Expression of androgen receptor in inflammatory breast cancer and its clinical relevance
Yun Gong MD,
Department of Pathology, The University of Texas MD Anderson Cancer Center, Houston, Texas
Morgan Welch Inflammatory Breast Cancer Research Program and Clinic, The University of Texas MD Anderson Cancer Center, Houston, Texas
Corresponding author: Yun Gong, MD, Department of Pathology, Unit 53, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030; Fax: (713) 794-5664; firstname.lastname@example.org
Presented at the 100th Annual Meeting of the United States and Canadian Academy of Pathology; San Antonio, Texas; February 26 to March 4, 2011.
We thank Tamara K. Locke from the Department of Scientific Publications at The University of Texas MD Anderson Cancer Center for editing this article.
Inflammatory breast cancer (IBC) is characterized by an aggressive clinical course with early metastasis and frequent resistance to conventional therapies. Identifying a novel therapeutic approach may improve the prognosis for patients with IBC. Because androgen receptor (AR)-expressing tumors may be targeted by anti-AR therapy, the authors examined the prevalence of AR expression in IBC tumors and explored its clinical relevance.
Tissue microarrays of 88 IBC tumors were stained immunohistochemically with monoclonal antibody against AR, and the results were correlated with clinicopathologic parameters and survival outcomes.
The median follow-up was 10.8 years. AR was positive in 39% of the IBC tumors and in approximately one-third of estrogen receptor (ER)-negative and progesterone receptor (PR)-negative tumors. AR positivity was significantly associated with lymphovascular invasion (P = .01) but not with other clinicopathologic parameters. There was a trend toward an association between AR expression and PR expression (P = .07). In univariate survival analysis, patients who had AR-negative/ER-negative tumors had significantly worse overall survival (P = .03) and disease-specific survival (P = .04) than patients who had tumors with other combinations of AR/ER status.
Inflammatory breast cancer (IBC) is a rare and very aggressive disease with characteristic symptoms that include redness, swelling, tenderness, and warmth in the breast because of dermal lymphatic occlusion by tumor emboli. Despite multimodality treatment approaches, the prognosis of patients with IBC remains poor, with a clinical outcome much worse than that of patients with noninflammatory breast cancer.[1, 2] Identifying novel therapeutic targets is highly desirable.
Emerging evidence is indicating that the androgen-signaling pathway plays a role in breast carcinogenesis through regulating estrogen-responsive genes. Androgen receptor (AR) expression in breast cancer has become an interesting topic of research, because AR can be targeted by antiandrogen agents such as bicalutamide and enzalutamide, which have been widely used to treat patients with AR-positive, advanced prostate cancers. Similar to estrogen receptor (ER) and progesterone receptor (PR), AR is a member of the steroid hormone receptor family. Depending on the study population, testing method, and cutoff values used, AR reportedly is expressed in 70% to 90% of invasive breast cancers.[4-10]
A significant association between AR and ER expression has been reported in breast carcinoma, not otherwise specified (NOS).[7, 10, 11] Peters et al observed coexpression of the 2 receptors in 80% to 90% of breast tumor cells. AR positivity also reportedly is associated with older age and/or postmenopausal status at diagnosis, smaller tumor size, lower tumor grade, negative human epidermal growth factor receptor 2 (HER2) status, and negative lymph node metastasis.[6-9, 12-17] However, AR positivity is also reported in 30% to 50% of ER-negative/PR-negative and triple-negative tumors.[6-8, 10, 12, 13, 16, 18]
The clinical significance of AR in breast cancer in the literature is somewhat inconsistent. In general, AR-positive status is significantly associated with better clinical outcomes than AR-negative tumors.[5, 6, 9, 11, 12, 15, 17, 19, 20] However, in some studies, the significant prognostic relevance of AR was observed in ER-positive tumors,[6, 9, 11] but not in ER-negative tumors[5, 6, 11, 18] or triple-negative tumors; whereas, in other studies, the prognostic significance of AR was observed in both ER-negative and triple-negative tumors.[12, 17, 20]
To date, the prevalence of AR expression in IBC tumors and its association with survival in patients with IBC have not been evaluated. The objectives of the current study were to examine the prevalence of AR expression in IBC tumors and to explore its prognostic value.
MATERIALS AND METHODS
This study included 88 patients with primary IBC who were treated at The University of Texas MD Anderson Cancer Center from September 1994 to August 2004 and for whom tumor tissue and clinical follow-up information were available. The diagnosis, preoperative and postoperative treatments of these patients, biomarker study (encompassing ER, PR, and HER2 status), and tissue microarray (TMA) construction have been previously reported. Notably, ER, PR, and HER2 status and tumor characteristics were evaluated on pretreated tumor samples obtained by core-needle biopsy, and a TMA was built up using postneoadjuvant residual tumors. This study was approved by the Institutional Review Board.
