Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Glucocorticoid Receptor Expression Predicts Good Outcome in response to Taxane-Free, Anthracycline-Based Therapy in Triple Negative Breast Cancer

Glucocorticoid Receptor Expression Predicts Good Outcome in response to Taxane-Free,... Hindawi Journal of Oncology Volume 2020, Article ID 3712825, 10 pages https://doi.org/10.1155/2020/3712825 Research Article Glucocorticoid Receptor Expression Predicts Good Outcome in response to Taxane-Free, Anthracycline-Based Therapy in Triple Negative Breast Cancer 1 2 1 2 Ahmed Elkashif , Victoria Bingham , Paula Haddock, Matthew P. Humphries , 2 2 1 1 Stephen McQuaid , Paul B. Mullan , Helen O. McCarthy , and Niamh E. Buckley School of Pharmacy, Queen’s University Belfast, Belfast BT9 7BL, UK Patrick G Johnston Centre for Cancer Research, Queen’s University Belfast, Belfast BT9 7AE, UK Correspondence should be addressed to Niamh E. Buckley; n.obrien@qub.ac.uk Received 30 January 2020; Accepted 16 April 2020; Published 20 May 2020 Academic Editor: Hakan Buyukhatipoglu Copyright © 2020 Ahmed Elkashif et al. )is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Triple negative breast cancer (TNBC) is a poor outcome subset of breast cancers characterised by the lack of expression of ER α, PR, and HER2 amplification. It is a heterogeneous group of cancers which fail to derive benefit from modern, more targeted treatments such as Tamoxifen and Herceptin. Current standard of care (SoC) is cytotoxic chemotherapy, which is effective for some patients, with other patients deriving little/no benefit and lacking alternative treatments. )is study has identified the glucocorticoid receptor (GR) as a potential predictive biomarker of response to anthracycline-based chemotherapy in triple negative breast cancer (TNBC). GR gene expression levels in patient samples were analysed through publicly available microarray datasets as well as protein expression through immunohistochemistry (IHC) and correlated with clinical/pathological outcomes, including survival. While the results confirmed previous observations that high GR expression is associated with poor outcome in response to taxane-based chemotherapy, this study shows for the first time that high GR expression is associated with improved outcomes in the context of anthracycline-based chemotherapy. GR therefore has the potential to be used as a predictive biomarker to guide treatment choices and ensure that patients derive the greatest benefit from first line treatment, avoiding unnecessary costs, side effects, and disease progression. amplification of the HER2 gene. )is subset is associated 1. Introduction with the poorest outcomes among breast cancers, with Breast cancer is the most common malignancy in females higher mortality rates compared to other subtypes, despite worldwide, with over 2 million new cases diagnosed in 2018 accounting for only 15–20% of cases [2]. )is is due, in part, and an estimated 600,000 deaths [1]. Following diagnosis of to the lack of molecular targets available for treatment. breast cancer, patients are stratified based on expression of Given the lack of biomarkers to guide treatment, first line estrogen receptor alpha (ERα), progesterone receptor (PR), therapy for TNBC is a regimen of adjuvant and increasingly and amplification of the human epidermal growth factor neoadjuvant, cytotoxic chemotherapy consisting of receptor 2 (HER2) gene. )is is used to guide treatments, anthracyclines and taxanes. Despite the high rates of mor- with targeted treatments including Tamoxifen and Trastu- tality, TNBC has higher response rates to chemotherapy zumab (Herceptin) used to treat ERα positive and HER2 compared to other breast cancers. )is is commonly referred positive cancers, respectively. to as the “TNBC paradox”. In the neoadjuvant setting, Triple negative breast cancer (TNBC) is a term used to patients who achieve a complete pathological response describe breast cancers which are ERα/PR negative and lack (pCR) tend to have a very good prognosis and survival rates 2 Journal of Oncology response to stress, and cell survival. It exists in the cell comparable to non-TNBC, accounting for around a third of patients [2]. )ose who do not show a response to first line cytoplasm, bound to heat shock proteins (HSPs). Its en- dogenous ligand is cortisol and upon ligand binding, GR chemotherapy or have residual disease (RD) following treatment tend to relapse in the first three years following dissociates from HSPs, dimerises, and translocates to the diagnosis and have a high risk of disease progression and nucleus. It binds to glucocorticoid response elements death [2, 3]. )is highlights the importance of biomarkers to (GREs) on DNA to transactivate or transrepress genes. GR is stratify patients and tailor treatment options accordingly involved in transrepression of genes such as those respon- and that an optimal response to first line treatment is sible for inflammation (e.g., AP1 and NFκB) while trans- paramount in assuring long term survival. Given the fact activating other genes such as those responsible for cell survival and DNA repair (e.g., STAT1 and Ets) [14–16]. that TNBC is a diagnosis of exclusion (based on the lack if receptors, as opposed to the presence), this variable response )is study implicates GR as a potential prognostic biomarker in the context of SoC which could be used to therapy is not surprising. Without the use of additional biomarkers, such diagnostic criteria leave a highly hetero- clinically to guide treatment choices and tailor chemo- therapy to individual patients. geneous population with variable treatment response rates that in theory should be able to be further stratified. )is is exemplified by the further subclassification of TNBC into the 2. Materials and Methods Vanderbilt subtypes by Lehmann et al. consisting of four 2.1. Gene Expression. )e in-house gene expression dataset subgroups, each with distinctive gene expression patterns, has been previously described [17]. )e publicly available responses to chemotherapy, and overall outcomes [4, 5]. datasets GSE5881 [18], GSE31519 [19], GSE7390 [20], )e development of personalised medicine approaches GSE10797 [21], and GSE9574 [22] were accessed online in TNBC is an area of increasing interest with the identi- using NCBI. fication and development of novel targeted treatments and/ or combinations as well as the associated companion bio- markers [6]. A major advancement in this field has been the 2.2. Tissue Microarray (TMA) and Immunohistochemistry development of immunotherapy and the use of immune (IHC). )e TMAs described in this study were constructed checkpoint inhibitors. TNBC is characterised by a high from formalin fixed paraffin embedded (FFPE) slides pre- expression of immune checkpoint proteins such as PD-L1 as pared by the Northern Ireland Biobank (NIB13-0043) and well as a degree of tumour infiltrating lymphocytes (TILs), the Breast Cancer Now Tissue Bank (TR-0055). Both the both of which are predictive of response to immunotherapy NIB (REC:11/NI/0013) and the Breast Cancer Now Tissue as well as conventional chemotherapies such as anthracy- Bank have ethical approval to use deidentified tissue samples clines [6–9]. Clinical trials have shown that response rates to from NHS tissue pathology archives (specifically the Belfast pembrolizumab, a monoclonal antibody targeting PD-1, are Health and Social Care Trust Cellular Pathology archive in increased from 5% to 18.5% by assessing PD-L1 status and the case of the NIB). )e first TMA with matched samples to only treating patients expressing this marker [10, 11]. Ad- the in-house gene expression cohort has been previously ditionally, the KEYNOTE-355 phase III clinical trial has described elsewhere [17, 23]. )e subsequent TMAs with shown the addition of pembrolizumab to chemotherapy various chemotherapy regimens have also been previously improves outcome in metastatic TNBC with the primary end described [9, 24–26]. Clinical information from these TMAs point of progression-free survival being met is summarised in Table 1. IHC - was performed in the (NCT02819518). Other targets being explored in TNBC Northern Ireland Molecular Pathology Laboratory, which include the androgen receptor (AR), epidermal growth has UK Clinical Pathology Accreditation. Tumour sections factor receptor (EGFR), vascular endothelial growth factor were cut from TMA blocks for H&E staining to check for (VEGF), and cyclin dependent kinases (CDKs). )e pre- quality. Following this, 4 µm sections were cut, dried clinical and clinical development of these areas has been overnight at 37 C, and stained using an automated immu- reviewed by ourselves and others; however, their incorpo- nostainer (Leica Bond-Max, Milton Keynes, UK). TNBC ration into routine clinical practice to date has been limited status was confirmed as previously described [27]. Epitope [6, 12, 13]. With research into these novel targets ongoing, retrieval solution was applied to the sections for 20 minutes the current reliance is still on cytotoxic chemotherapy, followed by a GR specific monoclonal antibody (Cell Sig- primarily anthracycline-based. )ese treatments have an nalling, #3660) used at a dilution of 1 : 50 for 15 minutes. important role, with many patients deriving benefit from Sections were visualised with diaminobenzidine, counter- their use and achieving long-term survival. We sought to stained with haematoxylin for 5 minutes, and then mounted identify a biomarker for SoC chemotherapy in TNBC to on a Sakura Autostainer. differentiate between good and poor responders which al- lows chemotherapy to be used effectively and alternative treatments to be offered where necessary. 