BMC - Breast Cancer Research

IntroductionExpression of the putative Wnt signaling inhibitor Dickkopf-3 (DKK3) is frequently lost in human cancer diseases due to aberrant 5'-cytosine methylation within the DKK3 gene promoter. Since other Wnt signaling inhibitors have been reported as targets of epigenetic inactivation in human breast cancer, we wondered whether DKK3 expression is epigenetically silenced during breast carcinogenesis as well and thus might contribute to oncogenic Wnt signaling commonly found in this disease. Methods: DKK3 mRNA expression and DKK3 promoter methylation were determined by RT-PCR, realtime PCR and methylation-specific PCR in breast cell lines (n = 9), normal breast tissues (n = 19), and primary breast carcinomas (n = 150), respectively. In vitro DNA demethylation was performed by incubating breast cell lines with 5-aza-2'-deoxycytidine and trichostatin A. DKK3 protein expression was analyzed by immunohistochemistry in breast carcinomas (n = 16) and normal breast tissues (n = 8). Methylation data were statistically correlated with clinical patient characteristics. All statistical evaluations were accomplished with SPSS 14.0 software. Results: DKK3 mRNA was downregulated in five of seven breast cancer cell lines and in 68% of primary breast carcinomas (n = 40) compared with benign cell lines and normal breast tissues, respectively. A DNA demethylating treatment of breast cell lines resulted in strong induction of DKK3 mRNA expression. In tumorous breast tissues, DKK3 mRNA downregulation was significantly associated with DKK3 promoter methylation (P

IntroductionThe amplification event occurring at chromosome locus 11q13, reported in several different cancers, includes a number of potential oncogenes. We have previously reported amplification of one such oncogene, CCND1, to be correlated with an adverse effect of tamoxifen in premenopausal breast cancer patients. Overexpression of cyclin D1 protein however, confers tamoxifen resistance but not a tamoxifen induced adverse effect. Potentially, co-amplification of an additional 11q13 gene, with a resulting protein overexpression, is required to cause an agonistic effect. Moreover, during 11q13 amplification a deletion of the distal 11q region has been described. In order to assess the potential impact of the deletion we have examined a selected marker for this event.MethodArray comparative genomic hybridization analysis was employed to identify and confirm changes in the gene expression of a number of different genes mapping to the 11q chromosomal region, associated with CCND1 amplification. The subsequent protein expression of these candidate genes was then examined in a clinical material of 500 primary breast cancers from premenopausal patients randomized to either tamoxifen or no adjuvant treatment. The protein expression was also compared to the gene expression data in a subset of 56 breast cancer samples. Results: Cortactin and FADD overexpression was linked to CCND1 amplification, determined by fluorescent in situ hybridization, but was not associated with a diminished effect of tamoxifen. However, deletion of distal chromosome 11q, defined as downregulation of the marker Chk1, was associated with an impaired tamoxifen response, and interestingly, also with low proliferative breast cancer of low grade. For Pak1 and cyclin D1 the protein expression corresponded to the gene expression data. Conclusion: The results indicate that many 11q13 associated gene products are overexpressed in conjunction with cyclin D1 but not linked to an agonistic effect of tamoxifen. Finally, the deletion of distal 11q, linked to 11q13 amplification, might be an important event affecting breast cancer outcome and tamoxifen response.

IntroductionThe PI-3 kinase/Akt pathway, operating downstream of EGFR and HER2, is implicated in cell migration and survival. EGFR and HER2 are expressed in circulating tumor cells, however, the activation status of downstream signaling molecules has not yet been reported. Methods: To investigate EGFR/HER2/PI-3 kinase/Akt expression in circulating tumor cells, we used peripheral blood mononuclear cells from 32 cytokeratin-19 mRNA-positive patients with early (n=16) and metastatic (n=16) breast cancer. PBMC cytospins were double stained with cytokeratin antibody along with either of the following: EGFR, phospho-EGFR, HER2, phospho-PI-3 kinase or phospho-Akt antibodies, respectively. Results: EGFR and HER2 were expressed in circulating tumor cells of 38% and 50% patients with early and 44% and 63% patients with metastatic disease, respectively. Interestingly, phospho-PI-3 kinase and phospho-Akt expressions were similar at 88%(14 out of 16) and 81% (13 out of 16), respectively, in circulating tumor cells of patients with early and metastatic disease. Phospho-EGFR was observed in circulating tumor cells of two (33%) early and six (86%) metastatic EGFR-positive patients. Immunomagnetic separation of peripheral blood mononuclear cells, using EpCAM antibody and subsequent double staining experiments of circulating tumor cells showed that EGFR was co-expressed with HER2, phospho-Akt and phospho-PI-3 kinases, indicating activation of the corresponding survival signaling pathway. Conclusions: Our findings demonstrate that circulating tumor cells express receptors and activated signaling kinases of the EGFR/HER2/PI-3 kinase/Akt pathway which could be used as targets for their effective elimination.

