BRCA1 germ-line mutations predispose to hereditary breast and ovarian cancers. Cells lacking functional BRCA1 protein are deficient in the homologous recombination DNA repair pathway. Base excision repair (BER) is essential for processing base damage induced by endogenous and exogenous sources. Recently, BRCA1 was shown to transcriptionally regulate expression of genes involved in BER. The primary aim of the work described in this thesis was to investigate whether targeting the double-strand break pathway in BRCA1-BER deficient cells using ATM or DNA-PKcs inhibitors would be synthetically lethal. DNA repair gene and protein expression in BRCA1 deficient and proficient cells were investigated. Initially 84 DNA repair genes were investigated. Data demonstrated down-regulation of several DNA repair mRNAs in BRCA1 mutant/knockdown cell lines as compared to proficient cell lines. RT-qPCR was performed for selected DNA repair genes and confirmed statistically significant down-regulation of these genes. Protein expression of the selected group was assessed and showed down-regulation. These results suggest that BRCA1 deficiency may be associated with a global defect in the BER pathway. BRCA1 deficient cells were targeted by ATM/DNA-PKcs inhibitors. BRCA1-BER deficient cells were sensitive to ATM and DNA-PKcs inhibitor treatment either alone or in combination with cisplatin and synthetic lethality was evidenced by DNA double strand breaks accumulation, cell cycle arrest and apoptosis. This in vitro study suggests that a potential synthetic lethality relationship exists between BRCA1 deficiency and ATM/DNA-PKcs inhibition. Moreover, results support the hypothesis that cisplatin increases the efficacy of ATM and DNA-PKcs inhibition in BRCA1 deficient cells. Taken together, this study provides the pre-clinical evidence that ATM and DNA-PKcs could be alternative synthetic lethality targets in BRCA1 deficient breast cancer.

Although global gene microarray studies have demonstrated the molecular heterogeneity of breast cancer (BC) and provided potential for clinical applications, the molecular subclassification of luminal/ER-positive tumours, which is the largest class of BC, remains unclear. Characterisation of luminal/ER-positive subtypes could have important implications in clinical decision-making and patient management. The patient study cohort is derived from a consecutive series of approximately 1902 cases of primary operable invasive breast carcinoma obtained from the Nottingham Tenovus Primary Breast Carcinoma Series, with patients presenting between 1986 and 1998. This is a well-characterized series of primary breast carcinoma that has been treated in a uniform way and previously used to study a wide range of proteins. Using gene microarray experiments in 128 frozen invasive BC derived from this series , 47,2 93 gene transcripts were analysed using a number of different bio-statistical models to identify a transcript signature for luminal/ER-positive BC, from which candidate genes were selected and that can be used to characterise ER-positive breast cancer. In addition, other biomarkers with strong relevance in ER-positive breast cancer were studied because the evidence strongly suggests an important role in the biology and molecular classification of ER-positive breast cancer. The selection criteria was based on published literature concentrating mainly on ER related pathways including ER coregulators (CARMI, PELPI), cellular proliferation (p27. TK1, cyclin B1), apoptosis (Bc12), Akt/PIK3 pathway (FOX03a), gene expression profiling (FOXA1, XBP1, TFF1) and endocrine resistance (CD71). Immunohistochemistry and high throughput tissue micro array technology were used to study the protein expression of 16 biomarkers with strong relevance to ER pathways in a well characterised consecutive series of invasive BC (n=1902) in addition to anther 9 markers that were available from the database of the breast cancer research group, University of Nottingham. The data were analysed using different clustering methods including K-means and Partitioning around Medoids. Kaplan Meier plots with Log-rank test (LR) were used to model clinical outcome. A transcript signature for ER positive BC was identified including RERG, GATA3 and other genes by a supervised classification analysis using 10-fold external cross-validation of the gene microarray data. Immunohistochemical validation studies confirmed their association with ER positive BC. Through a consensus approach using different clustering techniques applied to protein expression data 25 markers, three biological clusters (patient subclasses) in ER positive breast cancer showing significant difference in clinical outcome (LR= 28.185 & p<0.001) have been identified. Importantly, the poor prognosis cluster was significantly characterised by high tumour grade and frequent development of distant metastasis. In conclusion, our results emphasised the heterogeneity of luminal/ER-positive BC. Molecular profiling of breast cancer using protein biomarkers on TMAs can sub-classify ER-positive tumours into clinically and biologically relevant subgroups.