2.3. Assessment. All IHC cores were assessed by at least two In this study, we show that high expression of GR is experienced immunohistochemists blinded to clinical and indicative of patients who will respond well to anthracycline- pathological data. For the IHC analysis, an H score approach based chemotherapy regimens. GR is a steroid hormone to grading the level of GR expression was adopted. )is was receptor expressed ubiquitously in the body and is re- obtained by assigning a cell intensity score for each core on a sponsible for a range of functions, regulating inflammation, scale of 0–3. )is number was multiplied by the relative Journal of Oncology 3 Table 1: Clinical information for the TMAs: (A) TMA #1, (B) TMA significantly associated with improved relapse free survival #2, (C) TMA #3, and (D) TMA #4. (HR 0.27 (CI 0.1118–0.6519) P � 0.0064) (Figure 1(a) (ii)). We next expanded the analysis to a cohort which (A) TMA #1 clinical information consisted of 205 FEC treated patients spanning the mo- Median age (range) 50 (28–74) lecular subgroups of breast cancer as defined by the St. 1 0 Gallen classifications. )ese are IHC based subgroups of Grade 2 6 breast cancer based on ERα, PR, HER2 and Ki67% ex- 3 51 pression and are used as prognostic guides, to accurately Chemotherapy FEC predict disease features and survival [28]. No significant (B) TMA #2 clinical information difference in GR expression was observed between any of the Median age (range) 49 (28–74) 1 0 St. Gallen groups (Figure 1(b)). While no differences in GR Grade 2 7 expression within breast cancer was observed, we next 3 57 wanted to investigate whether GR expression varied between Chemotherapy FEC normal tissue and breast cancer. Using two publicly available (C) TMA #3 clinical information datasets (GSE10797 and GSE9574) [21, 22], there were no Median age (range) 45 (28–76) significant differences in GR expression between normal and 1 0 cancerous tissue, including epithelial and stromal cells 2 6 (Figure 1(c)). Grade 3 105 We went on to investigate whether the observed asso- Not stated 1 ciation between GR expression and improved outcome was CMF 77 specific to TNBC. High GR expression was only associated Chemotherapy FEC-docetaxel 18 with improved survival in the TNBC subset of patients, with Tamoxifen/radiotherapy 17 no significant difference observed in survival when com- (D) TMA #3 clinical information pared to all other subsets either combined (Figure 1(d)) or Median age (range) 54 (36–82) individually (Supplementary Figure 1(a)). 1 0 Grade 2 6 To validate these findings, we identified two independent 3 56 publicly available TNBC datasets with clinical follow-up and FEC 27 available gene expression information (GSE58812 and AC 14 GSE31519) [18, 19]. High GR expression was significantly Chemotherapy TACT-FEC 8 associated with improved metastasis free survival (MFS) None 13 (HR 0.4843 (CI 0.2392-0.9805) P � 0.0478) (Figure 2(a)) and improved event free survival (EFS) (HR 0.4108 (CI 0.1739–0.9703) P � 0.0432) (Figure 2(b)). percentage of positive cells, giving a range of scores from To verify whether GR was prognostic of TNBC pa- 0–300. Cores with significant fragmentation or cores with no thology and disease progression or whether it was a pre- identifiable tumour cells were excluded from analysis. Data dictive marker of response to chemotherapy, we identified from replicate cores were averaged to represent the case. A an untreated cohort of ER negative patients from the summary of assigned scores for each TMA can be found in TRANSBIG study (GSE7390) [20]. In this cohort, high GR Supplementary Table1. expression was associated with significantly poorer relapse free survival (RFS) (HR 2.553 (CI 1.267–5.142) P � 0.0087) 2.4. Survival Analysis and Statistics. Kaplan–Meier analysis and overall survival (OS) (HR 2.615 (CI 1.189–5.751) and hazard ratio calculations were carried out using P � 0.0196) (Figure 2(c)). )is indicates that high GR ex- GraphPad Prism (v8.2.1). Remaining data were analysed pression is indicative of worse overall disease progression using two-tailed t-tests or one-way ANOVA tests as ap- but predicts patients who may respond best to FEC based propriate. Data were deemed significant with a P value chemotherapy. ∗ ∗∗ of< 0.05 signified by and< 0.01 by . While gene expression provides valuable information, IHC examining protein expression fits more readily within the routine diagnostic pipeline. We therefore interrogated 3. Results GR protein expression through IHC analysis on tissue Preceding the present study, an in silico gene expression microarrays (TMAs) stained for GR in tumour samples. In the breast cancer TMAs, a range of expression was analysis was carried out to identify genes associated with good or poor outcome in TNBC [23]. Differentially observed in both the epithelial and stromal compartments expressed genes were identified in an in-house cohort of 30 (Figure 3(a)). At the cellular level, GR was localised to the good outcome (no relapse within 3 years) and 30 poor nucleus, consistent with its role as a nuclear receptor and outcome (relapse within 3 years) patients treated with FEC transcription factor. As expression varied by intensity and (fluorouracil, epirubicin, cyclophosphamide) based che- percentage of cells with expression, an H score was used to motherapy. One of the genes most significantly associated capture GR expression within the epithelial and stromal compartments separately. Following preliminary analysis, with good outcome was NR3C1, encoding GR (P � 0.0028 (Figure 1(a) (i)). Furthermore, expression of GR was only tumour scores were taken further for subsequent 4 Journal of Oncology ∗∗ 8 (i) (ii) 0 0 Good outcome Poor outcome 0 20 40 60 RFS (months) High Low (a) (b) (i) (ii) ns ns 10 4000 6 50 0 20 40 60 RFS (months) High Low (c) (d) Figure 1: (a) (i) Box and whisker plot showing GR expression in good (RFS> 3 years) and poor (RFS< 3 years) outcome patients from the in-house TNBC dataset. (ii) Kaplan–Meier curve of relapse free survival in the in-house TNBC dataset based on GR gene expression above (high) and below (low) the median. (b) Box and whisker plot showing GR expression in each of the St. Gallen subtypes in the in-house dataset. (c) Box and whisker plots showing GR expression in normal versus cancerous breast tissue in the publicly available datasets: (i) GSE10797 and (ii) GSE9574. 1 (d) Kaplan–Meier curve of relapse free survival of the entire in-house cohort dichotomised based on GR expression above (high) and below (low) the median. analyses as GR expression in stromal cells showed little As our discovery and validation datasets were FEC variation with the vast majority exhibiting strong positive treated, we next wanted to look at the relationship between staining. Representative images outlining the scoring GR and outcome in the context of other chemotherapies. We strategy are shown in Figure 3(b). therefore analysed two additional TMAs with a variety of )e influence of GR protein expression on survival was chemotherapies used within the cohorts (TMA #3 and TMA first analysed on a TMA with matched samples to the in- #4, respectively). )e first of these had primarily two regi- house gene expression analysis (TMA #1). Consistent with mens: one consisting of cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) and a second consisting of 5- previous findings, high expression of GR was associated with significantly improved OS (HR 0.2296 (CI 0.06689–0.7882) fluorouracil, epirubicin, cyclophosphamide, and docetaxel P � 0.0194) and an improved RFS which failed to reach (FEC-D). )ese were analysed separately to investigate the significance (Figure 4(a)). A second TNBC cohort was effect of removing the anthracycline (CMF) and adding the identified and scored for GR expression (TMA #2) with taxane (FEC-D) to therapy. In the CMF treated patients, similar results derived. High GR protein expression in tu- there was no association between GR expression and sur- mour cells was found to be associated with improved RFS vival (RFS and OS) (Figure 5(a)). However, in the patients and OS in this cohort (Figure 4(b)). Despite a strong as- with a taxane (docetaxel) added to FEC chemotherapy, high sociation, this did not reach significance likely due to the low GR expression was associated with decreased RFS sample number limiting the statistical power of the analysis. (Figure 5(b)). )is failed to reach significance as there were only 18 patients treated with this regimen, which led to a low Log median-centred intensity Log median-centred intensity Cancer epithelial Cancer stromal Normal epithelial Normal stromal Log median-centred intensity Percent survival Normal breast Breast cancer Percent survival Log median-centred intensity Luminal A Luminal B HER2 pos Luminal B HER2 neg HER2 enriched Triple negative/basal phenotype Journal of Oncology 5 100 100 50 50 0 0 0 2000 4000 6000 0 20 40 60 80 MFS (days) EFS (months) High High Low Low (a) (b) (i) (ii) 100 100 50 50 0 0 0 2000 4000 6000 8000 10000 0 2000 4000 6000 8000 10000 RFS (days) OS (days) High High Low Low (c) Figure 2: Kaplan–Meier curves of (a) metastasis and (b) event-free survival dichotomised based on GR expression above (high) or below (low) the median in the publicly available datasets GSE58812 (N � 107) and GSE31519 (N � 62), respectively. (c) Kaplan–Meier curves of (i) relapse free survival and (ii) overall survival dichotomised based on GR expression above (high) or below (low) the median in the publicly available dataset GSE7390 (N � 64). statistical power. )is finding, however, is interesting given with improved RFS (HR 0.4466 (CI 0.2225–0.8965), the previous link established between high GR expression P � 0.0442) (Supplementary Figure1(b)). and poor survival following taxane treatment [14, 29]. )e Summaries of statistical analyses carried out above are shown in Supplementary Tables 2–6. second TMA used to examine the effect of chemotherapy on the predictive power of GR consisted of 27 FEC and 14 AC (adriamycin, cyclophosphamide) treated patients out of a 4. Discussion total of 56 patients with treatments, also including FEC-D or no chemotherapy. In the total cohort, there is no association )e results of the present study indicate that high gene and/ between GR expression and outcome (Figure 5(c)). How- or protein expression of GR is indicative of patients who will ever, when the analysis is restricted to anthracycline treated respond well to anthracycline-based chemotherapy without patients (FEC/AC), it appears that high GR expression is the use of taxanes. )is appears to be specific to TNBC. again associated with improved