No Abstract

This is a letter regarding a recently published article in Breast Cancer Research by Tamimi et al (Breast Cancer Res 2008, 10(4):R67). Concerns regarding categorization of breast cancer molecular classes using surrogate immunohistochemical markers are discussed.

IntroductionWe have previously reported that tumour-specific expression of the rate-limiting enzyme in the mevalonate pathway, 3-hydroxy-3-methylglutharyl-coenzyme A reductase (HMG-CoAR), is associated with more favourable tumour parameters in breast cancer. In the present study, we examined the prognostic value of HMG-CoAR expression in a large cohort of primary breast cancer patients with long-term follow up. Methods: The expression of HMG-CoAR was assessed by immunohistochemistry (IHC) on tissue microarrays with tumour specimens from 498 consecutive breast cancer cases with a median follow-up of 128 months. Kaplan Meier analysis and Cox proportional hazards modelling were used to estimate the risk of recurrence free survival (RFS) and breast cancer specific survival (BCSS). Results: In line with our previous findings, tumour-specific HMG-CoAR expression was associated with low grade (p

IntroductionMenopausal hormone therapy has been reported to increase the risk of certain subtypes of breast cancer and to be associated with a favorable survival. This could either be due to an increased mammographic surveillance or a biological effect. We assessed these associations in a Swedish cohort of postmenopausal breast cancer patients holding information on mammographic examinations, menopausal hormone therapy use, other breast cancer risk factors, and cancer treatment. Methods: We analyzed 2,660 postmenopausal women aged 50-74 years, diagnosed with invasive breast cancer in 1993-1995 and followed until the end of year 2003 (median follow-up 9 years and 3 months). We assessed the influence of hormone therapy before diagnosis on tumor characteristics and breast cancer specific survival. We analyzed hormone therapy before diagnosis by regimen (estrogen-progestin therapy or estrogen alone therapy), recency (current or past), and duration of use (

Controversy surrounds the potential clinical importance of estrogen receptor (ER) beta in breast cancer and three recent papers have sought to resolve this. In this issue, Novelli and colleagues explored the significance of ERbeta1 expression in 936 breast cancer patients and showed diverse relationships according to lymph node status. A second paper examined 442 breast cancers in which ERbeta1 was an independent predictor of recurrence, disease-free and overall survival. Finally a third paper showed ERbeta2 was a powerful prognostic indicator in 757 breast cancers but this was dependent on cellular location with nuclear ERbeta2 expression predicting good survival whilst cytoplasmic expression predicted worse outcome. These papers point to a clinical role for ERbeta in breast cancer and will be discussed.

IntroductionDespite the fact that patients above 65 years of age have the highest incidence of developing breast cancer, these patients are excluded from clinical trials in most cases. Furthermore, most physicians tend to therapy regimens without the use of dose dense-highly active taxane based treatments because a lack of data regarding toxicities of these compounds in elderly patients. Methods: Pooled side-effect data were analyzed from four prospective, randomized clinical trials in which patients of different age groups (

IntroductionIsolated disseminated tumor cells (DTC) are regarded as surrogate markers for minimal residual disease in breast cancer. Characterization of these cells could help understand the known limitations of adjuvant therapy. Of particular interest is their estrogen receptor (ER) status as endocrine adjuvant therapy remains a cornerstone of breast cancer treatment. Methods: Bone marrow (BM) aspirates from 254 primary breast cancer patients were included in the study. A double immunofluorescence staining procedure was established for the identification of cytokeratin-positive (CK) / ERalpha positive cells. ERalpha status of the primary tumor was immunohistochemically assessed using the same antibody against ERalpha. Results: In 107 of 254 (42%) breast cancer patients CK-positive cells could be detected in the BM. More than one DTC in the BM was observed in 38 of the 107 patients. The number of detected cells ranged between 1 and 55 / cells per 2x106 mononuclear cells. Disseminated tumor cells demonstrated ERalpha positivity in 12% of the patients. The ERalpha expression was heterogeneous in 10 of the 38 (26%) patients with more than one DTC. Concordance rate of ERalpha status between primary tumor and DTC was 28%. Only 12 of 88 patients with ERalpha positive tumors also had ERalpha positive DTCs. Conclusions: Primary tumor and DTCs displayed a concordant ERalpha status in only 28% of cases. Most of the DTCs were ERalpha negative despite an ERalpha positive primary tumor. These findings further underline the distinct nature of DTCs and may explain the failure rates seen in conventional endocrine adjuvant therapy.

