Surgery (St Vincent's) - Theses
Now showing items 1-12 of 37
Colorectal cancer in rural regional Australia
Colorectal cancer (CRC) is the second commonest cancer in Australia. The survival outcome of colorectal cancer patients within Australia is reported to vary with population density and between health services, with some literature showing poorer prognosis in rural regional and remote patients. Chapter one aims to outline some key issues in CRC such as epidemiology, diagnosis, treatment, surveillance, and outcome whilst chapter two aims to present a systematic review of how geographical disparity influences CRC survival. There are many potential factors that contribute to a poorer prognosis in rural regional CRC patients though the literature is limited, and at times, inconsistent. Thus, there is a need for regular audit, reporting and benchmarking of outcomes in CRC patients against agreed standards. I reviewed the long-term outcomes of CRC at Barwon Health, which serves the South West Victoria, a region with a population of some 500,000. My aim was to determine whether changes introduced to the management of CRC translated into improved survival after surgery. The literature to date has suggested that patients living in rural and regional Australia (grouped together) have worse colorectal cancer survival rates than those living in metropolitan Australia. This thesis was based on a prospectively maintained registry kept over a period of thirteen years from 2002 to 2014, that had accumulated 1079 patients who had undergone surgery at the University Hospital Geelong for CRC (744 colon cancers and 335 rectal cancers). The overall number of operations per year increased over time (p=0.037) but with similar proportions of elective and emergency surgery (p=0.75) and tumour stage (P=0.21). This lack of change in the proportion of elective cases was in spite of the Federal Government introducing a National Bowel Cancer Screening Program in 2006. The proportion of patients with severe comorbidities did increase (p=0.015) over the study period. The median survival after surgery by stage was 123 months, 141 months, 76 months and 17 months for stages I to IV CRC respectively. Overall, there were improvements observed in both peri-operative mortality (POMR) (p=0.028) and long- term survival (p=0.0025) of CRC patients in this major regional centre. I then reviewed the outcome of patients with metastatic disease. The Geelong database included 843 patients who had undergone resection and primary anastomosis for their primary tumour (661 colon cancers, 182 rectal cancers). Metastatic disease was present in 16% (135 patients) and was associated with an increased risk of anastomotic leakage (13% vs. 5%, p=0.003) and a higher peri- operative mortality rate (9.6% vs. 2.8%, p=0.0003). Patients with anastomotic leakage had a reduction in the overall survival (121 months vs. 66 months, p=0.02). The fifth chapter aimed to perform a regional study to identify patients with colorectal cancer at higher risk of developing metastatic disease. There were 503 patients (345 colon and 158 rectal) with non-metastatic (stage I-III) CRC who had resections and were followed up for at least five years. Metastatic progression was, as expected, significantly higher for patients with stage III disease (aHR 4.42 for colon cancer 95% CI 1.74 to 11.23, aHR 3.34 for rectal cancer 95% CI 1.36 to 8.22), and those with lymphovascular invasion (aHR 2.94 95% CI 1.70 to 5.06). Metastatic disease was also more likely to eventuate in those with severe comorbidities (aHR 2.18, 95% CI 0.26 to 0.86), and in colon cancer patients with the lowest socioeconomic status (aHR 2.03 95% CI 1.23 to 3.34). Gender, tumour location and geographical location (rural or regional) was not associated with metastatic progression. Before determining surveillance strategies targeting higher-risk patient groups in regional Victoria, these findings would require confirmation from similar studies in other regions of rural/regional Victoria, such as Bendigo, Albury / Wodonga, Latrobe Valley, or Shepparton.