RFS and OS (Figure 5(d)). Conversely, high GR expression appears to be indicative of Similarly to the previous analyses, this fails to reach sig- patients who will respond poorly to regimens which include nificance due to the low patient numbers in this cohort. taxanes. However, when the data from TMA cohorts #2 and #3 GR is of particular interest in the context of oncology due consisting of taxane-free, anthracycline treated patients are to the widespread use of GR agonists in the treatment of combined, high GR expression is significantly associated chemotherapy induced nausea and vomiting. Percent survival Percent survival Percent survival Percent survival 6 Journal of Oncology (i) (ii) (i) (ii) (iii) (iv) (iii) (iv) (a) (b) Figure 3: (a) Representative images showing staining patterns of GR in breast cancer tissue, including (i) low tumour, high stroma; (ii) high tumour, low stroma; (iii) high tumour, high stroma; and (iv) low tumour low stroma. (b) Representative images showing scoring strategy including H scores of (i) 0, (ii) 100, (iii) 200, and (iv) 300. All TMA images taken at x20 magnification. (i) (ii) 100 100 50 50 0 0 0 20406080 0 20406080 OS (months) RFS (months) High High Low Low (a) (i) (ii) 100 100 50 50 0 0 0 20406080 0 20406080 OS (months) RFS (months) High High Low Low (b) Figure 4: Kaplan–Meier curves of (i) overall survival and (ii) relapse-free survival stratified by high and low GR IHC expression in (a) TMA #1 (N � 57) and (b) TMA #2 (N � 64). Percent survival Percent survival Percent survival Percent survival Journal of Oncology 7 (i) (ii) 100 100 50 50 0 0 0 100 200 300 400 0 100 200 300 400 OS (months) RFS (months) High High Low Low (a) (i) (ii) 100 100 50 50 0 0 0 20 40 60 0 20 40 60 OS (months) RFS (months) High High Low Low (b) (i) (ii) 100 100 50 50 0 0 0 20 40 60 0 20 40 60 OS (months) RFS (months) High High Low Low (c) (i) 100 (ii) 100 50 50 0 0 0 20 40 60 0 20 40 60 OS (months) RFS (months) High High Low Low (d) Figure 5: Kaplan–Meier curves of (i) overall survival and (ii) relapse-free survival stratified by GR IHC expression in the cohorts: (a) TMA #3 CMF treated (N � 77), (b) TMA #3 taxane treated (N � 18), (c) TMA #4 all chemo (N � 62), and (d) TMA #4 anthracycline treated (N � 41). Percent survival Percent survival Percent survival Percent survival Percent survival Percent survival Percent survival Percent survival 8 Journal of Oncology chemotherapy, DNA damaging, or otherwise. It has been Glucocorticoids such as dexamethasone are prescribed alongside chemotherapy for this purpose. Adherence to revealed that glucocorticoids may induce the production of reactive oxygen species (ROS) in breast cancer cells [36]. these medications is mainly symptom led and varies greatly between patients, although it is required before adminis- ROS can cause DNA damage and could have a synergistic tration of taxane-based treatments [30, 31]. Such agents may, effect when combined with DNA damaging chemother- through GR signalling, be affecting response to chemo- apies such as anthracyclines [37, 38]. Taxanes on the other therapy and therefore outcome in TNBC. )is highlights the hand produce low levels of ROS; thus, no synergy would be clinical relevance of studies such as this. expected [37]. A number of previous studies have made the link be- Another pathway that could be implicated in GR modulating chemotherapy response is the NFκB signal- tween GR signalling and disease progression/outcome/ chemotherapy response in TNBC and other cancers. High ling pathway. GR is a known regulator of NFκB [39], but the literature is conflicting on the nature of this rela- GR has been associated with decreased overall survival in ovarian cancer [32]. Additionally, it has been proposed that tionship. Two studies have demonstrated the beneficial effects of NFκB signalling on the efficacy of chemotherapy glucocorticoids may promote breast cancer metastasis through upregulation of pathways associated with metastasis in vitro and in vivo. One study found that the addition of such as epithelial mesenchymal transition, glucose meta- dexamethasone increased the cytotoxicity of cisplatin in bolism, and epidermal growth factor receptor signalling human cervical carcinoma cell line [40]. )e other [33]. )e role of GR in TNBC has also previously been showed increased antitumour activity of Adriamycin interrogated. Pan et al. have found that expression of GR was (doxorubicin), gemcitabine, and carboplatin against associated with poor outcome in TNBC [29]. )is was breast cancer cells when combined with dexamethasone in mice [41, 42]. Both studies postulated that this activity proposed to be caused by the inhibition of taxane-based chemotherapy induced apoptosis by GR signalling. GR re- was due to GR induced inhibition of NFκB leading to increased apoptosis following chemotherapy. However, sponsive genes such as serum and glucocorticoid inducible protein kinase-1 (SGK1) and mitogen-activated protein data from our research group has suggested that, in the presence of BRCA1 dysfunction, high NFκB signalling kinase phosphatase-1 (MKP1) were implicated in this process [34]. )is relationship appears to contradict our plays a role in improving outcome in TNBC by recruiting findings. However, this can be explained by the fact that our immune cells such as CD8+ cytotoxic T cells, creating an findings show a relationship between high GR expression antitumour microenvironment and preventing progres- and improved survival in the context of anthracycline-based sion [17]. As observed from the present study, context in chemotherapy regimens without the use of taxanes. Such terms of hormone receptor status and chemotherapy is regimens are the first line treatment of choice in TNBC. It key in deducing the role of such pathways, but the effect of GR on NFκB is potentially a key mechanism behind the would appear that the utility of GR as a biomarker is twofold, predicting which patients will respond well to anthracycline- predictive effect of this marker. based chemotherapy regimens, as well as those who will 5. Conclusions respond poorly to taxane-based treatments. We further explored the relationship between GR expression and re- In conclusion, we have identified that the expression of GR is sponse to taxanes in a small number of patients fitting these predictive of TNBC patients that will respond well to criteria, with the results agreeing with the findings of Pan anthracycline-based chemotherapy, which is the current et al. and Wu et al. [29, 34]. SoC in the UK and other countries. Such markers can be )is chemotherapy dependent role of GR can further be easily incorporated into routine IHC-based testing and can inferred from a 2018 study which examined the effect of be used to guide effective treatment choices at an early stage. glucocorticoid use on survival in patients with stages I–III )is also has potential implications for the use of gluco- breast cancer. )e use of glucocorticoids in patients re- corticoids alongside chemotherapy regimens as this could be ceiving no systemic chemotherapy was found to be asso- beneficial with anthracyclines and detrimental with taxanes. ciated with more aggressive clinical features such as higher )e development of such markers is essential in heteroge- histological grade and lymph node involvement [35]. Glu- neous populations such as TNBC to stratify patients into cocorticoid use was found to be associated with smaller clinically relevant populations in order to adopt a person- tumours and less lymph node involvement among anthra- alised approach to disease treatment. For the successful cycline treated patients. Additionally, glucocorticoid use was validation of such biomarkers, large TNBC populations significantly associated with prolonged OS in ERα negative must be analysed for expression of GR and response to patients and shorter OS in ERα positive patients. )e results chemotherapy, both anthracycline, and non-anthracycline- of our study show that hormone receptor status and choice based. Such studies will carry sufficient statistical power to of chemotherapy both influence the role that GR plays as a validate GR as a predictive biomarker. biomarker and its potential use as a treatment target. )ese are consistent with our findings that high GR expression Data Availability predicts good outcome in the context of ERα negative/TNBC and anthracycline-based chemotherapy. )e gene expression datasets analysed in the present study )ere are a number of GR related pathways that could are available from the NCBI repository (https://www.ncbi. explain how signalling could affect response to nlm.nih.gov/gds). All TMA samples are available upon Journal of Oncology 9 [3] R. Dent, M. Trudeau, K. I. Pritchard et al., “Triple-negative application from the Northern Ireland Biobank (http:// breast cancer: clinical features and patterns of recurrence,” www.nibiobank.org/) and the Breast Cancer Now Tissue Clinical Cancer Research, vol. 13, no. 15, 2007. Bank (https://www.breastcancertissuebank.org/). [4] B. D. Lehmann, J. A J. Bauer, X. Chen et al., “Identification of human triple-negative breast cancer subtypes and preclinical Conflicts of Interest models for selection of targeted therapies,” Journal of Clinical Investigation, vol. 121, no. 7, 2011. )e authors declare that they have no conflicts of interest. [5] B. D. Lehmann, B. Jovanovic, X. Chen et al., “Refinement of triple-negative breast cancer molecular subtypes: implications Acknowledgments for neoadjuvant chemotherapy selection,” PLoS One, vol. 11, no. 6, Article ID e0157368, 2016. )e authors acknowledge the role of the Breast Cancer Now [6] A. Elkashif, H. O. Mccarthy, and N. E. Buckley, “Personalising Tissue Bank in collecting and making available the samples the fight against triple negative breast cancer,” Journal of used in the generation of this publication. )is work was Breast Cancer Research and Advancements, vol. 1, no. 1, 2018. funded by Breast Cancer Now (2012MaySF112) and the [7] M. Ono, H. Tsuda, C. Shimizu et al., “Tumor- Department for the Economy (DfE). infiltrating lymphocytes are correlated with response to neoadjuvant chemotherapy in triple-negative breast cancer,” Breast