Lately, understanding the role of cancer stem cells in tumor initiation and progression became a major focus in stem cell biology and in cancer research. Considerable efforts, such as the recent studies by Honeth and colleagues, published in the June issue of Breast Cancer Research, are directed towards developing clinical applications of the cancer stem cell concepts. This work shows that the previously described CD44+CD24- stem cell phenotype is associated with basal-type breast cancers in human patients, in particular BRCA1 inherited cancers, but does not correlate with clinical outcome. These very interesting findings caution that the success of our efforts in translating cancer stem cell research into clinical practice depends on how thorough and rigorous we are at characterizing these cells.

The application of high-throughput genomic technologies has revealed that individual breast tumors display a variety of molecular features that require more personalized approaches to treatment. Several recent studies have demonstrated that a cross-species analytic approach provides a powerful means to filter through genetic complexity by identifying evolutionarily conserved genetic networks that are fundamental to the oncogenic process. Mouse-human tumor comparisons will provide insights into cellular origins of tumor subtypes, define interactive oncogenetic networks, identify potential novel therapeutic targets, and further validate as well as guide the selection of genetically engineered mouse models for preclinical testing.

IntroductionBreast cancers can be classified using whole genome expression into distinct subtypes that show differences in prognosis. One of these groups, the Basal-like subtype, is poorly differentiated, highly metastatic, genomically unstable, and contains specific genetic alterations like the loss of TP53. The loss of the retinoblastoma tumor suppressor encoded by the RB1 locus is a well-characterized occurrence in many tumor types, however, its role in breast cancer is less clear with many reports demonstrating loss of heterozygosity that does not correlate with loss of RB1 protein expression. Methods: We used gene expression analysis for tumor subtyping and polymorphic markers located at the RB1 locus to assess the frequency of loss of heterozygosity in 88 primary human breast carcinomas and their normal tissue genomic DNA samples. Results: RB1 loss of heterozygosity was observed at an overall frequency of 39%, with a high frequency in Basal-like (72%) and Luminal B (62%) tumors. These tumors also concurrently showed low expression of RB1 mRNA. p16INK4a was highly expressed in Basal-like tumors, presumably due to a previously reported feedback loop caused by RB1 loss. An RB1 loss of heterozygosity signature was developed and shown to be highly prognostic, and was potentially a predictive marker of response to neoadjuvant chemotherapy. Conclusions: These results suggest that the functional loss of RB1 is common in Basal-like tumors, which may play a key role in dictating their aggressive biology and unique therapeutic responses.

The hypothesis that cancer stem cells are responsible for the chemo-resistant and metastatic phenotypes of many breast cancers has gained support using cell sorting strategies to enrich for the tumor-initiating population of cells. However, the mechanisms regulating the cancer stem cell pool are less clear. Two recent publications suggest that loss of p53 permits expansion of presumptive cancer stem cells in mouse mammary tumors and human breast cell lines. These results add restriction of cancer stem cells as a new tumor suppressor activity attributed to p53.

IntroductionPatients with primary operable estrogen receptor negative (ER-) breast cancer account for approximately 30% of all cases and generally have worse prognosis than ER+ patients. Nevertheless, a significant proportion of ER- cases have favourable outcome and could potentially benefit from a less aggressive course of therapy. Identification of such good prognosis patients however remains difficult and at present only possible through histopathological factors. Methods: Building on a previously identified 7-gene prognostic IR-module for ER negative breast cancer, we develop a novel statistical tool based on Mixture Discriminant Analysis in order to build a classifier that can accurately identify ER- patients with good prognosis. Results: We report the construction of a 7-gene expression classifier that accurately predicts, across a training cohort of 183 ER- tumors and six independent test cohorts (a total of 469 ER- tumors), ER- patients of good prognosis (in test sets, average predictive value = 94% (range 85%-100%), average hazard ratio = 0.15 (0.07-0.36) P

