| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
1 Robert Bosch Hospital, Department of Clinical Chemistry, Stuttgart; 2 Dr. Margarete Fischer-Bosch Institute of Clinical Pharmacology, Stuttgart; 3 Department of Mathematics, University of Stuttgart, Stuttgart; and 4 Department of Hematology and Oncology, University Medical Center, Marburg, Germany
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design: Expression of TGF-ß receptor I (TßRI) and TGFß receptor II (TßRII) was retrospectively analyzed by immunohistochemistry in 246 breast cancer patients.
Results: Expression of TßRI was strongly correlated with tumor size (P < 0.001) and nodal status (P = 0.012) but only weakly with overall survival (P = 0.056). In contrast, TßRII was prognostic for overall survival in univariate analysis (P = 0.0370). In estrogen receptor (ER) -negative patients TßRII expression was correlated with highly reduced overall survival (P = 0.0083). In multivariate analysis TßRII proved to be an independent and highly significant prognostic marker with a hazard ratio of 6.8. Simultaneous loss of both ER and TßRII was associated with longer overall survival times comparable with those of ER-positive patients.
Conclusions: The results of this exploratory study show that TßRII is an independent, highly significant prognostic indicator for overall survival in ER-negative patients. In addition our results are supportive of a mechanism of breast cancer progression in which a selective loss of the tumor inhibitory action of TGFß takes place, whereas tumor- promoting aspects remain intact.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
TGF-ß is a pleiotropic growth factor, which affects many different cell functions such as proliferation, extracellular matrix synthesis, and immune responses (7) .
TGF-ß signals are mediated by specific transmembrane receptors. TGF-ß receptor type I and II (TßRI, TßRII) are serine-threonine kinases, which form a heteromeric signal transduction complex upon ligand binding. TßRII phosphorylates TßRI, which activates TßRI kinase and initiates downstream signaling (8) . Expression of TßRII is hormonally regulated, but expression of TßRI is not (9) .
Conflicting data exist about the influence of TGF-ß on the development and progression of breast cancer. TGF-ß is a very potent inhibitor of primary human mammary epithelial cells, and most human breast cancer cell lines are growth inhibited by TGF-ß as well (10 , 11) . However, in later tumor stages TGF-ß appears to become a promoter of progression (12 , 13) , and stimulation of angiogenesis, induction of extracellular matrix degradation, or the inhibition of antitumor immune responses prevail over the inhibitory effects on proliferation.
Loss of expression or functional inactivation of TßRI or TßRII leads to resistance against TGF-ß (14 , 15) . Furthermore, defects in downstream signaling components have been associated with altered sensitivity to TGF-ß and tumor progression in different tumor types (16 , 17) .
However, the clinical implications of such findings are still unclear. Therefore, in a retrospective immunohistochemical study we analyzed the correlation of TßRI and TßRII expression with overall survival in 246 breast cancer patients with a median follow-up of 5.7 years. To discriminate between hormone-dependent and -independent effects patients were additionally stratified for their ER status.
Our results show that expression of TßRII is correlated with poor prognosis and represents an independent prognostic marker in the subgroup of ER-negative patients.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 was reached (19)
. Of the 246 patients 148 were classified as ER positive and 87 as ER negative. There were 142 PR-positive patients and 92 PR negative. All of the patients had initially undergone either mastectomy or a breast-conserving resection of their primary carcinomas. There were 166 patients being treated in the adjuvant setting, and 44 received palliative treatment after relapse. Adjuvant treatment with tamoxifen was given to 71 patients, 49 patients received adjuvant treatment with cyclophosphamide, methotrexate and fluorouracil, 46 with anthracycline regimes (either mitoxantrone and cyclophosphamide; fluorouracil, doxorubicin and cyclophosphamide; or bonnadonna regimen), and 91 patients were treated with radiation. The sum of all treatments is >166, as some patients received more than one treatment.
Antibodies and Immunohistochemical Method.
Affinity-purified rabbit polyclonal antibodies raised against TßRI (R-20) and TßRII (C-16; Santa Cruz Biotechnology, Inc., Heidelberg, Germany) were used for immunostaining of the tissues. Both antibodies were used previously for immunohistochemical studies by other groups (20, 21, 22, 23)
.
