Strong androgen receptor expression can aid in distinguishing GATA3+ metastases

Metastatic Pleomorphic Lobular Carcinoma of the Breast to the Urinary Bladder: A Report of 10 Cases and Assessment of TRPS1 in the Differential Diagnosis With Plasmacytoid Urothelial Carcinoma

CONTEXT.—

Metastatic pleomorphic lobular carcinoma (MPLC) to the bladder is rare and has considerable histologic and immunohistochemical overlap with plasmacytoid urothelial carcinoma (PUC).

OBJECTIVE.—

To distinguish MPLC from PUC morphologically and immunohistochemically, including a newer marker, TRPS1.

DESIGN.—

Ten MPLCs to the bladder were reassessed and stained with estrogen, progesterone, and androgen receptors; GATA3; keratin 5/6; HMWK; GCDFP-15; and TRPS1. Sixteen PUCs constituted controls.

RESULTS.—

We studied 4 transurethral resections of bladder tumors and 6 biopsies from 10 women (median age, 69 years) who had breast cancer on average 15 years prior. Microscopic patterns included single cells and cords of cells (n = 4), nests/sheets of dyscohesive cells (n = 2), or both (n = 4). All tumors had cells with voluminous eosinophilic cytoplasm and eccentric nuclei mimicking PUC, and 7 of 10 tumors had signet ring cells. MPLCs were positive for estrogen (8 of 10), progesterone (3 of 7), and androgen (4 of 10) receptors; GCDFP-15 (7 of 10); GATA3 (9 of 10); HMWK (7 of 8); and TRPS1 (7 of 10). No MPLCs stained for keratin 5/6 (n = 9). Of 16 PUCs, 2 showed faint and 2 demonstrated strong TRSP1 staining; 7 of 16 were negative for p63.

CONCLUSIONS.—

MPLC to bladder often presents in patients with a remote history of breast cancer, exhibiting significant histologic and immunohistochemical overlap with PUC. Based on prior works and the current study, estrogen receptor (particularly SP-1), mammaglobin, and p63 help differentiate MPLC from PUC. Keratin 5/6 may aid in distinguishing a less frequent basal-type PUC because it is typically negative in MPLC. Some PUCs express TRPS1. Caution should be exercised because immunophenotypes of these tumors greatly overlap, and ramifications of misclassification are major.

The androgen receptor interacts with GATA3 to transcriptionally regulate a luminal epithelial cell phenotype in breast cancer

BACKGROUND

The androgen receptor (AR) is a tumor suppressor in estrogen receptor (ER) positive breast cancer, a role sustained in some ER negative breast cancers. Key factors dictating AR genomic activity in a breast context are largely unknown. Herein, we employ an unbiased chromatin immunoprecipitation-based proteomic technique to identify endogenous AR interacting co-regulatory proteins in ER positive and negative models of breast cancer to gain new insight into mechanisms of AR signaling in this disease.

RESULTS

The DNA-binding factor GATA3 is identified and validated as a novel AR interacting protein in breast cancer cells irrespective of ER status. AR activation by the natural ligand 5α-dihydrotestosterone (DHT) increases nuclear AR-GATA3 interactions, resulting in AR-dependent enrichment of GATA3 chromatin binding at a sub-set of genomic loci. Silencing GATA3 reduces but does not prevent AR DNA binding and transactivation of genes associated with AR/GATA3 co-occupied loci, indicating a co-regulatory role for GATA3 in AR signaling. DHT-induced AR/GATA3 binding coincides with upregulation of luminal differentiation genes, including EHF and KDM4B, established master regulators of a breast epithelial cell lineage. These findings are validated in a patient-derived xenograft model of breast cancer. Interaction between AR and GATA3 is also associated with AR-mediated growth inhibition in ER positive and ER negative breast cancer.

CONCLUSIONS

AR and GATA3 interact to transcriptionally regulate luminal epithelial cell differentiation in breast cancer regardless of ER status. This interaction facilitates the tumor suppressor function of AR and mechanistically explains why AR expression is associated with less proliferative, more differentiated breast tumors and better overall survival in breast cancer.