The role of C19 steroids in the human endometrium is at present unclear. In order to gain an insight into their action, radioimmunoassay procedures were developed which had sufficient specificity and accuracy to measure testosterone, 5α-DHT, oestradiol, progesterone and androstenedione in endometrial samples. Amounts of androstenedione were greater (range 1.2-20.8 ng/mg tissue) than other steroids. Samples were obtained from patients presenting with a variety of conditions: subfertility, postmenopausal bleeding, dysfunctional uterine bleeding and abdominal pain. Patients admitted for sterilisation were used as normal controls. A significant positive correlation (r = 0.80) was found between the levels of testosterone and 5α-DHT measured in the same tissue which suggests the presence of a 5α reductase enzyme. No relationship was observed in tissue steroid concentration and age of the patients. Steroid concentrations were found to be high in tissues obtained from patients with endometrial carcinomas whereas progesterone concentration being low in subfertiles. The oestrogen, progesterone and androgen receptor levels of endometrial tissues from subfertile women were also determined using the DCC technique and not the procedure based on protamine sulphate precipitation since endometrial tissue available was very small. No correlation was found between receptor binding sites and day of cycle for any of the three steroids analysed; nor was there any correlation between age and receptor binding sites. A cyclic variation followed by normal women was seen in the oestrogen and progesterone receptor concentrations in the menstrual cycle. Such a variation was also observed in subfertile women on clomiphene citrate therapy. It is concluded that normal endometrium contains measurable quantities of androgens and that a receptor for 5α-DHT is present. The difference in steroid concentrations between normal and pathological states suggest that C19 steroids may be induced in the development of abnormalities.

Breast cancer is the most common form of cancer among women in the United Kingdom, and there is considerable investment in research to identify the causes of breast cancer and the best means of diagnosis and treatments. The randomised controlled trial is the principal method used for evaluating diagnostic and treatment options. Trial organisers depend on recruitment of sufficient numbers of patients in order that the results are statistically significant and generalisable, but accrual to cancer clinical trials is poor. This research analyses factors affecting the accrual of women to two breast cancer trials, the British Association of Surgical Oncology (BASO) II trial (a treatment trial) and the International Breast cancer Intervention Study (IBIS) (a prevention trial). The aims were to identify the factors affecting the recruitment of women to breast cancer clinical trials from the surgeons' and multi-disciplinary teams' perspectives and, importantly, from the perspectives of women approached to participate in clinical trials, and their reasons for participation, or non-participation in the trials. There were three phases to the study using multiple methods. In the first phase quantitative methods were used in the form of a questionnaire, sent to consultant surgeons responsible for collecting audit data regarding breast cancer in the United Kingdom. The second and third phase incorporated qualitative methods of data collection; the second phase included in-depth interviews with multi-disciplinary teams; and the third phase involved focus group and individual interviews with women approached to join a breast cancer clinical trial. These three phases were carried out in both the trials examined. The findings contribute to the debate and knowledge of the recruitment of women to breast cancer clinical trials in a number of ways. Firstly, by including the views of all the key stakeholders concerned with breast cancer clinical trials. Secondly, by highlighting the factors affecting recruitment to these two breast cancer clinical trials. Thirdly, by making recommendations on methods to enhance recruitment.