Cancer Research and Treatment, vol. 132, no. 3, 2012. Supplementary Materials [8] P. Vikas, N. Borcherding, and W. Zhang, “)e clinical Supplementary Figure 1: A,Kaplan Meier Curve of relapse promise of immunotherapy in triple-negative breast cancer,” free survival in the in-house dataset based on GR gene Cancer Management and Research, vol. 10, 2018. [9] M. P. Humphries, S. Hynes, V. Bingham et al., “Automated expression above (high) and below (low) the median split tumour recognition and digital pathology scoring unravels into the St. Gallen subgroups, (i) Luminal A (N = 77), (ii) new role for PD-L1 in predicting good outcome in ER-/ Luminal B HER2 positive (N = 61), (iii) Luminal B HER2 HER2+ breast cancer,” Journal of Oncology, vol. 2018, Article negative (N = 30), and (iv) HER2 enriched (N = 37). B, ID 2937012, 14 pages, 2018. Kaplan Meier curve of relapse free survival in anthracycline [10] R. Nanda, E. M. Stringer-Reasor, P. Saha et al., “A randomized treated TNBC patients in TMA #2 and TMA #4 combined phase I trial of nanoparticle albumin-bound paclitaxel with or (N = 105). Supplementary Table 1: log rank hazard ratios, without mifepristone for advanced breast cancer,” Spring- 95% confidence intervals, and P values for survival analysis erplus, vol. 5, no. 1, 2016. of metastasis/event free survival dichotomised based on [11] S. Adams, P. Schmid, H. S. Rugo et al., “Phase 2 study of below (low) or above (high) median gene expression of GR pembrolizumab (pembro) monotherapy for previously in the publicly available datasets, GSE58812 and GSE31519, treated metastatic triple-negative breast cancer (mTNBC): respectively. Supplementary Table 2: log rank hazard ratios, KEYNOTE-086 cohort A,” Journal of Clinical Oncology, 95% confidence intervals, and P values for survival analysis vol. 35, no. 15_suppl, 2017. of metastasis/event free survival dichotomised based on [12] J.-R. Jhan and E. R. Andrechek, “Triple-negative breast cancer below (low) or above (high) median gene expression of GR and the potential for targeted therapy,” Pharmacogenomics, in the publicly available dataset, GSE7390. Supplementary vol. 18, no. 17, 2017. [13] G. Bianchini, J. M. Balko, I. A. Mayer, M. E. Sanders, and Table 3: log rank hazard ratios, 95% confidence intervals, and L. Gianni, “Triple-negative breast cancer: challenges and P values for survival analysis relapse free/overall survival opportunities of a heterogeneous disease,” Nature Reviews stratified by high or low GR IHC expression in (A) TMA #1 Clinical Oncology, vol. 13, no. 11, 2016. and (B) TMA #2. Supplementary Table 4: log rank hazard [14] M. N. Skor, E. L. Wonder, M. Kocherginsky et al., “Gluco- ratios, 95% confidence intervals, and P values for survival corticoid receptor antagonism as a novel therapy for triple- analysis relapse free/overall survival stratified by high or low negative breast cancer,” Clinical Cancer Research, vol. 19, GR IHC expression in (A) CMF treated patients in TMA #3 no. 22, 2013. and (B) taxane treated patients in TMA #3. Supplementary [15] P. Moutsatsou and A. G. Papavassiliou, “Breast Carcinoma: Table 5: log rank hazard ratios, 95% confidence intervals, and the glucocorticoid receptor signalling in breast cancer,” P values for survival analysis relapse free/overall survival Journal of Cellular and Molecular Medicine, vol. 12, no. 1, stratified by high or low GR IHC expression in (A) all patients in TMA #4 and (B) anthracycline treated patients in [16] M. J. Garabedian, C. A. Harris, and F. Jeanneteau, “Gluco- TMA #4. (Supplementary Materials) corticoid receptor action in metabolic and neuronal func- tion,” F1000Research, vol. 6, 2017. [17] N. E. Buckley, P. Haddock, R. De Matos Simoes et al., “A BRCA1 References deficient, NFκB driven immune signal predicts good outcome in triple negative breast cancer,” Oncotarget, vol. 7, no. 15, 2016. [1] F. Bray, J. Ferlay, I. Soerjomataram et al., “Global cancer [18] P. Jez ´ equel, ´ D. Loussouarn, C. Guerin-Charbonnel ´ et al., statistics 2018: GLOBOCAN estimates of incidence and “Gene-expression molecular subtyping of triple-negative mortality worldwide for 36 cancers in 185 countries,” Cancer Journal for Clinicians, vol. 68, no. 6, 2018. breast cancer tumours: importance of immune response,” Breast Cancer Research, vol. 17, 2015. [2] C. Liedtke, C. Mazouni, K. R. Hess et al., “Response to neoadjuvant therapy and long-term survival in patients with [19] A. Rody, T. Karn, C. Liedtke et al., “A clinically relevant gene triple-negative breast cancer,” Journal of Clinical Oncology, signature in triple negative and basal-like breast cancer,” vol. 26, no. 8, 2008. Breast Cancer Research, vol. 13, no. 5, 2011. 10 Journal of Oncology [20] C. Desmedt, F. Piette, S. Loi et al., “Strong time dependence of cancer,” Breast Cancer Research Treatment, vol. 171, no. 1, the 76-gene prognostic signature for node-negative breast cancer patients in the TRANSBIG multicenter independent [36] R. L. Flaherty, M. Owen, A. Fagan-Murphy et al., “Gluco- corticoids induce production of reactive oxygen species/re- validation series,” Clinical Cancer Research, vol. 13, no. 11, 2007. active nitrogen species and DNA damage through an iNOS mediated pathway in breast cancer,” Breast Cancer Research, [21] T. Casey, J. Bond, S. Tighe et al., “Molecular signatures suggest vol. 19, no. 1, 2017. a major role for stromal cells in development of invasive [37] K. A. Conklin, “Chemotherapy-associated oxidative stress: breast cancer,” Breast Cancer Research Treatment, vol. 114, impact on chemotherapeutic effectiveness,” Integrative cancer no. 1, 2009. therapy, vol. 3, 2004. [22] A. Tripathi, C. King, A. de la Morenas et al., “Gene expression [38] D. Gewirtz, “A critical evaluation of the mechanisms of action abnormalities in histologically normal breast epithelium of proposed for the antitumor effects of the anthracycline an- breast cancer patients,” International Journal of Cancer, tibiotics adriamycin and daunorubicin,” Biochemical Phar- vol. 122, no. 7, 2007. macology, vol. 57, no. 7, 1999. [23] P. B. Mullan, V. Bingham, P. Haddock et al., “NUP98–a novel [39] I. M. Adcock, Y. Nasuhara, D. A. Stevens, and P. J. Barnes, predictor of response to anthracycline-based chemotherapy in “Ligand-induced differentiation of glucocorticoid receptor triple negative breast cancer,” BMC Cancer, vol. 19, no. 1, (GR) trans-repression and transactivation: preferential tar- getting of NF-κB and lack of I-κB involvement,” British [24] K. Orr, N. E. Buckley, P. Haddock et al., “)romboxane A2 Journal of Pharmacology, vol. 127, no. 4, 1999. receptor (TBXA2R) is a potent survival factor for triple [40] Y.-S. Lu, P.-Y. Yeh, S.-E. Chuang et al., “Glucocorticoids negative breast cancers (TNBCs),” Oncotarget, vol. 7, no. 34, enhance cytotoxicity of cisplatin via suppression of NF-B activation in the glucocorticoid receptor-rich human cervical [25] R. D. A. Wilkinson, R. E. Burden, S. H. McDowell et al., “A carcinoma cell line SiHa,” Journal of Endocrinology, vol. 188, novel role for cathepsin S as a potential biomarker in triple no. 2, 2006. negative breast cancer,” Journal of Oncology, vol. 2019, Article [41] H. Wang, Y. Wang, E. R. Rayburn et al., “Dexamethasone as a ID 3980273, 12 pages, 2019. chemosensitizer for breast cancer chemotherapy: potentiation [26] C. Knowlson, P. Haddock, V. Bingham et al., “Pin1 plays a key of the antitumor activity of adriamycin, modulation of cy- role in the response to treatment and clinical outcome in triple tokine expression, and pharmacokinetics,” International negative breast cancer,” =erapeutic Advances in Medical Journal Oncology, vol. 30, no. 4, 2007. Oncology, vol. 12, 2020. [42] H. Wang, M. Li, J. J. Rinehart, and R. Zhang, “Pretreatment [27] D. P. Boyle, D. G. McArt, G. Irwin et al., “)e prognostic with dexamethasone increases antitumor activity of carbo- significance of the aberrant extremes of p53 immunophe- platin and gemcitabine in mice bearing human cancer xe- notypes in breast cancer,” Histopathology, vol. 65, no. 3, 2014. nografts: in vivo activity, pharmacokinetics, and clinical [28] I. Vasconcelos, A. Hussainzada, S. Berger et al., “)e St. Gallen implications for cancer chemotherapy,” Clinical Cancer Re- surrogate classification for breast cancer subtypes successfully search, vol. 10, no. 5, 2004. predicts tumor presenting features, nodal involvement, re- currence patterns and disease free survival,” =e Breast, vol. 29, 2016. [29] D. Pan, M. Kocherginsky, and S. D. Conzen, “Activation of the glucocorticoid receptor is associated with poor prognosis in estrogen receptor-negative breast cancer,” Cancer Research, vol. 71, no. 20, pp. 6360–6370, 2011. [30] S. Almazrou and L. Alnaim, “Evaluation of adherence to chemotherapy-induced nausea and vomiting guidelines. An observational study,” Journal of Cancer =erapy, vol. 3, no. 5, pp. 613–620, 2012. [31] H. Fujii, H. Iihara, M. Ishihara et al., “Improvement of ad- herence to guidelines for antiemetic medication enhances emetic control in patients with colorectal cancer receiving chemotherapy of moderate emetic risk,” Anticancer Research, vol. 33, no. 12, 2013. [32] J. T. Veneris, L. Huang, J. E. Churpek, S. D. Conzen, and G. F. Fleming, “Glucocorticoid receptor expression is asso- ciated with inferior overall survival independent of BRCA mutation status in ovarian cancer,” International Journal of Gynecologic Cancer, vol. 29, no. 2, pp. 357–364, 2019. [33] M. M. S. Obradovic, B. Hamelin, N. Manevski et al., “Glu- cocorticoids promote breast cancer metastasis,” Nature, vol. 567, no. 7749, pp. 540–544, 2019. [34] W. Wu, S. Chaudhuri, D. R. Brickley et al., “Microarray analysis reveals glucocorticoid-regulated survival genes that are associated with inhibition of apoptosis in breast epithelial cells,” Cancer Research, vol. 64, no. 5, 2004. [35] C.-H. Lin, P.-Y. Chuang, S.-L. You et al., “Effect of gluco- corticoid use on survival in patients with stage I–III breast http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

Glucocorticoid Receptor Expression Predicts Good Outcome in response to Taxane-Free, Anthracycline-Based Therapy in Triple Negative Breast Cancer

Download PDF
10 pages

Loading next page...
