Search

Evidence exists which suggests that the expression of certain genes is associated with the development of human malignancy. The aims of the work reported in this thesis were (i) to quantitate accurately the expression (ie transcription) of selected important genes in human breast cancer, to relate this to clinical and pathological factors and to other genetic changes in the tumours and (ii) to develop a xenograft model in which to study the effect of an anticancer agent on breast tumour gene expression in vivo and thus to establish whether these approaches migh have therapeutic implications. The particular approach used to achieve these objectives was the development of molecular techniques to measure gene expression and loss of genetic material at the nucleic acid level. Gene expression was measured in primary breast cancers from 80 female patients and compared to that in normal control tissues from other patients who had surgical removal of normal breast tissue. Ribonucleic acid (RNA) was extracted from the tumours and probed with radiolabelled gene probes. Gene expression (messenger RNA) was then quantified using laser densitometry on the resultant Northern blots. Increased expression of three oncogenes (p53, c-myc, c-erbB-2), a growth factor gene (TGF-beta) and an oestrogen regulated mRNA (pS2) was detected in varying proportions of the tumours. Overexpression of individual genes was found to correlate significantly to different clinical and pathological parameters. Since the loss of specific regions of deoxyribonucleic acid (DNA) is associated with oncogenesis, DNA was also extracted from the paired venous blood and tumour tissue of each patient. Allele loss specific to the tumour DNA was then determined using Southern blots probed with radiolabelled DNA. The p53 gene is located near the tip of the short arm of chromosome 17 at 17p13 and so this region was examined for loss of genetic material. Specific DNA loss in this region was demonstrated in 64% of tumours. Loss close to, but not including, the p53 locus correlated with p53 overexpression. Using the polymerase chain reaction, exons 5 and 6 of the p53 gene were selectively amplified and subsequently sequencing suggested the overexpressed p53 may be mutant and oncogenic in at least some tumours. These results suggest that chromosome 17p is important in breast cancer. Lack of normal p53 tumour suppressor gene expression or deletion of a nearby regulatory locus resulting in overexpression of mutant p53 are proposed as key events in the pathogenesis of breast cancer. It is concluded that the use of molecular approaches, such as those employed in this work, will help not only to elucidate the pathogenesis of breast cancer, but also clarify the mechanisms by which anticancer therapy is mediated and lead to new therapeutic approaches.

Aim To assess the change in expression of various tumour markers (p53, mcm2 and cyclin D1) with change in internal endocrine environment in patients with primary breast cancer and to study the expression and prognostic association of cyclin D1 and replication licensing proteins (mcm2 and geminin) in invasive breast cancer. Hypothesis The endocrine environment may be critical in the development and efficacy of treatment in breast cancer. Oestrogen has the potential to modulate the function of p53 network and cell cycle proteins (cyclin D1) in breast cancer, the consequences of which could profoundly affect treatment efficacy. It is proposed that the serum oestrogen level may effect the measurement of protein markers in primary breast cancer. It is hypothesised that the replication licensing protein family, mcm, may be a more consistent marker of proliferation than the conventional immunohistochemical marker, Ki-67, in primary breast cancer. Material and Methods Women with a diagnosis of primary, operable, symptomatic breast cancer were eligible and a total of 99 patients (47 premenopausal and 52 postmenopausal) were included in the study. Of these, 30 pre-menopausal and 33 post-menopausal women had paired samples of breast cancer biopsy and venous blood taken at diagnosis and later at the time of resectional surgery. The samples were collected at two time points to study the change in expression of tumour proteins, if any, with the change in serum oestradiol levels in pre-menopausal women, the postmenopausal women served as the control. The rest of the patients included had tissue samples and blood taken at one time point only i.e. at resectional surgey. Samples comprised formalin fixed breast cancer tissues from both time points for immunohistochemistry and FISH; serum was drawn for 17-? oestradiol, progesterone, FSH, LH. Tumour micro-arrays were prepared from the surgical specimens (corresponding to the second time point) and whole tissue sections were used to study the diagnostic specimen (corresponding to the first time point).Results Overexpression of cyclin D1 was seen in 63% of cancers studied whilst CCND1 amplification was identified in 30%. CCND1 amplification was significantly associated with cyclin D1 protein overexpression (p<0.001; Fisher’s exact test) and both CCND1 amplification and cyclin D1 protein expression with oestrogen receptor (ER) expression (p=0.003 and p<0.001; Fishers exact test). Neither CCND1 amplification nor cyclinD1 expression was associated with tumour size, pathological node status or HER2 amplification. High CCND1 amplification (CNG > 8) was significantly associated with high tumour grade and higher Nottingham Prognostic Index (p=0.001). Increased expression of both Mcm2 and geminin was associated with high tumour grade (p<0.05 ANOVA). Mcm2: Ki-67 and geminin: Ki-67 ratios were similar across different tumour grades suggesting similar G1 transit time in different grades of tumour, with higher grade tumours containing a larger proportion of in-cycle cells with proliferative potential. Expression of Mcm2 and geminin was associated with a high Nottingham Prognostic Index NPI (p<0.001, Spearman correlation) and inversely with ER-positive tumours (p<0.0001, t-test), hence correlated with poor prognosis. We explored a potential change in p53 expression with change in serum oestradiol levels in premenopausal women, following the generation of hypothesis from an animal study done in MCF-7 mouse xenograft model. The observation from MCF7 xenograft model suggesting an inverse relationship between p53 expression and serum oestradiol did not hold good in humans, although the numbers tested were small. Conclusion High level CCND1 amplification could be used to identify a subset of ER-positive breast cancers that are associated with poor prognosis. Such patients could be offered adjuvant chemotherapy, in keeping with evidence that CCND1 amplified cancers respond poorly to endocrine agents. This study demonstrates that replication licensing factors; Mcm2 and geminin correlate with poor prognosis (ER-negative and high NPI) breast cancers, and therefore merit consideration as poor prognostic markers in breast cancer. We did not observe any significant change in expression of p53 protein with a change in serum oestradiol levels in premenopausal patients with primary invasive breast cancer.