Background: Neoadjuvant chemotherapy (NAC) is being used as first line treatment in women with large and locally advanced breast cancers (LLABCs). However, the response to NAC is difficult to predict. Growing evidence suggests that these patients are immunosuppressed and that circulating immunosuppressive regulatory cells and humoral factors affect the response to NAC. We explored the possible role of the in situ tumour immune milieu in inducing and affecting the responses to NAC, and the contribution of concomitant systemic circulating regulatory cells. Methods: Paraffin-embedded breast cancers and ipsilateral axillary lymph nodes (ALNs) from pre- and post-NAC samples of a cohort of 33 women with LLABCs, 16 of whom had their blood regulatory cells previously investigated. Various immune cell infiltrations and expression of cytokines/biological molecules in the specimens were studied using appropriate monoclonal antibodies and immunohistochemistry. Statistical analysis was carried out using non-parametric tests with SPSS version 21. Results: High levels of pre-NAC tumour-infiltrating lymphocytes (TILs) (p < 0.001) and subsets of CD4⁺T cells (intratumoural, p=0.023; peritumoural, p=0.001), CD8+T cells (intratumoural, p=0.008; peritumoural, p=0.002) and CD56⁺NK cells (intratumoural, p=0.001; peritumoural, p < 0.001) were significantly associated with a pathological complete response (pCR). High levels of CD163⁺macrophages were also significantly associated with a good pathological response (p=0.004) and pCR (p=0.008). There was a positive correlation between the CD8:FOXP3 ratio and grade of pathological response. In multivariate analyses, TILs and peritumoural CD56+NK cells were found to be independent predictive factors for pCR. There was a significantly high expression of IL-10 in post-NAC breast specimens with poor responses to NAC (p < 0.001). NAC significantly reduced infiltrating T regulatory cells (Tregs) (p=0.001) and PD1⁺T cells (p=0.005), as well as expression of IL-4 (p=0.016). There was no significant difference between the percentages (%) of immune cells present in ALNs with or without metastases but there was a T helper-2 cytokine polarisation in metastatic ALNs. Metastatic ALNs with a high % of CD8+T cells (p=0.048) and low % of FOXP3+Tregs (p=0.019) were significantly associated with an ALN pCR. There was a significantly positive correlation between circulating and intratumoural infiltrating Tregs following NAC (p=0.003). Conclusions: The tumour immune microenvironment is a key factor in achieving a good pathological response with NAC. Tumour and blood immune parameters may be clinically useful in identifying women with LLABCs likely to respond to NAC. Our findings also suggest that the beneficial effects of NAC are mediated via modulation of anticancer immunity, in particular by reduction of T regulatory cells and immunosuppressive humoral factors.

Lymphovascular invasion, including both blood and lymphatic vessel invasion, is an important step in tumourigenesis and is a prerequisite event in the complex process of metastasis. In breast cancer, data suggest that lymphatic vessels plays a significant role by being the major route for lymphatic vessel invasion (LVI) and that inflammatory cells, may be involved in its regulation. Macrophage, a major component of inflammatory infiltrate, has been shown to be importance in tumour growth and metastasis. The biomolecular mechanisms by which macrophage can mediate dissemination of tumour cells through the lymphatic compartment are, however, not yet fully understood. In the first part of this project, CD68 (total) and CD163 (M2 subtype) macrophages were examined in two consecutive sections of 89 full-face invasive breast cancer samples. The density and localisation of macrophages were assessed using the Chalkley-grid counting method and their association with clinicopathological variables and clinical outcome was identified. A high count of CD68 macrophages was associated with the presence of lymphatic vessel invasion (LVI) and a high microvessel density (MVD), and tumour with high microvessel densities had high expression of CD163 macrophages, confirming a role for macrophage in mediating the process of angiogenesis and metastasis via lymphatic vessels. CD68 and CD163 macrophages were not associated with disease-free or breast cancer-specific survivals. In order to elucidate previous conflicting data by our lab, suggesting that recombinant, and macrophage-induced IL-1β can mediate in vitro LVI, with patient based IHC data showing that high expression of tumour IL-1β was associated with improved disease specific survival and was not associated with LVI, it was of interest to study the expression of IL-1β signalling-related proteins including caspase-1, IL-1R I, IL-1R II, IL-1RacP, and IL-1RA. This was conducted in 1902 early stage invasive breast cancer patients, with long-term follow-up, using IHC. In addition, serum IL-1β concentration was measured in a matched subset of patients using enzyme-linked immunosorbent assay (ELISA) and expression levels examined for associations with the above proteins as well as clinicopathological criteria and patient prognosis. Although pre-requisites for IL-1 signalling pathways are present in breast tumours, tumour IL-1β expression was not strongly correlated with the expression of caspase-1, IL-1R I, IL-1R II, IL-1RacP and IL-1RA. Caspase-1, IL-1R I, IL-1R II, IL-1RacP and IL-1RA were not independent prognostic factors. The analysis of the markers stained in the current study could not provide any further elucidation of the conflict between previous in vitro and IHC studies regarding the role of IL-1β in regulating LVI. The third part of the study dealt with IHC and in vitro investigations of macrophage-associated cytokines (IL-6 and IL-10) to