 
/lp/hindawi-publishing-corporation/glucocorticoid-receptor-expression-predicts-good-outcome-in-response-Xfs7dwwfLa

References (42)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Ahmed Elkashif et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2020/3712825
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Oncology Volume 2020, Article ID 3712825, 10 pages https://doi.org/10.1155/2020/3712825 Research Article Glucocorticoid Receptor Expression Predicts Good Outcome in response to Taxane-Free, Anthracycline-Based Therapy in Triple Negative Breast Cancer 1 2 1 2 Ahmed Elkashif , Victoria Bingham , Paula Haddock, Matthew P. Humphries , 2 2 1 1 Stephen McQuaid , Paul B. Mullan , Helen O. McCarthy , and Niamh E. Buckley School of Pharmacy, Queen’s University Belfast, Belfast BT9 7BL, UK Patrick G Johnston Centre for Cancer Research, Queen’s University Belfast, Belfast BT9 7AE, UK Correspondence should be addressed to Niamh E. Buckley; n.obrien@qub.ac.uk Received 30 January 2020; Accepted 16 April 2020; Published 20 May 2020 Academic Editor: Hakan Buyukhatipoglu Copyright © 2020 Ahmed Elkashif et al. )is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Triple negative breast cancer (TNBC) is a poor outcome subset of breast cancers characterised by the lack of expression of ER α, PR, and HER2 amplification. It is a heterogeneous group of cancers which fail to derive benefit from modern, more targeted treatments such as Tamoxifen and Herceptin. Current standard of care (SoC) is cytotoxic chemotherapy, which is effective for some patients, with other patients deriving little/no benefit and lacking alternative treatments. )is study has identified the glucocorticoid receptor (GR) as a potential predictive biomarker of response to anthracycline-based chemotherapy in triple negative breast cancer (TNBC). GR gene expression levels in patient samples were analysed through publicly available microarray datasets as well as protein expression through immunohistochemistry (IHC) and correlated with clinical/pathological outcomes, including survival. While the results confirmed previous observations that high GR expression is associated with poor outcome in response to taxane-based chemotherapy, this study shows for the first time that high GR expression is associated with improved outcomes in the context of anthracycline-based chemotherapy. GR therefore has the potential to be used as a predictive biomarker to guide treatment choices and ensure that patients derive the greatest benefit from first line treatment, avoiding unnecessary costs, side effects, and disease progression. amplification of the HER2 gene. )is subset is associated 1. Introduction with the poorest outcomes among breast cancers, with Breast cancer is the most common malignancy in females higher mortality rates compared to other subtypes, despite worldwide, with over 2 million new cases diagnosed in 2018 accounting for only 15–20% of cases [2]. )is is due, in part, and an estimated 600,000 deaths [1]. Following diagnosis of to the lack of molecular targets available for treatment. breast cancer, patients are stratified based on expression of Given the lack of biomarkers to guide treatment, first line estrogen receptor alpha (ERα), progesterone receptor (PR), therapy for TNBC is a regimen of adjuvant and increasingly and amplification of the human epidermal growth factor neoadjuvant, cytotoxic chemotherapy consisting of receptor 2 (HER2) gene. )is is used to guide treatments, anthracyclines and taxanes. Despite the high rates of mor- with targeted treatments including Tamoxifen and Trastu- tality, TNBC has higher response rates to chemotherapy zumab (Herceptin) used to treat ERα positive and HER2 compared to other breast cancers. )is is commonly referred positive cancers, respectively. to as the “TNBC paradox”. In the neoadjuvant setting, Triple negative breast cancer (TNBC) is a term used to patients who achieve a complete pathological response describe breast cancers which are ERα/PR negative and lack (pCR) tend to have a very good prognosis and survival rates 2 Journal of Oncology response to stress, and cell survival. It exists in the cell comparable to non-TNBC, accounting for around a third of patients [2]. )ose who do not show a response to first line cytoplasm, bound to heat shock proteins (HSPs). Its en- dogenous ligand is cortisol and upon ligand binding, GR chemotherapy or have residual disease (RD) following treatment tend to relapse in the first three years following dissociates from HSPs, dimerises, and translocates to the diagnosis and have a high risk of disease progression and nucleus. It binds to glucocorticoid response elements death [2, 3]. )is highlights the importance of biomarkers to (GREs) on DNA to transactivate or transrepress genes. GR is stratify patients and tailor treatment options accordingly involved in transrepression of genes such as those respon- and that an optimal response to first line treatment is sible for inflammation (e.g., AP1 and NFκB) while trans- paramount in assuring long term survival. Given the fact activating other genes such as those responsible for cell survival and DNA repair (e.g., STAT1 and Ets) [14–16]. that TNBC is a diagnosis of exclusion (based on the lack if receptors, as opposed to the presence), this variable response )is study implicates GR as a potential prognostic biomarker in the context of SoC which could be used to therapy is not surprising. Without the use of additional biomarkers, such diagnostic criteria leave a highly hetero- clinically to guide treatment choices and tailor chemo- therapy to individual patients. geneous population with variable treatment response rates that in theory should be able to be further stratified. )is is exemplified by the further subclassification of TNBC into the 2. Materials and Methods Vanderbilt subtypes by Lehmann et al. consisting of four 2.1. Gene Expression. )e in-house gene expression dataset subgroups, each with distinctive gene expression patterns, has been previously described [17]. )e publicly available responses to chemotherapy, and overall outcomes [4, 5]. datasets GSE5881 [18], GSE31519 [19], GSE7390 [20], )e development of personalised medicine approaches GSE10797 [21], and GSE9574 [22] were accessed online in TNBC is an area of increasing interest with the identi- using NCBI. fication and development of novel targeted treatments and/ or combinations as well as the associated companion bio- markers [6]. A major advancement in this field has been the 2.2. Tissue Microarray (TMA) and Immunohistochemistry development of immunotherapy and the use of immune (IHC). )e TMAs described in this study were constructed checkpoint inhibitors. TNBC is characterised by a high from formalin fixed paraffin embedded (FFPE) slides pre- expression of immune checkpoint proteins such as PD-L1 as pared by the Northern Ireland Biobank (NIB13-0043) and well as a degree of tumour infiltrating lymphocytes (TILs), the Breast Cancer Now Tissue Bank (TR-0055). Both the both of which are predictive of response to immunotherapy NIB (REC:11/NI/0013) and the Breast Cancer Now Tissue as well as conventional chemotherapies such as anthracy- Bank have ethical approval to use deidentified tissue samples clines [6–9]. Clinical trials have shown that response rates to from NHS tissue pathology archives (specifically the Belfast pembrolizumab, a monoclonal antibody targeting PD-1, are Health and Social Care Trust Cellular Pathology archive in increased from 5% to 18.5% by assessing PD-L1 status and the case of the NIB). )e first TMA with matched samples to only treating patients expressing this marker [10, 11]. Ad- the in-house gene expression cohort has been previously ditionally, the KEYNOTE-355 phase III clinical trial has described elsewhere [17, 23]. )e subsequent TMAs with shown the addition of pembrolizumab to chemotherapy various chemotherapy regimens have also been previously improves outcome in metastatic TNBC with the primary end described [9, 24–26]. Clinical information from these TMAs point of progression-free survival being met is summarised in Table 1. IHC - was performed in the (NCT02819518). Other targets being explored in TNBC Northern Ireland Molecular Pathology Laboratory, which include the androgen receptor (AR), epidermal growth has UK Clinical Pathology Accreditation. Tumour sections factor receptor (EGFR), vascular endothelial growth factor were cut from TMA blocks for H&E staining to check for (VEGF), and cyclin dependent kinases (CDKs). )e pre- quality. Following this, 4 µm sections were cut, dried clinical and clinical development of these areas has been overnight at 37 C, and stained using an automated immu- reviewed by ourselves and others; however, their incorpo- nostainer (Leica Bond-Max, Milton Keynes, UK). TNBC ration into routine clinical practice to date has been limited status was confirmed as previously described [27]. Epitope [6, 12, 13]. With research into these novel targets ongoing, retrieval solution was applied to the sections for 20 minutes the current reliance is still on cytotoxic chemotherapy, followed by a GR specific monoclonal antibody (Cell Sig- primarily anthracycline-based. )ese treatments have an nalling, #3660) used at a dilution of 1 : 50 for 15 minutes. important role, with many patients deriving benefit from Sections were visualised with diaminobenzidine, counter- their use and achieving long-term survival. We sought to stained with haematoxylin for 5 minutes, and then mounted identify a biomarker for SoC chemotherapy in TNBC to on a Sakura Autostainer. differentiate between good and poor responders which al- lows chemotherapy to be used effectively and alternative treatments to be offered where necessary. 