IntroductionThe purpose of this study was to investigate the effect of CAD prompts on reader behaviour in a large sample of breast screening mammograms by analysing the relationship of the presence and size of prompts to the recall decision. Methods: Local research ethics committee approval was obtained; informed consent was not required. Mammograms were from women attending routine mammography at two breast screening centres in 1996. Films, previously double read, were re-read by a different reader using CAD. The study material included 315 cancer cases comprising all screen detected cancer cases, all subsequent interval cancers and 861 normal cases randomly selected from 10,267 cases.. Ground truth data were used to assess the efficacy of CAD prompting. Associations between prompt attributes and tumour features or reader recall decisions were assessed by Chi squared tests. Results: There was a highly significant relationship between prompting and a decision to recall for cancer cases and for a random sample of normal cases (p

IntroductionRelaxin is increased in human breast cancer and was shown to promote cancer cell migration in carcinoma cells of the breast, prostate and thyroid. In estrogen receptor alpha-negative MDA-MB-231 human breast cancer cells, relaxin was shown to down-regulate the metastasis-promoting protein S100A4 (metastasin), a highly significant prognostic factor for poor survival in breast cancer patients. The cellular mechanisms of relaxin exposure in breast cancer cells are not fully understood. The aim of this study was to investigate short-term and long-term effects of relaxin on cancer cell motility and S100A4 expression and to determine the long-term effects of relaxin on in vivo tumor growth in an estrogen-independent context. Methods: We have established stable transfectants of highly invasive estrogen receptor alpha-negative MDA-MB-231 human breast cancer cells with constitutive expression of bioactive H2-relaxin (MDA/RLN2). RLN2 secretion was determined by ELISA. Relaxin receptor RXFP1 (Relaxin-family-peptide) was detected by RT-PCR and its activation was assessed by induction of cyclic-AMP. Stable MDA/RLN2 clones and RLN2 treated MDA-MB-231 cells were subjected to motility and in vitro-invasion assays. Proliferation was assessed in bromodeoxyuridine (BrdU) and MTT assays. S100A4 expression was determined by RT-PCR and Western Blot. Specific small interfering RNA was employed to down-regulate relaxin receptor and S100A4. MDA/EGFP vector control and two MDA/RLN2 clones were injected subcutaneously in nude mice to determine tumor growth and cancer cell invasiveness in vivo. Xenograft tumor tissues were assessed by histology and immunohistochemistry and frozed tissues were used for the detection of S100A4 and RLN2. Results: Short-term exposure to relaxin for 24 hours increased cell motility in a relaxin receptor-dependent manner. This increase in cell motility was mediated by S100A4. Long-term exposure to relaxin secreted from stable transfectants reduced cell motility and in vitro invasiveness. Relaxin decreased cell proliferation and down-regulated cellular S100A4 levels in MDA-MB-231 and T47D breast cancer cells. Stable MDA/RLN2 transfectants produced smaller xenograft tumors containing reduced S100A4 protein levels in vivo. Conclusions: Our results indicate that long-term exposure to relaxin confers growth inhibitory and anti-invasive properties in estrogen-independent tumors in vivo which may in part be mediated through a down-regulation of S100A4.

