Vleiminckx
K, Vakaet l Jr, Mareel M, Fiers W, Van Roy F (1991) Genetic manipulation of E-Cadherin expression by epithelial
tumor cells reveals an invasion suppressor role. Cell 66, 107-119.
Cristina Faleiro-Rodrigues