assess their role in regulating LVI in breast cancer. Scratch wound migration assays were conducted to examine the influence of L-6 and IL-10 on breast cancer cell migration with results showing that low concentrations of IL-6 induced migration, whereas high concentrations inhibited it. Similarly, a high concentration of IL-10 also inhibited breast cancer cell migration. In addition, the effect of these cytokines, IL-6 and IL-10, on the tumour- endothelial (blood and lymphatic) adhesive process was also studied. Static adhesion assays showed that the adhesion patterns of breast cancer cell lines to the endothelial cells did not change following pre-stimulation of either blood or lymphatic endothelial cells with either IL-6 or IL-10. Thus, in vitro data suggest that both cytokines may not play a significant role in regulating LVI other than their effect on migration. The important of IHC expression of IL-6 and IL-10 were examined in a large cohort of early stage invasive breast cancer patients. Data showed that high expressions of IL-6 and IL-10 in breast cancer tissues were not associated with the presence of LVI. A significant association was found between high expression of IL-6, and longer disease-free interval, but it was not associated with improved disease-free survival. However, high IL-10 expression was not associated with improved disease-free survival, and breast cancer-specific survival. In the final part of the project, the correlation between macrophage-associated cytokines including IL-1β, IL-6, and IL-10 and their downstream signalling elements, and target genes was assessed using Spearman’s rank correlation test. Expression data for downstream signalling elements was provided via collaboration with the Breast Pathology group. No strong correlations were observed between these cytokines. In addition, their expressions in breast cancer tissues were not strongly correlated with downstream signalling compartments, or target genes. In conclusion, macrophages seem to play an important role in regulating LVI, with such LVI being almost entirely invasion of lymphatic vessels. Macrophage and macrophage-associated cytokines (IL-6 and IL-10) have been found to potential play role in breast cancer progression, with preliminary in vitro data suggesting that this may be via tumour cell migration rather than influencing tumour-endothelial cell interaction and LVI. This project has shed some light on the role of macrophage-associated cytokines in regulating LVI however more studies are needed to determine the mechanisms whereby IL-1β, IL-6, and IL-10 regulate the progression and prognosis of breast cancer.

The evidence suggests that alteration of the endometrial environment in women with previous failed in vitro fertilization (IVF) or intra-cytoplasmic sperm injection (ICSI) by inflicting endometrial injury improves the outcome of the subsequent treatment cycle. The selected population assessed in the studies supporting this statement prevents from generalizing this to all women undergoing IVF or ICSI treatment. The mechanisms responsible for alteration of the endometrial environment following biopsy remain still unclear. If proven effective, endometrial injury could serve as a beneficial adjunct for all couples undergoing assisted reproductive treatment (ART). The hypothesis forming the basis of the work reported in this thesis was that endometrial injury in the cycle directly preceding an embryo transfer cycle, be it fresh or frozen, improves the outcome of that treatment irrespective of previous reproductive history. In order to support or refute this hypothesis, a clinical trial of endometrial biopsy prior to IVF or ICSI treatment has been designed. Additional objectives that allow examining the mechanisms responsible for the beneficial effects of the biopsy on the endometrium included utilization of two-dimensional, three- dimensional and Spatio-Temporal Image Correlation (STIC) ultrasound and assessment of the predictive value of any sonographic indices on ART outcome. Same sonographic modalities were used to determine the effect of endometrial injury on endometrial and subendometrial blood flow. Similarly, an examination of the value of uterine natural killer (uNK) cell numbers and expression of selected molecular markers of endometrial function as a predictors of ART outcome was explored. In this work it was not possible to clearly demonstrate a benefit of routine endometrial biopsy in all women undergoing ART on treatment outcomes, though clues, as to which population might benefit from the procedure, were identified. Extensive sonographic analysis of endometrial factors did not produce results allowing for unequivocal non- invasive identification of a receptive endometrial milieu. Triple layer endometrial pattern at oocyte collection was correlated with positive outcome and endometrial biopsy. Spatio-Temporal Image Correlation did not fulfil expectations as a non-invasive marker of endometrial receptivity and was not able to identify women that would go on to have a successful ART outcome. In a small number of patients, STIC indices were able to predict first trimester miscarriage with relatively high sensitivity and specificity. uNK cell numbers were not associated nor predictive of ART outcome, and as such not useful as a routine diagnostic tests prior to ART. An observed significant decrease in uNK cell levels following endometrial biopsy indicates a possible mechanism of action of this intervention. Limited (3) molecular cues were not able to differentiate between a receptive and non-receptive endometrium. This work, however extensive, indicates that the endometrium is a complex microenvironment requiring further investigation in order to understand and influence the mechanisms related to pregnancy establishment and development.