Tissue microarrays (TMAs) are a high-throughput technique that facilitates the survey of very large numbers of tumours, important both in clinical and research applications. However, the assessment of stained TMA sections is laborious and still needs to be carried manually, constituting a bottleneck in the pathologist?s work-flow. This process is also prone to perceptual errors and observer variability.Thus, there is strong motivation for the development of automated quantitative analysis of TMA image data. The analysis of breast TMA sections subjected to nuclear immunostaining begins with the classification of each spot as to the maintype of tissue that it contains, namely tumour, normal, stroma, or fat. Tumour and normal spots are then assigned a so-called quickscore composed of a pair or integer values, the first reflecting the proportion of epithelial nuclei that are stained, and the second reflecting the strength of staining of those nuclei. In this work, an approach was developed to analyse breast TMA spots subjectedto progesterone receptor immunohistochemistry. Spots were classified into their four main types through a method that combined a bag of features approachand classifiers based on either multi-layer perceptrons or latent Dirichlet allocation models. A classification accuracy of 74.6 % was achieved. Tumour and normal spots were scored via an approach that involved the computation of global features formalising the quickscore values used by pathologists, and the use of Gaussian processes for ordinal regression to predict actual quickscores based on global features. Mean absolute errors of 0.888 and 0.779 were achieved in the prediction of the first and second quickscore values, respectively. By setting thresholds on prediction confidence, it was possible to classify and score fractions of spots with substantially higher accuracies and lower mean absolute errors. Amethod for the segmentation of TMA spots into regions of different types was also investigated, to explore the generative nature of latent Dirichlet allocation models.

Lung cancer has developed from a rare condition into the leading cause of cancerrelated death in the United Kingdom. Lung cancer patients face a disease with a high symptom burden, increased psychosocial needs and a high mortality. Supportive care needs are often relevant from diagnosis. Despite this there are no clear follow-up structures for lung cancer patients that address both cancer management and supportive care. The aims of this study were to evaluate supportive care needs, assess predictors of such needs and identify factors which could aid service provision within Stobhill lung cancer services. Methods Supportive care needs were measured using an adapted Palliative Outcome Scale (POS), incorporated within a larger questionnaire. All lung cancer patients attending the clinic could complete this questionnaire. Respiratory symptoms, performance status, service usage, preferences and satisfaction were also assessed. Data were stratified to allow evaluation of three clinical groupings: all patients, newly diagnosed patients and patients in the last three months of life. Analyses were phased: descriptive analyses, univariate tests of association and multivariate regression. Results Three hundred and fifty three lung cancer patients completed questionnaires. The high symptom burden in lung cancer was confirmed. Anxiety, pain and dyspnoea were identified as the key issues. Poor performance status was identified to be an independent predictor of increased POS score, increased anxiety, increased pain and increased dyspnoea. There was no independent relationship between POS and survival. Although the majority of patients were satisfied with the care received, there was uncertainty regarding who was in charge of care and some disparity in preferred structure for follow-up. Conclusions Despite recent advances in lung cancer management, improvements are still required to address unmet supportive care needs of patients. Particular attention should be given to those with poorer performance status to effectively identify and meet such needs.