| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Pathology, Division of Molecular Biology, and2 Department of Medical Oncology, Erasmus MC Rotterdam, Rotterdam, The Netherlands;3 Department of Chemical Endocrinology, University Medical Centre Nijmegen, Nijmegen, The Netherlands
aaddress correspondence to this author at: Department of Chemical Endocrinology, University Medical Centre Nijmegen, PO Box 9101, 6500 HB Nijmegen, The Netherlands; fax 31-243541484, e-mail F.Sweep{at}ace.umcn.nl
Advanced breast cancer is frequently treated with the antiestrogen tamoxifen. Response is observed in approximately one half of estrogen receptor (ER)-positive patients but ultimately fails because the cancer becomes resistant. The mechanisms for resistance of breast cancer to tamoxifen treatment are not yet understood (1). The breast cancer anti-estrogen resistance protein 1 (BCAR1) gene was identified in a search for genes that cause proliferation of estrogen-dependent breast cancer cells in the presence of an antiestrogen drug (2). To assess the role of the BCAR1 protein in breast cancer progression, we developed a specific and sensitive ELISA. Here we report the use of this BCAR1 ELISA to measure BCAR1 protein concentrations in human breast cancer cytosols to predict success of tamoxifen treatment for advanced disease.
Primary invasive breast tumors were collected between 1978 and June 1995. Our study design was approved by the medical ethics committee of the Erasmus MC Rotterdam. Selection of samples for analysis of response to tamoxifen in recurrent breast cancer was based on the availability of stored cytosol extracts (in liquid nitrogen), which remained after routine ER and progesterone receptor analyses (3). Tissue specimens that were sampled after neoadjuvant treatment or obtained from a biopsy were not included. In addition, the samples selected were only from ER-positive (>10 fmol/mg of protein) tumors of patients who developed recurrent disease and were treated with tamoxifen as first-line therapy. Of these 592 patients, 133 had received adjuvant chemotherapy (89 receiving cyclophosphamide–methotrexate–5-fluorouracil and 44 receiving 5-fluorouracil–epirubicin–cyclophosphamide), and none had received adjuvant hormonal therapy. The median age at the start of tamoxifen treatment for recurrent disease was 61 years (range, 28–91 years); 144 patients were premenopausal and 448 were postmenopausal. During the follow-up period, 527 of 592 patients (89%) showed tumor progression, and 433 patients (73%) died. Median follow-up of patients alive after primary surgery (n = 159) was 95 months (range, 10–196 months), and median follow-up after start of first-line tamoxifen treatment was 36 months (range, 2–131 months). Objective response to tamoxifen therapy was classified by the standard criteria for complete remission (complete disappearance of all lesions) and partial remission (tumor reduction of 
50%) established by the Union International Contra Cancer (4). The patients with no change of disease (i.e., between a tumor size reduction of 50% or a size increase of 25%) were divided in patients with no change for >6 months [considered as having a similar prognosis as partial remission (5)] and those with no change for 
6 months (considered as progressive disease). Therefore, overall response (clinical benefit) was defined as objective response plus no change for >6 months, as has been done previously (6)(7)(8). BCAR1 concentrations were measured by the BCAR1 ELISA (9) in 592 cytosols from primary invasive breast tumors, prepared and processed as recommended by the European Organization for Research and Treatment of Cancer (10).
The strength of the associations of BCAR1 protein concentrations with continuous variables was estimated by Spearman rank correlation (rs) using the STATA statistical package, release 8.2 (STATA Corp.). The strength of the association of BCAR1 protein with other variables (used as grouping variable) was tested with the nonparametric Wilcoxon rank-sum test or the Kruskal–Wallis test, followed by a Wilcoxon-type test for trend across ordered groups where appropriate. All P values were two-sided. A P value <0.05 was considered statistically significant. BCAR1 protein concentrations ranged from 0.02 to 19.8 µg/g of cytosolic protein (median, 3.7 µg/g of protein) and exhibited an approximately log-normal distribution. Within this selection of ER-positive breast tumor specimens, we observed a weak positive correlation between the concentrations of ER and BCAR1 (rs = 0.15; n = 592; P <0.001); however, we observed no significant association of BCAR1 concentrations with adjuvant chemotherapy, disease-free interval, site of relapse, age, or menopausal status.
In the present study, 373 patients (63%) showed clinical benefit on tamoxifen therapy, whereas 219 patients showed tumor progression soon after start of treatment or no change for <6 months. The relationship with the type of response to tamoxifen therapy was examined with logistic regression analysis. Odds ratios (ORs) were calculated and are presented with their 95% confidence intervals. The likelihood ratio test in multivariable logistic regression models was used to test for differences and interactions. To analyze the relationship of BCAR1 with benefit of tamoxifen therapy, the 592 patients were divided into four equal-sized groups based on the quartile protein concentrations of BCAR1. From quartile 1 to quartile 4, the proportion of patients who showed clinical benefit on tamoxifen therapy ranged from 68% to 59%. In addition to univariate analysis, we performed multivariate logistic regression analysis in which a predictive base model (
2 = 47.82; df = 5; P <0.0001) was defined that included the traditional predictive factors for response to treatment, i.e., menopausal status, disease-free interval, dominant site of relapse, and ER concentrations (Table 1
). BCAR1 was subsequently added separately as a log-transformed continuous variable or as a categorized variable.