Immunohistochemical staining for AR was performed on 4 μm-thick paraffin sections of TMA slides. Antigen retrieval was conducted by steaming the slides in 10 mM citrate buffer (pH 6.0) for 25 minutes. The sections were then incubated with monoclonal antibody against AR441 (dilution 1:30; Dako, Carpinteria, Calif) for 30 minutes at room temperature. The Refine Polymer Detection Kit (Leica Microsystems, Buffalo Grove, Ill) was used for the detection of bound antibody, with 3,3′-diaminobenzidine as the chromogen. The tissues were then counterstained with Mayer hematoxylin solution and observed under a lightfield microscope. The results were evaluated with known positive and negative tissue controls. Nuclear staining in greater than 10% of invasive tumor cells was defined as positive.
Summary statistics of patient characteristics are provided in frequency tables and percentages. The Fisher exact test was used to evaluate associations between AR expression and clinicopathologic variables. We used the Kaplan-Meier method to estimate overall survival (OS) and disease-specific survival (DSS). Comparisons between patient groups with respect to survival endpoints were made using log-rank tests. All tests were 2-sided, and P-values ≤.05 were considered statistically significant. Statistical analysis was carried out using SAS version 9 (SAS Institute, Cary, NC).
The median follow-up was 10.8 years (range, 0.7-14.5 years), and the 5-year OS and DSS rate were 46% and 49%, respectively. AR was positive in 39% of IBC tumors overall, and in 33.3% of ER-negative tumors, 29.3% of PR-negative tumors, and 42.6% of triple-negative tumors (Table 1, Fig. 1A-C). Positive AR expression was significantly associated with lymphovascular invasion (P = .01). However, there was no significant association between AR positivity and ER, PR, or HER2 status or other pathologic parameters listed in Table 1, but there was a trend toward an association between AR and PR expression (P = .07) (Table 1).
Table 1. Relation Between Androgen Receptor Status and Clinicopathologic Factors in Inflammatory Breast Cancer
AR Expression Status in Patients With IBC: No. of Patients (%)
Abbreviations: AR, androgen receptor; ER, estrogen receptor; HER2, human epidermal growth factor receptor 2; IBC, inflammatory breast cancer; PR, progesterone receptor.
P values were calculated using the Fisher exact test.
Age at diagnosis, y
Lymph node status
Univariate survival analysis identified no significant association between AR positivity and OS (P = .14) or DSS (P = .18). When evaluating AR in ER-positive and ER-negative groups, respectively, we observed that, in the ER-negative tumors, there was a trend toward an association between positive AR expression and better OS (P = .07) and DSS (P = .06). However, in the ER-positive tumors, there was no significant or trend association between AR expression and survival outcome.
When evaluating various combinations of AR and ER status (ie, AR-negative/ER-negative, AR-negative/ER-positive, AR-positive/ER-negative, or AR-positive/ER-positive tumors), we observed significant differences among the 4 groups of tumors for OS and DSS (P = .03 and P = .04, respectively), and the patients with AR-negative/ER-negative tumor had the worst survival outcome compared with other groups (Fig. 2).
In this study, we observed AR expression in 39% of IBC tumors and a trend toward an association between AR expression and PR expression. In addition, in patients with ER-negative tumors, there was a trend toward an association between positive AR expression and better survival outcome. Furthermore, patients who had AR-negative/ER-negative tumors had a significantly worse survival outcome compared with patients who had tumors that exhibited other combinations of AR/ER status.
Our findings were similar to those from some of the published studies that used breast carcinoma-NOS.[12, 17, 20] Agoff et al demonstrated that AR positivity was significantly associated with better disease-free survival in ER-negative tumors. In another study by Rakha et al, AR negativity was associated with a greater frequency of recurrence and distant metastasis in triple-negative tumors. These findings, together with our current results, suggest that AR expression in ER-negative breast carcinoma, including IBC, may have therapeutic significance, because these patients do not respond to conventional anti-ER therapies and often have a significantly worse prognosis than patients with ER-positive disease. In our study, AR was expressed in approximately one-third of ER-negative and PR-negative tumors and in 42.6% of triple-negative tumors, similar to the range of 30% to 50% reported in the literature.[6-8, 10, 12, 13, 16, 18]
However, different results regarding the clinical relevance of AR expression also have been reported in the literature. For example, Micello et al studied 232 ER-negative/PR-negative breast cancers and observed no significant correlation between AR expression with either OS or disease-free survival. Other investigators observed a significant correlation between AR expression and survival outcome in a group of patients with tumors that were unsorted for ER status, but such a correlation was not significant in patients who had ER-negative tumors or triple-negativetumors. Two other studies demonstrated that the association of AR expression with a better prognosis was significant in patients with ER-positive tumors but not in patients with ER-negative tumors.[6, 11] Perplexingly, 1 study reported that AR expression in triple-negative tumors was associated with increased breast cancer-associated mortality.