Sections of 3 µm were prepared from the paraffin block of each patient tumor. The tissue sections were deparaffinized and rehydrated in descending alcohol dilutions. Before staining on a TechMate instrument (Dako, Glostrup, Denmark) the tissue sections were subjected to antigen retrieval in microwave oven using a citrate buffer solution (Dako). Endogenous biotin, biotin receptors, or avidin binding sites present in the tissue were blocked using an Avidin/Biotin Blocking kit (Vector Laboratories, Inc., Burlingame, CA). Endogenous peroxidase activity was blocked by treatment with hydrogen peroxide. Staining was performed with the Dako ChemMate Detection kit, Peroxidase/3,3'-diaminobenzidine. The slides were incubated with primary antibodies for 25 min at room temperature. The optimum concentration of antibodies for staining, as determined by previous experiments, was 5 µg/ml for anti-TßRI and 2 µg/ml for anti-TßRII. Tissues were then incubated with biotinylated goat antirabbit IgG (diluted 1:100) for 25 min at room temperature, followed by incubation with peroxidase-conjugated streptavidin for 25 min. The immunoreaction was visualized by using diaminobenzidine in the presence of H2O2 resulting in brown colored final reaction products. Tissues were counterstained with hematoxylin.
Skin tissue was used as positive control. Negative controls were obtained by omission of the primary antibody.
The stained tumor tissues and surrounding normal tissues were scored blindly with respect to clinical patient data.
Staining intensity was visually scored in four degrees: absent (0), weak (1) , moderate (2) , and strong (3) . The percentage of TßRI- or TßRII-positive tumor cells was graded as absent (0), 1–20% (1) , 21–50% (2) , 51–80% (3) , and 81–100% (4) . An immunoreactive score (IRS) index was calculated as the product of both values. With respect to the well-known heterogeneity of breast tumors initially the IRS was used for analysis of TßRI or TßRII expression. However, further analysis revealed that heterogeneity in the samples was not sufficient to justify the continuous use of IRS for clinical application. Therefore, we reduced our assessment to two grades, because it has been already shown for scoring of erbB-2 staining in terms of the decision of treatment with Herceptin (24) . For these reasons tissue samples were classified into either IRS = 0 (negative) or IRS >0 (positive) staining for TGF-ß receptors.
Statistical Methods.
Data assessment was made using the statistics software program SPSS (SPSS Software GmbH, Munich, Germany). Survival curves were established by the Kaplan-Meier method, and comparisons between survival curves were performed by the log-rank test. Patients who died from unrelated causes were considered as censored by the time of their death.
Multivariate analysis in the subgroup of ER-negative patients was performed using Cox regression analysis in a model, which included tumor size, nodal status, distant metastases, and grading.
Differences in the TGF-ß receptor expression between normal and tumor tissues, and associations between TGF-ß receptor expression and other parameters such as tumor size, nodal status, grading, and hormonal status were assessed by 
2- or Fisher’s exact-test.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. As expected there was a significant association between the classical prognostic factors (tumor size, nodal status, distant metastases, grading, and ER) and outcome. No correlation was observed for menopausal and PR status.
|
TßRI was detected in 54% of the normal tissue samples. In tumor tissue expression of TßRI was significantly more frequent. Of the tumor samples, 78% showed positive staining for the receptor (P = 0.0012). TßRII on the other hand was detected in the majority of the samples, and no significant differences between normal (64%) and tumor (72%) tissues were observed (Table 2)
.
|
shows a representative immunohistochemical staining for TßRI and TßRII in tumor and adjacent normal tissue.
|
.
|
TßRII showed no significant correlations with any of the parameters analyzed.
Expression of the TGF-ß Receptors and Overall Survival.
Loss of TGF-ß receptor expression has been associated with loss of TGF-ß growth-inhibitory effects and progression to more aggressive tumor types (21
, 25)
. However, these studies did not consider overall patient survival. To assess the influence of TGF-ß receptor expression on prognosis of breast cancer patients Kaplan-Meier survival curves were calculated and log-rank analysis performed.
Surprisingly, patients with detectable expression of TßRII had significantly shorter overall survival times in comparison with patients with undetectable receptor expression (P = 0.0370; Fig. 2B
). The mean overall survival time was 11.5 years [95% confidence interval (CI), 10.3–12.7] in patients without TßRII expression and 10.4 years (95% CI, 9.5–11.4) in patients with detectable expression of TßRII.
|
) and probably due to the strong correlation of TßRI expression with tumor size and nodal status (Table 3)
.
ER Subgroup Analysis.
To take hormonal influences into consideration the tissue specimen were additionally subgrouped into either ER positive or ER negative, and Kaplan-Meier survival curves were calculated.