The overall aim of this thesis was to examine the role of social support networks and the significant other in breast cancer survivors. This thesis begins with a scoping review of current scientific literature to assess social support in the breast cancer context. The findings from this review showed that studies which have investigated social support networks in women with breast cancer do not often consider whether the actual support needs of the patient are met. In addition, there is little evidence from the support sources about the support that they may attempt to provide. Past research has attempted to measure social support and coping but is limited when attempting to understand interactions between the woman with breast cancer and their significant other as a dyadic process. Furthermore, the review indicated the lack of relevant research based in the UK. In Study 1, the perceptions held regarding the male partner were explored in five in-depth interviews with breast cancer survivors to answer the research question of “What support do male partners provide?”. Data were thematically analysed and findings indicated that male partners were perceived to provide instrumental support, whereas female support sources were perceived to provide more emotional support. In addition, the male partner was not always perceived to be the significant other as previous literature has suggested. This finding led to Studies 2A and 2B which aimed to find out more about different social support sources, the types of support provided and identify the significant other. Studies 2A and 2B mark one of the first examples of using ecomaps in this research area 1) as an elicitation tool in semi-structured interviews to collect data regarding an individual’s social support network and 2) as a method of visually presenting social support networks. Studies 2A and 2B illustrate the variety of sources and support provided. The final study, Study 3, presents three case studies of dyads, consisting of a breast cancer survivor and their nominated significant other, who they perceived to have provided them with the most support along the illness trajectory. This thesis contributes to the research literature in several ways. First, it outlines the research gaps in the current scientific research. Second, it provides a novel methodology for investigating the social support networks of breast cancer survivors through the use of ecomapping. Third, it contributes to the emerging knowledge on dyadic coping. The new knowledge generated is of importance when considering the post-treatment phase of the breast cancer trajectory. Finally the limitations and strengths of this work are discussed.