|
For the BCAR1 concentrations treated as log-transformed continuous variables [ln(µg/g of protein)], the multivariable OR was 0.68 (95% confidence interval, 0.47–0.99; P = 0.044). The multivariable ORs for BCAR1 concentrations in quartiles decreased from 0.81 for quartile 2, to 0.64 for quartile 3, to 0.61 for quartile 4, but this trend was not statistically significant. There were no statistically significant interactions of BCAR1 protein concentrations and any of the factors of the base model with the efficacy of tamoxifen treatment. We did not observe an association of BCAR1 protein concentrations with duration of response to first-line tamoxifen therapy in these analyses. Isotonic regression analysis (6)(11) was used to search for a BCAR1 protein concentration cut point in an exploratory analysis. An optimized cut point was defined that identified a minor group of 54 patients (9% of total) with low BCAR1 protein concentrations (<1.43 µg/g of protein) who received relatively high clinical benefit (81%) from tamoxifen therapy. Using this cut point, when we compared the patients with high BCAR1 concentrations with those with low concentrations, the 538 patients with high BCAR1 concentrations performed poorly on tamoxifen in both univariate (OR = 0.36; P = 0.004) and multivariate logistic regression analysis (OR = 0.31; P = 0.002). However, within these 538 patients, we observed no significant trend between the BCAR1 concentration and the rate of response. In an attempt to further optimize the relationship, we used the defined cut point to subdivide the first quartile (presented in Table 1
). From the univariate and multivariate analyses, it was clear that all subgroups with BCAR1 concentrations >1.43 µg/g of protein in the primary tumor had a statistically significant increased risk for no clinical benefit from tamoxifen treatment for recurrent disease.
This study shows that BCAR1 protein concentrations in the primary tumor can predict the efficacy of tamoxifen therapy in recurrent breast cancer. BCAR1, the human homolog of the murine p130Cas protein, belongs to a small family of adaptor proteins shown to participate in many cellular processes (12). The involvement of BCAR1/p130Cas (and its numerous interaction partners) in the regulation of actin cytoskeleton, cell migration, and regulation of growth and apoptosis has been investigated (12)(13). In addition, BCAR1/p130Cas has been implicated in pathogen internalization, which depends on actin cytoskeleton rearrangement. A possible role in cancer was suggested by the observation that p130Cas is the prime, and likely a crucial, phosphorylation target in v-SRC and v-CRK transformed cells (12). We have identified BCAR1/p130Cas in a functional screen for proliferation of human breast cancer cells resistant to an antiestrogen drug and provided evidence that high concentrations of BCAR1 protein in primary breast tumors are associated with poor prognosis (2)(14). Using the sensitive and specific BCAR1 ELISA, we strengthened the previous conclusions that low concentrations of BCAR1 protein in the primary tumors predict favorable response of advanced breast cancers to tamoxifen therapy. This association is independent of the traditional markers of response in multivariate analysis, including ER concentrations. In addition to its predictive value for tamoxifen resistance in recurrent breast cancer, as shown here, the prognostic value of BCAR1 in primary breast cancer has been confirmed separately (L.C.J. Dorssers et al., manuscript submitted). BCAR1/p130Cas has also been associated with malignant melanoma and some types of leukemia (12). Recently, overexpression of BCAR1/p130Cas in vitro was shown to interfere with KAI1/CD82-mediated suppression of metastasis in prostate cancer cells (15). Further evaluation of the role of BCAR1 in various kinds of malignancies and other diseases may benefit from our quantitative detection methodology for BCAR1 protein.
Acknowledgments
This research was supported in part by the Revolving Fund of the Erasmus MC (Grant 99.760), the Dutch Cancer Society, and the Association for International Cancer Research (Grant 00-38). We acknowledge the contributions of Simone van Broekhoven, Daniëlle de Jong, Harry Peters, Henk Portengen, Marion Meijer-van Gelder, and Doorlène van Tienoven.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  T. van Agthoven, A. M. Sieuwerts, M. E. Meijer-van Gelder, M. P. Look, M. Smid, J. Veldscholte, S. Sleijfer, J. A. Foekens, and L. C.J. Dorssers Relevance of Breast Cancer Antiestrogen Resistance Genes in Human Breast Cancer Progression and Tamoxifen Resistance J. Clin. Oncol., February 1, 2009; 27(4): 542 - 549. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. C. J. Dorssers, N. Grebenchtchikov, A. Brinkman, M. P. Look, S. P. J. van Broekhoven, D. de Jong, H. A. Peters, H. Portengen, M. E. Meijer-van Gelder, J. G. M. Klijn, et al. The Prognostic Value of BCAR1 in Patients with Primary Breast Cancer Clin. Cancer Res., September 15, 2004; 10(18): 6194 - 6202. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