Expression of TßRII defined a subset of patients in the ER-negative subgroup with strongly reduced overall survival times (P = 0.0083; Fig. 3B
). The mean overall survival time was 11.3 years (95% CI, 10.2–12.4) for TßRII-negative patients compared with only 8.3 years (95% CI, 6.8–9.7) for TßRII-positive patients. Additional stratification for treatment regimens (cyclophosphamide, methotrexate, and fluorouracil; and anthracyclin containing) gave similar results. In both treatment groups patients with detectable expression of TßRII had a considerably worse prognosis than patients without detectable expression of TßRII (data not shown).
|
). The mean overall survival time was 11.1 years (95% CI, 9.6–12.6) for TßRII-negative patients and 11.4 years (95% CI, 10.3–12.6) for TßRII-positive patients. TßRI expression had no effect on overall survival after stratification for the ER status. In the ER-positive as well as in the ER-negative subgroup patients with detectable TßRI expression had shorter survival times but differences to TßRI-negative patients were not significant (data not shown).
Multivariate Analysis.
A multivariate Cox proportional hazards regression analysis was carried out to establish if expression of TßRII was an independent prognostic marker in the subgroup of ER-negative patients. The model initially included all of the parameters that were predictive of overall survival in the univariate analysis of the entire study group as presented in Table 1
(tumor size, nodal status, distant metastases, and tumor grading). A forward stepwise procedure was adopted to obtain the final model of significant predictors for overall survival consisting of the factors distant metastases, nodal status, and expression of TßRII. Inclusion of tumor size or grading into this model did not improve the log partial likelihood significantly. In the subgroup of ER-negative patients expression of TßRII was strongly associated with poor outcome, with a hazard ratio of 6.8 (Table 4)
.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Thus far only a few studies have examined the role of the TGF-ß receptors in breast cancer tissues; none of these studies considered the influence of receptor expression on disease outcome. In this retrospective exploratory study we have therefore analyzed the expression of TßRI and TßRII in 246 human breast cancer tissues and adjacent normal tissues, and evaluated their association with prognosis. To assess hormonal influences on the TGF-ß system we additionally included the ER status of the tumors into the analysis.
It has been suggested previously that loss of TßRII expression may contribute to breast cancer progression and may be associated with a more aggressive phenotype (21)
. Our data show only a slight loss of TßRII expression with increasing tumor grade, which does not reach statistical significance. TßRII was expressed in a large part of the normal (64%) as well as the tumor tissue (72%; Table 2
). The number of tumors with detectable expression of TßRII decreased only slightly with high tumor grading (G1 + G2 75%, G3 64%). TßRII was not correlated with any of the clinical parameters analyzed, including tumor size, node status, distant metastasis, grading, ER, PR, and menopausal status (Table 3)
. However, expression of TßRII had a strong negative influence on prognosis. Patients with detectable expression of TßRII had significantly decreased survival times (P = 0.0370; Fig. 2B
). These data indicate that in breast cancer, loss of TGF-ß growth-inhibitory effects is not caused by loss of TßRII expression. Similar results were obtained previously for pancreatic cancer in which enhanced expression of TßRII was significantly correlated with reduced overall survival. In pancreatic cancer increased expression of TßRII was associated with high expression of plasminogen activator inhibitor 1 and matrix metalloproteinase 9 indicating a dissociation between TGF-ß signals leading to growth inhibition and extracellular matrix production (31)
.
Because TGF-ß signals are mediated by a heteromeric complex of TßRI and TßRII another possible explanation for the absence of TGFß growth-inhibitory effects and tumor progression in breast cancer could be a loss of TßRI expression. In colon and prostate cancer loss of TßRI expression was associated with increased malignant potential (25
, 32
, 33)
. No comparable studies exist for breast cancer. In our patient collective, TßRI was expressed in most of the tumor tissues (78%; Table 2
), and expression was more often detected in tumors of lower grade (81%; Table 3
). Furthermore, expression of TßRI had a weak negative effect on prognosis. However, this correlation with survival is probably because TßRI expression was significantly more often detected in larger and node-positive tumors (Table 3)
, two factors known to have a strong negative impact on survival (1)
.
In summary our results suggest that loss of TGF-ß growth-inhibitory effects in breast cancer cannot be attributed to either loss of TßRI or TßRII expression.
It has been implicated before that in advanced tumor stages the growth-inhibitory component of TGF-ß signaling selectively gets lost, whereas tumor promoting effects gain importance (13 , 31) . In some tumor types active TGF-ß signaling seems to play an important role in the progression to more aggressive phenotypes. Several reports exist that link TGF-ß signaling to cell invasiveness and formation of metastasis (13 , 34, 35, 36) .
Our data support a role for TGF-ß in breast cancer progression, because both receptors were expressed in the majority of tumor tissues and expression was associated with poor outcome. It can be assumed that the TGF-ß receptors detected in this study are able to transduce signals, because mutational inactivation of TßRI or TßRII appears to be a rare event in breast cancer. Thus far the only mutation in TßRI has been found in breast cancer metastases, but this mutation does not seem to occur frequently (37) . Four inactivating mutations in TßRII have been found recently in recurrent breast tumors, but no TßRII mutations have been detected in primary breast cancers (38) . In addition, an immunohistochemical study on the expression of Smad2, phosphorylated Smad2, and Smad4 suggests, that the majority of invasive breast carcinomas are able to actively mediate TGF-ß signals (17) .