Pre-mRNA alternative splicing in the epithelial to mesenchymal transition of breast cancer cells
Pre-mRNA alternative splicing in the epithelial to mesenchymal transition of breast cancer cells Edwin Widodo, Eva Tomaskovic-Crook, Bryce van Denderen, Erik W. Thompson Summary Alternative pre-messenger RNA splicing is a process that generates multiple variants of a single gene by virtue of the alternative exons that are transcribed. In breast cancer progression and metastasis, alternative splice events (ASE) are regulated during epithelial to mesenchymal transition (EMT). EMT occurs naturally during embryonic development as epithelial-derived cells become transiently mesenchymal and move around the embryo to generate the body plan. EMT status can be determined by expression of specific markers for EMT. E-cadherin is recognized as the archetypical marker for the epithelial phenotype. During carcinoma EMT, E-Cadherin is reduced by transcriptional repression and/or translocation away from the membrane junctions, and the cytokeratin intermediate filament network is reduced or lost while vimentin expression is enhanced. EMT manipulation can be implemented by inducing overexpression of EMT-regulating driver genes, including Twist1 and Snail1 (Mani et al., 2008), and is prominently driven by transforming growth factor beta (TGFbeta). These genes act by transcriptional repression of E-Cadherin. The nuclear factor kappa B (NF-kB) pathway has also been shown to be involved in EMT in MCF10A breast cancer cells. Human breast cancer cell lines are mostly divided into 5 categories based on their characteristics defined in clinical breast cancer datasets. The categories are Luminal A, Luminal B, Basal A, Basal B, and HER2+ types. Hierarchical clustering of high throughput array studies conducted on 34 (Charafe-Jauffret et al., 2006) and 51 (Neve et al., 2006) measuring RNA expression on human breast cancer cell lines grouped those cell lines into Luminal and Basal subgroups. Luminal cells often express estrogen receptor (ER+) and progesterone receptor (PR+) while Basal cell lines lack expression of ER, PR and HER2 (triple negative) and are more resistant to adjuvant chemotherapy. The Luminal group was further divided into Luminal A with low Ki67, a marker of proliferating cells, and Luminal B with high Ki67. The Basal group of cell lines was further divided into 2 groups, Basal A and Basal B (Neve et al., 2006). Most cell lines in the Basal B group have a more invasive phenotype and exhibit a mesenchymal gene signature. We applied a panel of cell lines from those different molecular subgroups: Luminal (MCF7, which has epithelial features), Basal A (MDA-MB-468) and Basal B (MDA-MB-231, which has mesenchymal properties). The EMT features in the PMC42 system include down-regulation of CDH1 and up-regulation of Vimentin for PMC42-ET and PMC42-LA cell lines (Ackland et al., 2001, Ackland et al., 2003). PMC42 system consists of the parental PMC42-ET and the epithelial subtype, PMC42-LA, the PMC42 system provides us with a spectrum of EMT associated changes. The PMC42-LA cell line contains a low number (10-15%) of Vimentin-positive cells, whereas the PMC42-ET cells are 100% Vimentin positive, with commensurate CDH1 differences (Hugo et al., 2007). Further, in response to EGF PMC42-LA cells undergo EMT-like changes (Ackland et al., 2003). ASE in the PMC42 human breast cancer EMT were investigated by (i) comparisons between the more mesenchymal parental PMC42-ET (ET) cells and the more epithelial PMC42-LA (LA) subline, and (ii) in response to epidermal growth factor (EGF), which stimulates EMT-like changes at the mRNA and protein level in both PMC42 variants. We assessed these effects in 2D monolayer culture as well as 3D cultures in Matrigel or Collagen (Vitrogen) and found very similar results in all three culture conditions. ASE are regulated by Epithelial Splice Regulatory Proteins (ESRP) 1 and 2. The expression of both, ESRP1 and ESRP2, was found to be suppressed in Basal B but not in Luminal or Basal A cell lines (Warzecha et al., 2009a). This suggests that their suppression could be involved in EMT-related events. ESRP1 and 2 mRNA levels were constitutively lower in the mesenchymal ET cells compared to LA, but showed little EGF regulation. ESRP1 mRNA levels in epithelial MCF-7 cells were similar to LA, while mesenchymal MDA-MB-231 cells were similar to ET. For ESRP2, MCF-7 levels were higher than LA. Mammalian Ena Homolog (MENA) levels in both PMC42 variants resembled MCF-7 cells, however both