2.3. Assessment. All IHC cores were assessed by at least two In this study, we show that high expression of GR is experienced immunohistochemists blinded to clinical and indicative of patients who will respond well to anthracycline- pathological data. For the IHC analysis, an H score approach based chemotherapy regimens. GR is a steroid hormone to grading the level of GR expression was adopted. )is was receptor expressed ubiquitously in the body and is re- obtained by assigning a cell intensity score for each core on a sponsible for a range of functions, regulating inflammation, scale of 0–3. )is number was multiplied by the relative Journal of Oncology 3 Table 1: Clinical information for the TMAs: (A) TMA #1, (B) TMA significantly associated with improved relapse free survival #2, (C) TMA #3, and (D) TMA #4. (HR 0.27 (CI 0.1118–0.6519) P � 0.0064) (Figure 1(a) (ii)). We next expanded the analysis to a cohort which (A) TMA #1 clinical information consisted of 205 FEC treated patients spanning the mo- Median age (range) 50 (28–74) lecular subgroups of breast cancer as defined by the St. 1 0 Gallen classifications. )ese are IHC based subgroups of Grade 2 6 breast cancer based on ERα, PR, HER2 and Ki67% ex- 3 51 pression and are used as prognostic guides, to accurately Chemotherapy FEC predict disease features and survival [28]. No significant (B) TMA #2 clinical information difference in GR expression was observed between any of the Median age (range) 49 (28–74) 1 0 St. Gallen groups (Figure 1(b)). While no differences in GR Grade 2 7 expression within breast cancer was observed, we next 3 57 wanted to investigate whether GR expression varied between Chemotherapy FEC normal tissue and breast cancer. Using two publicly available (C) TMA #3 clinical information datasets (GSE10797 and GSE9574) [21, 22], there were no Median age (range) 45 (28–76) significant differences in GR expression between normal and 1 0 cancerous tissue, including epithelial and stromal cells 2 6 (Figure 1(c)). Grade 3 105 We went on to investigate whether the observed asso- Not stated 1 ciation between GR expression and improved outcome was CMF 77 specific to TNBC. High GR expression was only associated Chemotherapy FEC-docetaxel 18 with improved survival in the TNBC subset of patients, with Tamoxifen/radiotherapy 17 no significant difference observed in survival when com- (D) TMA #3 clinical information pared to all other subsets either combined (Figure 1(d)) or Median age (range) 54 (36–82) individually (Supplementary Figure 1(a)). 1 0 Grade 2 6 To validate these findings, we identified two independent 3 56 publicly available TNBC datasets with clinical follow-up and FEC 27 available gene expression information (GSE58812 and AC 14 GSE31519) [18, 19]. High GR expression was significantly Chemotherapy TACT-FEC 8 associated with improved metastasis free survival (MFS) None 13 (HR 0.4843 (CI 0.2392-0.9805) P � 0.0478) (Figure 2(a)) and improved event free survival (EFS) (HR 0.4108 (CI 0.1739–0.9703) P � 0.0432) (Figure 2(b)). percentage of positive cells, giving a range of scores from To verify whether GR was prognostic of TNBC pa- 0–300. Cores with significant fragmentation or cores with no thology and disease progression or whether it was a pre- identifiable tumour cells were excluded from analysis. Data dictive marker of response to chemotherapy, we identified from replicate cores were averaged to represent the case. A an untreated cohort of ER negative patients from the summary of assigned scores for each TMA can be found in TRANSBIG study (GSE7390) [20]. In this cohort, high GR Supplementary Table1. expression was associated with significantly poorer relapse free survival (RFS) (HR 2.553 (CI 1.267–5.142) P � 0.0087) 2.4. Survival Analysis and Statistics. Kaplan–Meier analysis and overall survival (OS) (HR 2.615 (CI 1.189–5.751) and hazard ratio calculations were carried out using P � 0.0196) (Figure 2(c)). )is indicates that high GR ex- GraphPad Prism (v8.2.1). Remaining data were analysed pression is indicative of worse overall disease progression using two-tailed t-tests or one-way ANOVA tests as ap- but predicts patients who may respond best to FEC based propriate. Data were deemed significant with a P value chemotherapy. ∗ ∗∗ of< 0.05 signified by and< 0.01 by . While gene expression provides valuable information, IHC examining protein expression fits more readily within the routine diagnostic pipeline. We therefore interrogated 3. Results GR protein expression through IHC analysis on tissue Preceding the present study, an in silico gene expression microarrays (TMAs) stained for GR in tumour samples. In the breast cancer TMAs, a range of expression was analysis was carried out to identify genes associated with good or poor outcome in TNBC [23]. Differentially observed in both the epithelial and stromal compartments expressed genes were identified in an in-house cohort of 30 (Figure 3(a)). At the cellular level, GR was localised to the good outcome (no relapse within 3 years) and 30 poor nucleus, consistent with its role as a nuclear receptor and outcome (relapse within 3 years) patients treated with FEC transcription factor. As expression varied by intensity and (fluorouracil, epirubicin, cyclophosphamide) based che- percentage of cells with expression, an H score was used to motherapy. One of the genes most significantly associated capture GR expression within the epithelial and stromal compartments separately. Following preliminary analysis, with good outcome was NR3C1, encoding GR (P � 0.0028 (Figure 1(a) (i)). Furthermore, expression of GR was only tumour scores were taken further for subsequent 4 Journal of Oncology ∗∗ 8 (i) (ii) 0 0 Good outcome Poor outcome 0 20 40 60 RFS (months) High Low (a) (b) (i) (ii) ns ns 10 4000 6 50 0 20 40 60 RFS (months) High Low (c) (d) Figure 1: (a) (i) Box and whisker plot showing GR expression in good (RFS> 3 years) and poor (RFS< 3 years) outcome patients from the in-house TNBC dataset. (ii) Kaplan–Meier curve of relapse free survival in the in-house TNBC dataset based on GR gene expression above (high) and below (low) the median. (b) Box and whisker plot showing GR expression in each of the St. Gallen subtypes in the in-house dataset. (c) Box and whisker plots showing GR expression in normal versus cancerous breast tissue in the publicly available datasets: (i) GSE10797 and (ii) GSE9574. 1 (d) Kaplan–Meier curve of relapse free survival of the entire in-house cohort dichotomised based on GR expression above (high) and below (low) the median. analyses as GR expression in stromal cells showed little As our discovery and validation datasets were FEC variation with the vast majority exhibiting strong positive treated, we next wanted to look at the relationship between staining. Representative images outlining the scoring GR and outcome in the context of other chemotherapies. We strategy are shown in Figure 3(b). therefore analysed two additional TMAs with a variety of )e influence of GR protein expression on survival was chemotherapies used within the cohorts (TMA #3 and TMA first analysed on a TMA with matched samples to the in- #4, respectively). )e first of these had primarily two regi- house gene expression analysis (TMA #1). Consistent with mens: one consisting of cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) and a second consisting of 5- previous findings, high expression of GR was associated with significantly improved OS (HR 0.2296 (CI 0.06689–0.7882) fluorouracil, epirubicin, cyclophosphamide, and docetaxel P � 0.0194) and an improved RFS which failed to reach (FEC-D). )ese were analysed separately to investigate the significance (Figure 4(a)). A second TNBC cohort was effect of removing the anthracycline (CMF) and adding the identified and scored for GR expression (TMA #2) with taxane (FEC-D) to therapy. In the CMF treated patients, similar results derived. High GR protein expression in tu- there was no association between GR expression and sur- mour cells was found to be associated with improved RFS vival (RFS and OS) (Figure 5(a)). However, in the patients and OS in this cohort (Figure 4(b)). Despite a strong as- with a taxane (docetaxel) added to FEC chemotherapy, high sociation, this did not reach significance likely due to the low GR expression was associated with decreased RFS sample number limiting the statistical power of the analysis. (Figure 5(b)). )is failed to reach significance as there were only 18 patients treated with this regimen, which led to a low Log median-centred intensity Log median-centred intensity Cancer epithelial Cancer stromal Normal epithelial Normal stromal Log median-centred intensity Percent survival Normal breast Breast cancer Percent survival Log median-centred intensity Luminal A Luminal B HER2 pos Luminal B HER2 neg HER2 enriched Triple negative/basal phenotype Journal of Oncology 5 100 100 50 50 0 0 0 2000 4000 6000 0 20 40 60 80 MFS (days) EFS (months) High High Low Low (a) (b) (i) (ii) 100 100 50 50 0 0 0 2000 4000 6000 8000 10000 0 2000 4000 6000 8000 10000 RFS (days) OS (days) High High Low Low (c) Figure 2: Kaplan–Meier curves of (a) metastasis and (b) event-free survival dichotomised based on GR expression above (high) or below (low) the median in the publicly available datasets GSE58812 (N � 107) and GSE31519 (N � 62), respectively. (c) Kaplan–Meier curves of (i) relapse free survival and (ii) overall survival dichotomised based on GR expression above (high) or below (low) the median in the publicly available dataset GSE7390 (N � 64). statistical power. )is finding, however, is interesting given with improved RFS (HR 0.4466 (CI 0.2225–0.8965), the previous link established between high GR expression P � 0.0442) (Supplementary Figure1(b)). and poor survival following taxane treatment [14, 29]. )e Summaries of statistical analyses carried out above are shown in Supplementary Tables 2–6. second TMA used to examine the effect of chemotherapy on the predictive power of GR consisted of 27 FEC and 14 AC (adriamycin, cyclophosphamide) treated patients out of a 4. Discussion total of 56 patients with treatments, also including FEC-D or no chemotherapy. In the total cohort, there is no association )e results of the present study indicate that high gene and/ between GR expression and outcome (Figure 5(c)). How- or protein expression of GR is indicative of patients who will ever, when the analysis is restricted to anthracycline treated respond well to anthracycline-based chemotherapy without patients (FEC/AC), it appears that high GR expression is the use of taxanes. )is