The present study indicates that the ER status of a tumor is an important marker for the transition of TGF-ß from tumor suppressor to tumor promoter. In ER-negative tumors expression of TßRII was associated with a subset of tumors that seemed to be highly aggressive leading to strongly reduced overall survival times (8.3 years; Fig. 3B
). Simultaneous loss of both ER and TßRII, on the other hand, was associated with longer overall survival times (11.3 years) comparable with those of ER-positive patients with and without TßRII expression (11.4 years and 11.1 years, respectively). Expression of TßRII proved to be an independent prognostic marker in the subgroup of ER-negative patients (Table 4)
.
Differences in treatment appear not to influence the impact of TßRII expression on overall survival in ER-negative patients. We distinguished between the two most commonly used treatment regimens, namely cyclophosphamide, methotrexate, and fluorouracil, and anthracycline-containing regimens. In each treatment group prognosis was worse for patients with detectable expression of TßRII.
It has already been shown that a cross-talk between growth factors and steroid hormone receptors may be relevant to the regulation of growth and differentiation processes in hormone responsive tissues: the effect of epidermal growth factor seems to be partially mediated through the ER even in the absence of estradiol (39 , 40) . A similar cross-talk appears to exist for TGF-ß signaling and the ER. The transcriptional activity of Smad3 can be suppressed by ER, whereas ER-mediated transcriptional activity can be increased by activation of TGF-ß signaling (41) .
In ER-positive tumors, TGF-ß seems to act in an autocrine inhibitory fashion because a rapid increase in TGF-ß2 levels under treatment with tamoxifen was correlated previously with clinical remission in patients with metastatic breast cancer (6) . Loss of ER could evoke distinct changes in the cellular response to TGF-ß and represent the starting point for a loss of TGF-ß growth-inhibitory effects.
Therefore, a therapeutic approach that inhibits TGF-ß signal transduction might turn out to be specifically effective for ER-negative patients with detectable expression of TßRII. A soluble TGF-ß receptor type II protein that interferes with TGF-ß binding to endogenous TGF-ß receptors has been shown previously to reduce tumor cell motility, intravasation, and distant metastasis in a mouse model (34) .
In conclusion, in our study TßRII proved to be an independent prognostic marker in the subgroup of ER-negative patients. According to the criteria of the American Society of Clinical Oncology (1) this study can be classified as evidence level III. Our results support a mechanism for breast cancer progression in which a selective loss of the growth-inhibitory action of TGF-ß takes place, whereas tumor-promoting aspects of TGF-ß signaling remain intact. We identified the ER as an important marker for changes in the tumor response to TGF-ß. The precise interactions between ER and TGF-ß signaling on the molecular level remain to be clarified. However, analysis of TßRII expression in ER-negative patients could be relevant for treatment decisions, as patients without detectable expression of TßRII had a prognosis similar to that of ER-positive patients.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Cornelius Knabbe, Department of Clinical Chemistry, Robert Bosch Hospital, Auerbachstr. 110, 70376 Stuttgart, Germany. Phone: 49-711-81013500; Fax: 49-711-81013618. E-mail: cornelius.knabbe{at}rbk.de
Received 2/25/03; revised 10/14/03; accepted 10/20/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. A. Teicher Transforming Growth Factor-{beta} and the Immune Response to Malignant Disease Clin. Cancer Res., November 1, 2007; 13(21): 6247 - 6251. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. B BUCK and C. KNABBE TGF-Beta Signaling in Breast Cancer Ann. N.Y. Acad. Sci., November 1, 2006; 1089(1): 119 - 126. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Beisner, M. B. Buck, P. Fritz, J. Dippon, M. Schwab, H. Brauch, G. Zugmaier, K. Pfizenmaier, and C. Knabbe A Novel Functional Polymorphism in the Transforming Growth Factor-{beta}2 Gene Promoter and Tumor Progression in Breast Cancer. Cancer Res., August 1, 2006; 66(15): 7554 - 7561. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Volinia, G. A. Calin, C.-G. Liu, S. Ambs, A. Cimmino, F. Petrocca, R. Visone, M. Iorio, C. Roldo, M. Ferracin, et al. A microRNA expression signature of human solid tumors defines cancer gene targets PNAS, February 14, 2006; 103(7): 2257 - 2261. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Cell Growth & Differentiation |