Background: HER2 gene amplification and protein overexpression defined as HER2 positivity (HER2+) breast cancer (BC) is encountered in 15-25% of cases and is characterised by an aggressive behaviour and poor outcome. Despite the high clinical efficacy of anti-HER2 targeted therapy, the response and clinical behaviour of HER2+ tumours is variable. There is evidence indicating that the response of HER2+ BC to anti-HER2 targeted therapy and chemotherapy is related to Oestrogen Receptor (ER) expression. In addition global gene expression profiling studies have demonstrated that ER and HER2 are the main determinant of BC molecular profiles and that HER2+/ER+ (luminal B) are molecularly distinct from HER2+/ER- (HER2 positive) tumours. It is hypothesised that ER+/HER2+ BC is also a distinct molecular class when compared to tumours with single positive or double negative expression. Therefore, this study aimed to investigate the biological impact of ER expression in HER2+ BC with consideration of the molecular classification of BC and key pathways related to expression and behaviour of both proteins in an attempt to understand their variable biological significance and relationship to treatment response and potentially to identify new therapeutic targets. Methods: Methods included assessment of proteins with known associations with HER2 and ER status and correlating their expression with clinicopathological variables, molecular classes, different key BC proteins and outcome. For this purpose, Immunohistochemistry (IHC) was used to stain a number of key targets, including Mitogen Activated Protein Kinases (MAPKs), Phosphatidylinositol 3 kinase (PI3K)/Akt/mammalian target of Rapamycin (mTOR) pathway members and other proteins related to HER2 and ER and proliferation in a large well-characterised uniformly treated and annotated cohort of 1835 patients with primary BC. In addition, a cohort of 197 primary BC patients treated with Trastuzumab between 2003 and 2012, were also included. Reverse Phase Protein Array (RPPA) was used to quantify protein expression in six BC cell lines. To assess the effect of HER2 on cell lines with and without ER expression, two HER2 negative cell lines (MCF-7 and MDA-MB-231) were transfected with HER2. Results: The majority of MAPKs pathway members (pan Extracellular Signal- Regulated Kinase (ERK1/2), nuclear phosphorylated (p)-ERK1/2, p-c-jun-N terminal Kinase (JNK1/2), pan p38, p-p38 and p-ATF2 and p-C-JUN) showed positive associations with good prognostic variables and longer survival in the whole (unselected) cohort and in ER+ tumours but many of these associations were lost with HER2 co-expression. Such associations were infrequently observed within ER-HER2+ cases. HER2 overexpression was associated with downregulation of phosphorylated MAPKs within the whole cohort and within ER+ BC (significant for nuclear p-ERK1/2, p-ATF2 and p-p38), but ERK1/2 and p-p38 were associated with HER2 positivity within ER- tumours implying their context specific function. In addition, pan ERK1/2, p-p38 and p-ATF2 were independent predictors of better survival in BC and in ER+ BC. RPPA confirmed the IHC findings and showed similar association where the expression of MAPKs was different in ER+HER2+ cell lines compared to ER-HER2+ and ER+HER2- ones. Regarding the PI3K/Akt/mTOR pathway, p-mTORC1 and Phosphatase and Tensin homolog (PTEN) were negatively associated with HER2 overexpression in ER+ tumours but were (in addition to Akt and PI3K) positively associated with HER2 in ER- tumours. Meanwhile, mTOR exhibited positive associations with favourable prognostic factors within ER+ BC which were decreased with HER2 co-expression and with ER loss. Additionally, p-mTORC1 was associated with prolonged breast cancer specific survival (BCSS) within Akt+ tumours but not within the whole cohort or other subgroups. In this study, using RPPA, mTOR and PTEN were positively associated with ER and negatively with HER2 in ER+ cell lines and p-mTORC1 was positively associated with HER2 in ER- cell lines in addition to other members. Importantly, PI3K, Akt, p-mTORC1 and its downstream p-S6K showed increased expression within ER+HER2+ cell lines compared to ER-HER2+ cell lines but PTEN expression was increased in ER-HER2+ vs ER+HER2+ cell lines. When the biological significance of HER2 and KI67-LI was investigated in ER+ tumours, both HER2 and KI67-LI were associated with poor prognostic variables and adverse outcome in the ER+ tumours. Although KI67-LI rather than HER2 was associated with downregulation of luminal associated biomarkers, HER2 positivity was associated with worse outcome in ER+ tumours, indicating that HER2+ BC are distinct aggressive tumours regardless of their proliferative activity. Investigation of other proteins related to HER2 and ER pathways revealed that nuclear form of both the Carboxyl-terminus of Hsp-70-Interacting Protein (CHIP) and the stem cell protein, Sry-Related HMG Box 9 (SOX9), were negatively associated with HER2. CHIP was positively associated with ER, ER-associated proteins and prolonged outcome in the unselected BC and in ER+ BC but not in ER+/HER2+ and ER-/HER2+ tumours. The phosphorylated form of ER at Serine (SER) 118 was positively associated with good prognostic variables and negatively with HER2 in the unselected series and in the ER+ BC group with an observed decrease in these associations within ER+/HER2+ tumours. Increased loss of association was encountered and even some unfavourable associations were observed within ER-HER2+. Furthermore, it was associated with prolonged survival in ER+ tumours and was a predictor of prolonged survival in patients receiving tamoxifen therapy. Clustering analysis to predict class memberships based on HER2 and ER expressions using a large panel of biomarkers related to ER, HER2 and key pathways’ proteins, generated a decision tree that could be a future model for patients’ stratification which indicated the overwhelming driving effect of HER2 expression. Conclusions: ER+/HER2+ BC is a distinct biological group, having some luminal features but is associated with worst outcome owing to the co-expression of HER2 independently from ER influence. The investigation of MAPKs, PI3K/Akt/mTOR pathways and other proteins highlighted their differential expressions and associations (with key proteins related to ER and HER2) within different BC subgroups based on ER and HER2 expressions indicating that ER+HER2+ stands as a group with unique features from those with single positive or double negative expression. Finally, development of a decision tree is a potentially promising tool for patients’ stratification. Breast cancer cell line studies using the high throughput technique, RPPA, showed good concordance with IHC results, implying that further in vitro studies using relevant cell line models could be possible.