Despite the inconsistent findings in the literature regarding the biologic and clinical significance of AR in breast cancer, it has been suggested that AR serves as a potential therapeutic target for subgroups of breast carcinomas that are AR-positive but ER-negative/PR-negative. Bicalutamide is a nonsteroidal antiandrogen agent that competitively inhibits the action of androgens by binding to cytosol ARs in target tissues. Currently, a phase II clinical trial of bicalutamide is ongoing that is evaluating the efficacy of bicalutamide in women with AR-positive/ER-negative/PR-negative, metastatic breast cancers (ClinicalTrials.gov Identifier NCT00468715).[22, 23] Clinical trials with another AR inhibitor, enzalutamide, also are currently underway. The data on breast cancer survival in patients who receive AR-targeted therapy have not been published.
The functional role of AR in breast cancer seems complicated and remains to be elucidated. Existing data based on preclinical experiments indicate that an interaction between AR and ER seemingly regulates ER function in ER-positive breast cancer cells. Peters et al observed that AR bound with response elements of estrogen target genes, by which AR prevented the activation of ER-α signaling in breast cancer cells and, thus, inhibited the ER-α-mediated growth-stimulatory effect. Studies in breast cancer cell lines have demonstrated that androgen dehydroepiandrosterone sulfate (DHEA-S) can stimulate cell proliferation through ER in cells that are AR-positive/ER-positive but inhibit growth if the cells are AR-positive/ER-negative,[24-29] indicating that AR is responsible for the inhibitory effect of androgen in AR-positive/ER-negative tumors and that such tumors may respond to appropriate androgen-modulating therapy as long as they are AR-positive.
With respect to the association between AR expression and clinicopathologic variables, there are also conflicting results in the literature.[5, 8, 9, 12-14, 16, 18, 30-32] Our current study was similar to that of Gonzalez et al, who demonstrated that there was no significant association of AR expression with age, tumor size, tumor grade, lymph node involvement, stage of disease, vascular invasion, ER status, or PR status. Surprisingly, our study demonstrated a significant association between AR expression and lymphovascular invasion in IBC tumors. Notably, we previously observed the same association in another study with 91 cases of triple-negative breast carcinoma-NOS. The underlying mechanism remains to be explored. Pang et al previously reported that AR/androgen was able to up-regulate matrix metalloproteinase 13 (MMP-13) in prostate carcinoma. Gonzalez et al subsequently observed a positive and significant correlation between AR expression and the expression of MMP-1, MMP-7, and MMP-13 in breast cancer. Because MMPs are a family of degradative enzymes that contribute to tumor invasion and metastasis by destruction of the basement membrane and extracellular matrix, causing lymphovascular penetration,[5, 34, 35] it is possible that the expression of MMPs may play some role in lymphovascular invasion of AR-positive tumors. Further studies will be required to elucidate the mechanism.
To our knowledge, the current study is the first to evaluate AR expression in IBC tumors. The series had a long clinical follow-up. Because all patients with IBC received neoadjuvant treatment, it is difficult to obtain pretreated tumor samples for research purposes. We used post-treated tumor tissues built into a TMA. Therefore, our study has some limitations, because the use of a TMA may increase the risk of false-negative interpretation of AR expression. In addition, it is of concern that the use of postneoadjuvant, residual IBC tumor tissues potentially could lead to selection bias of the study population. A study by Loibl et al appears to address the relevant issue by evaluating the prognostic value of AR in 626 patients who received neoadjuvant chemotherapy. Those authors observed that, in the group of patients whose tumors achieved a pathologic complete response (pCR), there was no difference in the prognosis according to AR expression in pretreated tumors; whereas, in the group of patients whose tumors did not achieve a pCR, those with AR-positive tumors had significantly better disease-free survival and OS than those with AR-negative tumors.
In summary, the current study demonstrated that AR is commonly expressed in post-treated IBC tumors. AR positivity was significantly associated with lymphovascular invasion, and there was a trend toward an association between AR expression and PR expression. Patients who had IBC with AR-negative/ER-negative tumors had significantly worse OS and DSS than patients who had tumors that exhibited other combinations of AR/ER status. For patients with AR-positive IBC tumors, an AR-modulated therapeutic approach may add to the existing treatment to improve patients' outcomes. Further study with a larger series will be required to delineate the biologic mechanisms of AR and their clinical significance in IBC tumors.
This study was supported by the faculty research fund of The University of Texas MD Anderson Cancer Center (to Y.G.).
CONFLICT OF INTEREST DISCLOSURES
Dr. Uneo reports possible new grant support from Astellas Scientific and Medical Affairs, Inc.