Evaluation of Metastasis-inducing Proteins as Biomarkers of Patient Outcome in Human Cancers

Human cancer in the Western World affects about a quarter of the population and about 40% will die of this disease. Although the primary tumour can be removed at surgery, in a good proportion of the patients the malignant tumour can disseminate or metastasise to other parts of the body. It is the process of metastasis that is largely responsible for the death of the patient, since conventional therapeutic agents are not selective enough to kill the cancer cells without harming growing normal cells. Much effort has been expended on the identification of genes which initiate the uncontrolled growth of the primary tumour, but little or no comparable effort has been expended in understanding the process of metastasis. To tackle the problem of metastasis, Prof Rudland and his group have utilised their unique breast stem cell lines to identify four genes capable of inducing metastasis in their animal model system. The level of the protein products of these genes were then assessed with specific antibodies in thin sections cut from routinely collected clinical specimens of the primary tumour of one set of patients with breast cancer. The presence of these metastasis-inducing proteins showed a much stronger correlation with death of the patient than biomarkers in routine use in hospitals today. It is proposed to repeat this work in many of the commonly occurring human cancers, including another set of breast cancers, so as to establish whether these metastasis-inducing proteins either singly or in combination can predict the outcome of the patient better than conventional biomarkers and to test the usefulness of the best one in a simple blood test. The identification of groups of patients at high risk of death by their overproduction of biomarkers for metastasis will enable them to be treated more vigorously and spare the remainder from harmful side effects of the vigorous treatment. It will also help validate the next phase of the work in developing suitable inhibitors that interfere in the life threatening aspect of cancer, that of metastasis.

Metastasis is the process in cancer which is responsible for most human deaths. Professor Rudland and colleagues have utilised their unique rat breast stem cell lines to identify a series of genes that can induce metastasis in the syngeneic rat system; the most well characterised are the invasion-inducing S100A4, S100P, and the adhesion-inducing osteopontin and anterior gradient protein 2 (AGR2). Overproduction of these metastasis-inducing proteins in the primary tumours of one group of patients with breast cancer is strongly correlated with early patient demise, with single risk factors of an order of magnitude greater than for single pathological and other molecular biomarkers in this particular group of breast cancer patients. This project will now utilise clinical specimens of other commonly-occurring carcinomas, e.g. colorectal, head and neck, stomach, kidney and pancreas as well as a second independent group of breast cancers to assess the prognostic significance of the four metastasis-inducing proteins S100A4, S100P, osteopontin and AGR2, either singly or in combination, in comparison with existing biomarkers for patient outcomes. The clinical specimens and anonomised patient data have been collected over a period of 14 years by the University of Liverpool?s Cancer Tissue Bank Research Centre (CTBRC) with full legal authorisation, informed patient consent and ethical approval. The levels of metastasis-inducing proteins will be assessed by immunocytochemistry underpinned by Western blotting on paraffin-embedded sections/tissue microarrays and protein extracts, respectively, already collected and retained within CTBRC. Positive immunocytochemical staining will be correlated with patient outcomes, particularly patient survival and relative risks associated with the metastasis-inducing proteins computed either singly or in combinations and compared with those of standard pathological and other biomarkers. The most highly correlated in breast cancer so far, AGR2, will be assessed in blood using an in-house ELISA assay for its clinical utility as a biomarker of patient outcome. Results of these experiments will validate the hypothesis that if a metastasis-inducing gene is overexpressed in the primary carcinoma then the patient will have a poor prognosis. The identification of high risk groups of cancer patients by overexpression of markers for metastasis will enable them to be selectively targeted for high dose conventional therapy and spare the remainder from harmful side effects. It will also help validate the next phase of the work in developing suitable inhibitors that interfere in the establishment and maintenance of metastases.