appears to be specific to TNBC. again associated with improved RFS and OS (Figure 5(d)). Conversely, high GR expression appears to be indicative of Similarly to the previous analyses, this fails to reach sig- patients who will respond poorly to regimens which include nificance due to the low patient numbers in this cohort. taxanes. However, when the data from TMA cohorts #2 and #3 GR is of particular interest in the context of oncology due consisting of taxane-free, anthracycline treated patients are to the widespread use of GR agonists in the treatment of combined, high GR expression is significantly associated chemotherapy induced nausea and vomiting. Percent survival Percent survival Percent survival Percent survival 6 Journal of Oncology (i) (ii) (i) (ii) (iii) (iv) (iii) (iv) (a) (b) Figure 3: (a) Representative images showing staining patterns of GR in breast cancer tissue, including (i) low tumour, high stroma; (ii) high tumour, low stroma; (iii) high tumour, high stroma; and (iv) low tumour low stroma. (b) Representative images showing scoring strategy including H scores of (i) 0, (ii) 100, (iii) 200, and (iv) 300. All TMA images taken at x20 magnification. (i) (ii) 100 100 50 50 0 0 0 20406080 0 20406080 OS (months) RFS (months) High High Low Low (a) (i) (ii) 100 100 50 50 0 0 0 20406080 0 20406080 OS (months) RFS (months) High High Low Low (b) Figure 4: Kaplan–Meier curves of (i) overall survival and (ii) relapse-free survival stratified by high and low GR IHC expression in (a) TMA #1 (N � 57) and (b) TMA #2 (N � 64). Percent survival Percent survival Percent survival Percent survival Journal of Oncology 7 (i) (ii) 100 100 50 50 0 0 0 100 200 300 400 0 100 200 300 400 OS (months) RFS (months) High High Low Low (a) (i) (ii) 100 100 50 50 0 0 0 20 40 60 0 20 40 60 OS (months) RFS (months) High High Low Low (b) (i) (ii) 100 100 50 50 0 0 0 20 40 60 0 20 40 60 OS (months) RFS (months) High High Low Low (c) (i) 100 (ii) 100 50 50 0 0 0 20 40 60 0 20 40 60 OS (months) RFS (months) High High Low Low (d) Figure 5: Kaplan–Meier curves of (i) overall survival and (ii) relapse-free survival stratified by GR IHC expression in the cohorts: (a) TMA #3 CMF treated (N � 77), (b) TMA #3 taxane treated (N � 18), (c) TMA #4 all chemo (N � 62), and (d) TMA #4 anthracycline treated (N � 41). Percent survival Percent survival Percent survival Percent survival Percent survival Percent survival Percent survival Percent survival 8 Journal of Oncology chemotherapy, DNA damaging, or otherwise. It has been Glucocorticoids such as dexamethasone are prescribed alongside chemotherapy for this purpose. Adherence to revealed that glucocorticoids may induce the production of reactive oxygen species (ROS) in breast cancer cells [36]. these medications is mainly symptom led and varies greatly between patients, although it is required before adminis- ROS can cause DNA damage and could have a synergistic tration of taxane-based treatments [30, 31]. Such agents may, effect when combined with DNA damaging chemother- through GR signalling, be affecting response to chemo- apies such as anthracyclines [37, 38]. Taxanes on the other therapy and therefore outcome in TNBC. )is highlights the hand produce low levels of ROS; thus, no synergy would be clinical relevance of studies such as this. expected [37]. A number of previous studies have made the link be- Another pathway that could be implicated in GR modulating chemotherapy response is the NFκB signal- tween GR signalling and disease progression/outcome/ chemotherapy response in TNBC and other cancers. High ling pathway. GR is a known regulator of NFκB [39], but the literature is conflicting on the nature of this rela- GR has been associated with decreased overall survival in ovarian cancer [32]. Additionally, it has been proposed that tionship. Two studies have demonstrated the beneficial effects of NFκB signalling on the efficacy of chemotherapy glucocorticoids may promote breast cancer metastasis through upregulation of pathways associated with metastasis in vitro and in vivo. One study found that the addition of such as epithelial mesenchymal transition, glucose meta- dexamethasone increased the cytotoxicity of cisplatin in bolism, and epidermal growth factor receptor signalling human cervical carcinoma cell line [40]. )e other [33]. )e role of GR in TNBC has also previously been showed increased antitumour activity of Adriamycin interrogated. Pan et al. have found that expression of GR was (doxorubicin), gemcitabine, and carboplatin against associated with poor outcome in TNBC [29]. )is was breast cancer cells when combined with dexamethasone in mice [41, 42]. Both studies postulated that this activity proposed to be caused by the inhibition of taxane-based chemotherapy induced apoptosis by GR signalling. GR re- was due to GR induced inhibition of NFκB leading to increased apoptosis following chemotherapy. However, sponsive genes such as serum and glucocorticoid inducible protein kinase-1 (SGK1) and mitogen-activated protein data from our research group has suggested that, in the presence of BRCA1 dysfunction, high NFκB signalling kinase phosphatase-1 (MKP1) were implicated in this process [34]. )is relationship appears to contradict our plays a role in improving outcome in TNBC by recruiting findings. However, this can be explained by the fact that our immune cells such as CD8+ cytotoxic T cells, creating an findings show a relationship between high GR expression antitumour microenvironment and preventing progres- and improved survival in the context of anthracycline-based sion [17]. As observed from the present study, context in chemotherapy regimens without the use of taxanes. Such terms of hormone receptor status and chemotherapy is regimens are the first line treatment of choice in TNBC. It key in deducing the role of such pathways, but the effect of GR on NFκB is potentially a key mechanism behind the would appear that the utility of GR as a biomarker is twofold, predicting which patients will respond well to anthracycline- predictive effect of this marker. based chemotherapy regimens, as well as those who will 5. Conclusions respond poorly to taxane-based treatments. We further explored the relationship between GR expression and re- In conclusion, we have identified that the expression of GR is sponse to taxanes in a small number of patients fitting these predictive of TNBC patients that will respond well to criteria, with the results agreeing with the findings of Pan anthracycline-based chemotherapy, which is the current et al. and Wu et al. [29, 34]. SoC in the UK and other countries. Such markers can be )is chemotherapy dependent role of GR can further be easily incorporated into routine IHC-based testing and can inferred from a 2018 study which examined the effect of be used to guide effective treatment choices at an early stage. glucocorticoid use on survival in patients with stages I–III )is also has potential implications for the use of gluco- breast cancer. )e use of glucocorticoids in patients re- corticoids alongside chemotherapy regimens as this could be ceiving no systemic chemotherapy was found to be asso- beneficial with anthracyclines and detrimental with taxanes. ciated with more aggressive clinical features such as higher )e development of such markers is essential in heteroge- histological grade and lymph node involvement [35]. Glu- neous populations such as TNBC to stratify patients into cocorticoid use was found to be associated with smaller clinically relevant populations in order to adopt a person- tumours and less lymph node involvement among anthra- alised approach to disease treatment. For the successful cycline treated patients. Additionally, glucocorticoid use was validation of such biomarkers, large TNBC populations significantly associated with prolonged OS in ERα negative must be analysed for expression of GR and response to patients and shorter OS in ERα positive patients. )e results chemotherapy, both anthracycline, and non-anthracycline- of our study show that hormone receptor status and choice based. Such studies will carry sufficient statistical power to of chemotherapy both influence the role that GR plays as a validate GR as a predictive biomarker. biomarker and its potential use as a treatment target. )ese are consistent with our findings that high GR expression Data Availability predicts good outcome in the context of ERα negative/TNBC and anthracycline-based chemotherapy. )e gene expression datasets analysed in the present study )ere are a number of GR related pathways that could are available from the NCBI repository (https://www.ncbi. explain how signalling could affect response to nlm.nih.gov/gds). All TMA samples are available upon Journal of Oncology 9 [3] R. Dent, M. Trudeau, K. I. Pritchard et al., “Triple-negative application from the Northern Ireland Biobank (http:// breast cancer: clinical features and patterns of recurrence,” www.nibiobank.org/) and the Breast Cancer Now Tissue Clinical Cancer Research, vol. 13, no. 15, 2007. Bank (https://www.breastcancertissuebank.org/). [4] B. D. Lehmann, J. A J. Bauer, X. Chen et al., “Identification of human triple-negative breast cancer subtypes and preclinical Conflicts of Interest models for selection of targeted therapies,” Journal of Clinical Investigation, vol. 121, no. 7, 2011. )e authors declare that they have no conflicts of interest. [5] B. D. Lehmann, B. Jovanovic, X. Chen et al., “Refinement of triple-negative breast cancer molecular subtypes: implications Acknowledgments for neoadjuvant chemotherapy selection,” PLoS One, vol. 11, no. 6, Article ID e0157368, 2016. )e authors acknowledge the role of the Breast Cancer Now [6] A. Elkashif, H. O. Mccarthy, and N. E. Buckley, “Personalising Tissue Bank in collecting and making available the samples the fight against triple negative breast cancer,” Journal of used in the generation of this publication. )is work was Breast Cancer Research and Advancements, vol. 1, no. 1, 2018. funded by Breast Cancer Now (2012MaySF112) and the [7] M. Ono, H. Tsuda, C. Shimizu et al., “Tumor- Department for the Economy (DfE). infiltrating lymphocytes are