variants predominantly expressed the mesenchymal-associated form, as was the case with Cluster of Differentiation 44 (CD44), Ral GEF with PH Domain and SH3 Binding Motif 2 (RALGPS2) and Membrane-associated guanylate kinase, WW and PDZ domain-containing protein 1 (MAGI1). Thus, EMT-associated-ASE revealed predominantly mesenchymal-specific splicing patterns despite the ESRP1 differential, perhaps due to a lack of ESRP2. The results confirm an ESRP1/2-related mesenchymal shift from PMC42-LA to ET cells. In general, we also confirmed a high level of ESRP1 and Fibroblast growth factor receptor 2 – exon IIIb (FGFR2 IIIb) in Luminal and Basal A cells, and reduced level of ESRP1 and higher Fibroblast growth factor receptor 2 – exon IIIc (FGFR2 IIIc) in Basal B cells. The shift from FGFR2-IIIb to FGFR2-IIIc in EMT showed alternatively spliced variants from the same gene, FGFR2. Expression of FGFR2-IIIb measured by splice-specific RT-PCR followed the same pattern as ESRP1, while both PMC42 variants were higher than MCF-7 cells for mesenchymal-associated FGFR2-IIIc. Zinc finger E-box-binding homeobox 1 (ZEB1) mRNA levels, a transcription factor that binds E-box motifs in promoters, were reduced by expression of a short hairpin RNA (shZEB1) in PMC42-ET cells. Lack of ZEB1 resulted in a significant reduction (p=0.0018) of ESRP1, but not ESRP2, consistent with the E-box in the ESRP1 proximal promoter. Although FGFR2 IIIb was upregulated after ZEB1 silencing (p=0.058), FGFR2 IIIc, which was supposed to be alternatively spliced, remained at the same level after ZEB1 silencing (p=0.6263). This suggests a direct role of ZEB1 in ESRP1 expression. Total Enabled Homolog (ENAH) was not reduced significantly after ZEB1 knockdown, (p=0.366). In summary, ZEB1-knockdown in PMC42-ET cells caused enhanced levels of ESRP1 and FGFR2 IIIb expression. Partek Genomic Suite analysis of the Affymetrix data indicated a selective upregulation of a 3’-truncated isoform of Laminin subunit alpha 3 (LAMA3 variant 2 or LAMA3v2) by EGF. qRT-PCR analysis revealed that both the long variant (LAMA3v1) and shorter variant (LAMA3v2) showed enhanced levels along the Luminal to Basal B spectrum (as explained in Chapter 1.1.3.), although the Basal B MDA-MB-231 cells appeared to under-express LAMA3v2. LAMA3v2 was particularly highly expressed in the PMC42 variants and was upregulated by EGF in PMC42-LA cells but not in the PMC42-ET cells. LAMA3v1 levels in PMC42-LA and –ET cells both resembled the MDA-MB-231. In the MDA-MB-468 model of EGF-induced EMT, LAMA3v1 was stimulated by EGF treatment (7 days) but not hypoxia (3 days), whereas LAMA3v2 expression was stimulated by either EGF (7 days) or hypoxia (3 days) treatments. Our group has conducted an experiment by xenografting MDA-MB-468 in mice. In MDA-MB-468 xenografted tumours, LAMA3v2 was expressed significantly higher than LAMA3v1. Gene silencing using small interference RNA (siRNA) techniques provide a sight on the short term effects of knocking down gene(s) in cells. In assessing the effectiveness of silencing subunit alpha 3, subunit beta 3 and subunit gamma 2 of Laminin (LAMA3, LAMB3 and LAMC2), we used short interference RNA (siRNA) targeting LAMA3 (siLAMA3), LAMB3 (siLAMB3) and LAMC2 (siLAMC2) and combination of those three siRNAs (siLAMA3B3C2). On the one hand, inhibition of LAM using the siLAMs, as confirmed with inhibition of LAMB3 LAMC2, and laminin v2, inhibited the expression of EMT markers: Vimentin. On the other hand, siLAM increased expression of Zeb1, ENAH and laminin v1. This may suggest that targeting Laminins using siRNA could reduce the EMT properties of PMC42 cell lines. RNA Seq results showed several top genes as potential candidates for EMT in PMC42 breast cancer cells. We applied Multivariate Analysis of Transcript Splicing (MATS) and Differential Exon Usage in RNASeq (DEXSeq) to identify several top candidates, which have upregulated transcript variants during EMT in the PMC42 system. Those top candidates were inspected using SeqMonk to visualize RNASeq reads. Several transcripts were listed on MATS and DEXSeq results as having one of their exon upregulated after EGF treatment in PMC42-LA, including Ladinin-1 (LAD1), Tenascin-C (TNC), Cleft Lip and Palate Transmembrane Protein-1 Like (CLPTM1L), Serine / Arginin-rich Splicing Factor 1 (SRSF1). In this investigation, only LAD1 showed a pronounced up-regulated transcript variant in SeqMonk visualization. Thus, LAD1 is a good candidate as a target for inhibiting EMT in PMC42 breast cancer cell line.