IntroductionErbB2, a member of the epidermal growth factor receptor (EGFR) family, is overexpressed in 20% to 30% of human breast cancer cases and forms oncogenic signalling complexes when dimerised to ErbB3 or other EGFR family members. Methods: We crossed mouse mammary tumour virus (MMTV)-myr-Akt1 transgenic mice (which express constitutively active Akt1 in the mammary gland) with MMTV-c-ErbB2 transgenic mice to evaluate the role of Akt1 activation in ErbB2-induced mammary carcinoma using immunoblot analysis, magnetic resonance spectroscopy and histological analyses. Results: Bitransgenic MMTV-c-ErbB2, MMTV-myr-Akt1 mice develop mammary tumours twice as fast as MMTV-c-ErbB2 mice. The bitransgenic tumours were less organised, had more mitotic figures and fewer apoptotic cells. However, many bitransgenic tumours displayed areas of extensive necrosis compared with tumours from MMTV-c-ErbB2 mice. The two tumour types demonstrate dramatically different expression and activation of EGFR family members, as well as different metabolic profiles. c-ErbB2 tumours demonstrate overexpression of EGFR, ErbB2, ErbB3 and ErbB4, and activation/phosphorylation of both ErbB2 and ErbB3, underscoring the importance of the entire EGFR family in ErbB2-induced tumourigenesis. Tumours from bitransgenic mice overexpress the myr-Akt1 and ErbB2 transgenes, but there was dramatically less overexpression and phosphorylation of ErbB3, diminished phosphorylation of ErbB2, decreased level of EGFR protein and undetectable ErbB4 protein. There was also an observable attenuation in a subset of tyrosine-phosphorylated secondary signalling molecules in the bitransgenic tumours compared with c-ErbB2 tumours, but Erk was activated/phosphorylated in both tumour types. Finally, the bitransgenic tumours were metabolically more active as indicated by increased glucose transporter 1 (GLUT1) expression, elevated lactate production and decreased intracellular glucose (suggesting increased glycolysis). Conclusion: Expression of activated Akt1 in MMTV-c-ErbB2 mice accelerates tumourigenesis with a reduced requirement for signalling through the EGFR family, as well as a reduced requirement for a subset of downstream signaling molecules with a metabolic shift in the tumours from bitransgenic mice. The reduction in signalling downstream of ErbB2 when Akt is activated suggest a possible mechanism by which tumour cells can become resistant to ErbB2-targeted therapies, necessitating therapies that target oncogenic signalling events downstream of ErbB2.

The insulin-like growth factor axis, which has been shown to protect cells from apoptosis, plays an essential role in normal cell physiology and in cancer development. The family of insulin-like growth factor binding proteins (IGFBPs) has been shown to have a diverse spectrum of functions in cell growth, death, motility, and tissue remodeling. Among the six IGFBP family members, IGFBP-5 has recently been shown to play an important role in the biology of breast cancer, especially in breast cancer metastasis; however, the exact mechanisms of action remain obscure and sometimes paradoxical. An in-depth understanding of IGFBP-5 would shed light on its potential role as a target for breast cancer therapeutics.

IntroductionCirculating tumor cells (CTCs) are detectable in most cancer patients and they can meet an existing medical need to monitor cancer patients during a course of treatment and to help determine recurrent disease. CTCs are rarely found in the blood of cancer patients and enrichment is necessary for sensitive CTC detection. Most CTC enrichment technologies are anti-EpCAM antibody based even though CTC identification criteria are cytokeratin positive (CK+), CD45 negative (CD45-) and 4'6-diamidino-2-phenylindole (nuclear stain) positive (DAPI+). However, some tumor cells express low or no EpCAM. Here we present a highly sensitive and reproducible enrichment method that is based on binding to anti-CK alone or a combination of anti-CK and anti-EpCAM antibodies. Methods: Blood samples from 49 patients with metastatic breast cancer were processed using the CellSearch™ system (Veridex, LLC, Raritan, NJ, USA), in parallel with our CTC assay method. We used anti-CK alone or in combination with anti-EpCAM antibodies for CTC enrichment. Brightfield and fluorescence labeled anti-CK, anti-CD45 and DAPI (nuclear stain) images were used for CTC identification. The Ariol® system (Genetix USA Inc, San Jose, CA, USA) was used for automated cell image capture and analysis of CTCs on glass slides. Results: Our method has the capability to enrich three types of CTCs including CK+&EpCAM+, CK+&EpCAM-/low, and CK-/low&EpCAM+ cells. In the blind method comparison, our anti-CK antibody enrichment method showed a significantly higher CTC positive rate (49% vs. 29%) and a larger dynamic CTC detected range (1 to 571 vs. 1 to 270) than that of the CellSearch™ system in the total of 49 breast cancer patients. Our method detected 15 to 111% more CTCs than the CellSearch™ method in patients with higher CTC counts (>20 CTCs per 7.5 ml of blood). The three fluorescent and brightfield images from the Ariol® system reduced the number of false-positive CTC events according to the established CTC criteria. Conclusion: Our data indicate that the tumor-specific intracellular CK marker could be used for efficient CTC enrichment. Enrichment with anti-CK alone or combined with anti-EpCAM antibodies significantly enhances assay sensitivity. The three fluorescent and brightfield superior images with the Ariol® system reduced false-positive CTC events.