Background. In vitro studies have shown that ER+ acquired tamoxifen resistant MCF7 breast cancer cell lines can show elevated levels of EGFR expression with an increase in its subsequent signalling pathway(s) and that these are growth inhibited by gefitinib, an EGFR tyrosine kinase inhibitor. This thesis examines the effect of gefitinib on tamoxifen resistant human breast cancer in the clinical setting and in an ‘in-vivo’ mouse model. Patients and Methods. This phase 2 clinical study recruited 54 patients. 28 were oestrogen receptor positive and had progressed on tamoxifen treatment(acquired resistance), the other 26 (48.1%) were oestrogen receptor negative(de novo resistance). Patients were given a loading dose of 1000mg gefitinib on Day 1 and then gefitinib 500mg as a once daily oral dosing until evidence of disease progression. Clinical data were recorded. Sequential tumour biopsies were taken pre-treatment, after 8 weeks therapy and at the development of resistance and analysed immunocytochemically to identify predictive factors for response to treatment and also to see the effect of treatment and resistance on tumour biology, encompassing monitoring steroid receptors, EGFR, HER2 and IGFR, downstream kinases MAPK and AKT, and the proliferation marker Ki67. In parallel with the clinical study, ER+ acquired tamoxifen resistant MCF7 xenografts (TAMR) were grown in nude mice in the presence of tamoxifen and treated with gefitinib 50mg per day orally (designated 3 treatment) or tamoxifen alone (designated control) and monitored for impact on tumour growth. Results. In the phase 2 study gefitinib treatment was well tolerated with an overall clinical benefit rate of 33.3% (n=18/54). Pre-treatment oestrogen receptor positivity was associated with tumour response to gefitinib (p=0.015, longer TTP (p = 0.015), and with clinical benefit (CB) in 53.6 % of the ER+ acquired tamoxifen resistant patients. In contrast, the clinical benefit rate was minimal in the steroid receptor negative patient cohort (11.5%). All patients in this series expressed detectable levels of EGFR, but high pre-treatment levels of EGFR predicted a poorer outcome (p=0.075) Only patients achieving CB had a significant fall in Ki67 staining as measured at 8 weeks versus pretreatment levels (p=0.024), and that Ki67 levels were lower in CB than PD patients at this time. We observed lower levels of EGFR phosphorylation at this time point in some CB patients. Further examination of the CB pts who showed a >10% decline in EGFR phosphorylation revealed decreases in phosphorylation of MAPK and also in Ki67. TAMR xenografts expressed high levels of EGFR as previously observed in vitro. Their growth was significantly inhibited by gefitinib (p=0.039) over the study period while after only 2 weeks of gefitinib treatment tumours showed a decrease in the level of Ki67 staining (p = 0.068). Conclusion. Acquired tamoxifen resistance in vivo both in patients and in a xenograft model appears to be in part mediated through EGFR pathway signalling and this can be blocked and growth inhibited with gefitinib. In ER 4 negative tumours the effects of gefitinib were less striking, suggesting alternative signalling pathways are dominant in promoting their growth despite obvious overexpression of EGFR.