Clinical Implications of Androgen-Positive Triple-Negative Breast Cancer

This review summarizes the recent findings of a vast array of studies conducted on androgen receptor-positive triple-negative breast cancer (AR-positive TNBC) to provide a better understanding of this specific breast cancer subgroup. AR expression is correlated with higher age, lower histological grade, lower proliferation index Ki-67, spiculated masses, and calcifications on mammography. Studies investigating the correlation between AR expression and lymph node metastasis are highly discordant. In addition, results regarding prognosis are highly contradictory. AR antagonists are a promising novel therapeutic approach in AR-positive TNBC. However, AR signaling pathways should be more investigated in order to understand the influence of AR expression on TNBC more thoroughly.

Androgen Receptor mRNA levels determine the prognosis in triple-negative breast cancer patients

BACKGROUND

Anti-Androgen Receptor (AR) therapy holds promise for a subset of AR expressing triple-negative breast cancer (TNBC) patients. However, current AR assays are suboptimal in detecting the dynamic range of AR expression, contributing to its controversial role in TNBC disease prognosis. This study is aimed at evaluating the feasibility of qRT-PCR to sensitively and robustly detect AR mRNA levels for prognostication.

METHODS

mRNA expression profiling was performed on FFPE blocks from a retrospective cohort of 101 TNBC patients using qRT-PCR and compared with AR protein expression by immunohistochemistry . Statistical analyses included Spearman's rank correlation, Chi-square and Kaplan-Meier analyses. Distant Metastasis Free Survival was used as the end point in survival analysis.

RESULTS

AR mRNA expression was observed in 34/101 patients (34%) whereas 12/80 cases (15%) were positive by IHC. qRT-PCR could thus detect more AR positive patients as compared to IHC, with 75% (9/12) concordance between the two methods. Co-expression of GATA3 and FOXA1 mRNA was observed in 85 and 88% of AR mRNA positive tumors, respectively. AR mRNA positivity was significantly correlated with age at disease onset (p = 0.02), high FOXA1/GATA3 (p < 0.05) and distant recurrence. AR mRNA positive patients had poorer DMFS (43%; p = 0.002). DMFS dropped further to 26% (p = 0.006) in AR (+)/high FOXA1/GATA3 patients. AR mRNA expression together with node positivity had the worst DMFS (23%; p < 0.0001) compared to patients who were either positive for any one of these, or negative for both AR and node status. Low Ki67 mRNA with AR mRNA positivity also had poorer DMFS (39%; p = 0.001) compared to patients expressing low Ki67 with no AR mRNA expression.

CONCLUSION

qRT-PCR was more sensitive and reliable in detecting the dynamic expression levels of AR compared to IHC and this variation could be explained by the higher sensitivity of the former method. High AR mRNA expression was strongly associated with expression of AR protein, high FOXA1/GATA3 mRNA, and with poor prognosis. qRT-PCR was more efficient in detecting the AR positive cases compared to IHC. A distinct signature involving high GATA3/FOXA1, low Ki67, and node positivity in AR mRNA positive tumors correlated with poor prognosis. Thus, AR mRNA screening can serve as an effective prognostic marker along with offering potential targeted therapy options for TNBC.

Strong androgen receptor expression is not useful in distinguishing GATA3 + metastases

GATA binding protein 3 (GATA3) immunohistochemical expression is commonly considered to be a sensitive and specific diagnostic marker for breast and urothelial carcinomas in surgical pathology practice. However, since its expression has been also demonstrated in other tumors, GATA3 should be better used in conjunction with other immunohistochemical markers to establish tumor primitivity in metastatic setting. Interestingly, GATA3 expression seems to be significantly correlated with androgen receptor (AR) expression in breast carcinoma. In addition, strong AR expression -defined as immunohistochemical positivity in more than 60% of tumor cells- was suggested to be 100% specific for breast origin in GATA3+ metastases. The aim of this study was to verify whether strong AR expression may actually be useful to determine primivity in GATA3+ metastatic setting. Thus, we investigated AR and GATA3 immuno-expression in a cohort of metastatic tumors derived from urothelial, breast, endometrial and salivary gland carcinomas. We did not find any GATA3 or AR expression in the metastases from endometrial or salivary gland carcinomas, while GATA3 expression was seen in the majority of metastases from urothelial or breast carcinomas. In addition, strong AR expression was seen in 73% and in 47% of metastatic breast and urothelial carcinomas, respectively. On the whole, our findings confirm that GATA3 is sensitive and specific for breast and urothelial origin in metastatic setting. According to our results, strong AR expression is not useful to distinguish breast from urothelial primitivity, as previously suggested.