correlated with response to neoadjuvant chemotherapy in triple-negative breast cancer,” Breast Cancer Research and Treatment, vol. 132, no. 3, 2012. Supplementary Materials [8] P. Vikas, N. Borcherding, and W. Zhang, “)e clinical Supplementary Figure 1: A,Kaplan Meier Curve of relapse promise of immunotherapy in triple-negative breast cancer,” free survival in the in-house dataset based on GR gene Cancer Management and Research, vol. 10, 2018. [9] M. P. Humphries, S. Hynes, V. Bingham et al., “Automated expression above (high) and below (low) the median split tumour recognition and digital pathology scoring unravels into the St. Gallen subgroups, (i) Luminal A (N = 77), (ii) new role for PD-L1 in predicting good outcome in ER-/ Luminal B HER2 positive (N = 61), (iii) Luminal B HER2 HER2+ breast cancer,” Journal of Oncology, vol. 2018, Article negative (N = 30), and (iv) HER2 enriched (N = 37). B, ID 2937012, 14 pages, 2018. Kaplan Meier curve of relapse free survival in anthracycline [10] R. Nanda, E. M. Stringer-Reasor, P. Saha et al., “A randomized treated TNBC patients in TMA #2 and TMA #4 combined phase I trial of nanoparticle albumin-bound paclitaxel with or (N = 105). Supplementary Table 1: log rank hazard ratios, without mifepristone for advanced breast cancer,” Spring- 95% confidence intervals, and P values for survival analysis erplus, vol. 5, no. 1, 2016. of metastasis/event free survival dichotomised based on [11] S. Adams, P. Schmid, H. S. Rugo et al., “Phase 2 study of below (low) or above (high) median gene expression of GR pembrolizumab (pembro) monotherapy for previously in the publicly available datasets, GSE58812 and GSE31519, treated metastatic triple-negative breast cancer (mTNBC): respectively. Supplementary Table 2: log rank hazard ratios, KEYNOTE-086 cohort A,” Journal of Clinical Oncology, 95% confidence intervals, and P values for survival analysis vol. 35, no. 15_suppl, 2017. of metastasis/event free survival dichotomised based on [12] J.-R. Jhan and E. R. Andrechek, “Triple-negative breast cancer below (low) or above (high) median gene expression of GR and the potential for targeted therapy,” Pharmacogenomics, in the publicly available dataset, GSE7390. Supplementary vol. 18, no. 17, 2017. [13] G. Bianchini, J. M. Balko, I. A. Mayer, M. E. Sanders, and Table 3: log rank hazard ratios, 95% confidence intervals, and L. Gianni, “Triple-negative breast cancer: challenges and P values for survival analysis relapse free/overall survival opportunities of a heterogeneous disease,” Nature Reviews stratified by high or low GR IHC expression in (A) TMA #1 Clinical Oncology, vol. 13, no. 11, 2016. and (B) TMA #2. Supplementary Table 4: log rank hazard [14] M. N. Skor, E. L. Wonder, M. Kocherginsky et al., “Gluco- ratios, 95% confidence intervals, and P values for survival corticoid receptor antagonism as a novel therapy for triple- analysis relapse free/overall survival stratified by high or low negative breast cancer,” Clinical Cancer Research, vol. 19, GR IHC expression in (A) CMF treated patients in TMA #3 no. 22, 2013. and (B) taxane treated patients in TMA #3. Supplementary [15] P. Moutsatsou and A. G. Papavassiliou, “Breast Carcinoma: Table 5: log rank hazard ratios, 95% confidence intervals, and the glucocorticoid receptor signalling in breast cancer,” P values for survival analysis relapse free/overall survival Journal of Cellular and Molecular Medicine, vol. 12, no. 1, stratified by high or low GR IHC expression in (A) all patients in TMA #4 and (B) anthracycline treated patients in [16] M. J. Garabedian, C. A. Harris, and F. Jeanneteau, “Gluco- TMA #4. (Supplementary Materials) corticoid receptor action in metabolic and neuronal func- tion,” F1000Research, vol. 6, 2017. [17] N. E. Buckley, P. Haddock, R. De Matos Simoes et al., “A BRCA1 References deficient, NFκB driven immune signal predicts good outcome in triple negative breast cancer,” Oncotarget, vol. 7, no. 15, 2016. [1] F. Bray, J. Ferlay, I. Soerjomataram et al., “Global cancer [18] P. Jez ´ equel, ´ D. Loussouarn, C. Guerin-Charbonnel ´ et al., statistics 2018: GLOBOCAN estimates of incidence and “Gene-expression molecular subtyping of triple-negative mortality worldwide for 36 cancers in 185 countries,” Cancer Journal for Clinicians, vol. 68, no. 6, 2018. breast cancer tumours: importance of immune response,” Breast Cancer Research, vol. 17, 2015. [2] C. Liedtke, C. Mazouni, K. R. Hess et al., “Response to neoadjuvant therapy and long-term survival in patients with [19] A. Rody, T. Karn, C. Liedtke et al., “A clinically relevant gene triple-negative breast cancer,” Journal of Clinical Oncology, signature in triple negative and basal-like breast cancer,” vol. 26, no. 8, 2008. Breast Cancer Research, vol. 13, no. 5, 2011. 10 Journal of Oncology [20] C. Desmedt, F. Piette, S. Loi et al., “Strong time dependence of cancer,” Breast Cancer Research Treatment, vol. 171, no. 1, the 76-gene prognostic signature for node-negative breast cancer patients in the TRANSBIG multicenter independent [36] R. L. Flaherty, M. Owen, A. Fagan-Murphy et al., “Gluco- corticoids induce production of reactive oxygen species/re- validation series,” Clinical Cancer Research, vol. 13, no. 11, 2007. active nitrogen species and DNA damage through an iNOS mediated pathway in breast cancer,” Breast Cancer Research, [21] T. Casey, J. Bond, S. Tighe et al., “Molecular signatures suggest vol. 19, no. 1, 2017. a major role for stromal cells in development of invasive [37] K. A. Conklin, “Chemotherapy-associated oxidative stress: breast cancer,” Breast Cancer Research Treatment, vol. 114, impact on chemotherapeutic effectiveness,” Integrative cancer no. 1, 2009. therapy, vol. 3, 2004. [22] A. Tripathi, C. King, A. de la Morenas et al., “Gene expression [38] D. Gewirtz, “A critical evaluation of the mechanisms of action abnormalities in histologically normal breast epithelium of proposed for the antitumor effects of the anthracycline an- breast cancer patients,” International Journal of Cancer, tibiotics adriamycin and daunorubicin,” Biochemical Phar- vol. 122, no. 7, 2007. macology, vol. 57, no. 7, 1999. [23] P. B. Mullan, V. Bingham, P. Haddock et al., “NUP98–a novel [39] I. M. Adcock, Y. Nasuhara, D. A. Stevens, and P. J. Barnes, predictor of response to anthracycline-based chemotherapy in “Ligand-induced differentiation of glucocorticoid receptor triple negative breast cancer,” BMC Cancer, vol. 19, no. 1, (GR) trans-repression and transactivation: preferential tar- getting of NF-κB and lack of I-κB involvement,” British [24] K. Orr, N. E. Buckley, P. Haddock et al., “)romboxane A2 Journal of Pharmacology, vol. 127, no. 4, 1999. receptor (TBXA2R) is a potent survival factor for triple [40] Y.-S. Lu, P.-Y. Yeh, S.-E. Chuang et al., “Glucocorticoids negative breast cancers (TNBCs),” Oncotarget, vol. 7, no. 34, enhance cytotoxicity of cisplatin via suppression of NF-B activation in the glucocorticoid receptor-rich human cervical [25] R. D. A. Wilkinson, R. E. Burden, S. H. McDowell et al., “A carcinoma cell line SiHa,” Journal of Endocrinology, vol. 188, novel role for cathepsin S as a potential biomarker in triple no. 2, 2006. negative breast cancer,” Journal of Oncology, vol. 2019, Article [41] H. Wang, Y. Wang, E. R. Rayburn et al., “Dexamethasone as a ID 3980273, 12 pages, 2019. chemosensitizer for breast cancer chemotherapy: potentiation [26] C. Knowlson, P. Haddock, V. Bingham et al., “Pin1 plays a key of the antitumor activity of adriamycin, modulation of cy- role in the response to treatment and clinical outcome in triple tokine expression, and pharmacokinetics,” International negative breast cancer,” =erapeutic Advances in Medical Journal Oncology, vol. 30, no. 4, 2007. Oncology, vol. 12, 2020. [42] H. Wang, M. Li, J. J. Rinehart, and R. Zhang, “Pretreatment [27] D. P. Boyle, D. G. McArt, G. Irwin et al., “)e prognostic with dexamethasone increases antitumor activity of carbo- significance of the aberrant extremes of p53 immunophe- platin and gemcitabine in mice bearing human cancer xe- notypes in breast cancer,” Histopathology, vol. 65, no. 3, 2014. nografts: in vivo activity, pharmacokinetics, and clinical [28] I. Vasconcelos, A. Hussainzada, S. Berger et al., “)e St. Gallen implications for cancer chemotherapy,” Clinical Cancer Re- surrogate classification for breast cancer subtypes successfully search, vol. 10, no. 5, 2004. predicts tumor presenting features, nodal involvement, re- currence patterns and disease free survival,” =e Breast, vol. 29, 2016. [29] D. Pan, M. Kocherginsky, and S. D. Conzen, “Activation of the glucocorticoid receptor is associated with poor prognosis in estrogen receptor-negative breast cancer,” Cancer Research, vol. 71, no. 20, pp. 6360–6370, 2011. [30] S. Almazrou and L. Alnaim, “Evaluation of adherence to chemotherapy-induced nausea and vomiting guidelines. An observational study,” Journal of Cancer =erapy, vol. 3, no. 5, pp. 613–620, 2012. [31] H. Fujii, H. Iihara, M. Ishihara et al., “Improvement of ad- herence to guidelines for antiemetic medication enhances emetic control in patients with colorectal cancer receiving chemotherapy of moderate emetic risk,” Anticancer Research, vol. 33, no. 12, 2013. [32] J. T. Veneris, L. Huang, J. E. Churpek, S. D. Conzen, and G. F. Fleming, “Glucocorticoid receptor expression is asso- ciated with inferior overall survival independent of BRCA mutation status in ovarian cancer,” International Journal of Gynecologic Cancer, vol. 29, no. 2, pp. 357–364, 2019. [33] M. M. S. Obradovic, B. Hamelin, N. Manevski et al., “Glu- cocorticoids promote breast cancer metastasis,” Nature, vol. 567, no. 7749, pp. 540–544, 2019. [34] W. Wu, S. Chaudhuri, D. R. Brickley et al., “Microarray analysis reveals glucocorticoid-regulated survival genes that are associated with inhibition of apoptosis in breast epithelial cells,” Cancer Research, vol. 64, no. 5, 2004. [35] C.-H. Lin, P.-Y. Chuang, S.-L. You et al., “Effect of gluco- corticoid use on survival in patients with stage I–III breast

Journal

Journal of Oncology – Hindawi Publishing Corporation

Published: May 20, 2020

There are no references for this article.