Combining TIL and CAR for adoptive cell therapy in metastatic melanoma
Background Metastatic melanoma is a highly lethal disease, and until recently patients had limited therapeutic options. Knowledge and understanding of the role the immune system plays in tumour development and its therapeutic potential has recently gained momentum and immunotherapeutic agents have emerged as the gold standard of therapy in treating this cancer. Adoptive cell therapy (ACT) has been shown to have high rates of tumour regression with durable, complete responses and potential 'cure'. Tumour-infiltrating Lymphocytes (TIL) and Chimeric Antigen Receptor (CAR) therapies are examples of ACT. Each has their own advantages, limitations and toxicities. As the complexity of the immune system and its targets is increasingly appreciated, combining immunotherapies is emerging as a promising avenue for improving patient oncological outcomes. This project explores the efficacy of dual specific T cells by combining TIL and CAR therapies. Aim To establish a model system transducing TIL with anti-Her2 CAR (TIL-CAR) and assessing function against autologous melanoma tumour cells that express Her2 antigen. Method TIL were generated from patient derived metastatic melanoma tumours and tumour cell lines were established in a biobank. TIL were thawed and activated using CD3/28 beads and transduced with second generation anti-Her2 CAR (scFv-erbB2-CD28-zeta) using a retronectin protocol. Patient matched PBMCs were transduced for functional comparison. Melanoma tumour lines in the biobank were found to innately express Her2 antigen to varying degrees. Some melanoma tumour lines were transduced and sorted to create higher expressing Her2 antigen lines for functional comparison. Flow cytometry was used to confirm cell phenotype and antigen/CAR expression. Functional testing was performed using ELISA and chromium release assays. An in vivo ACT model in NSG mice was performed comparing TIL and TIL-CAR. Results TIL were successfully cultured from metastatic melanoma tumour pieces. Despite TIL proliferating at lower rates than PBMCs, both were successfully transduced to express anti-Her2 CAR. When TIL were transduced to express anti-Her2 CAR they were functionally active through both TCR and CAR and produced greater amounts of interferon gamma against Her2 expressing tumour lines. TIL-CAR had greater cytotoxic activity when cultured against autologous melanoma tumour lines, but the benefit transduced TIL over PBMCs varied in response between tested patients. The advantage of TIL-CAR over PBMC-CAR did not demonstrate consistent trends across this limited group of patients. The functional activity may be influenced by the level of Her2 expression in the co-cultured tumour cells as well as by the phenotype of T cell populations. Results of an in vivo pilot study in mice demonstrated reduction in tumour size when TIL-CAR were used in an ACT protocol. The primary limitation of this study was the low proliferation rate of TIL following transduction which required extended periods in culture. Conclusions Combination of TIL-CAR is a novel concept. TIL can be transduced to express anti-Her2 CAR. Metastatic melanoma cells in our biobank constitutively express Her2 antigen. TIL-CAR tend to show greater activity in interferon gamma and cytotoxic functions compared to parental (non-transduced) TIL when cultured against Her2 expressing tumour lines. When compared to the activity of PBMCs transduced with the same CAR the additional benefit conferred by TIL-CAR is inconsistent. Protocols would benefit from further optimisation to generate a 'younger' phenotype capable of more rapid and sustainable proliferative potential and facilitate earlier delivery of therapy if used in a clinical setting.
Strategies for engineering skeletal muscle: an important link in the neuroprosthetic interface of bionic limbs
Limb amputation is a major cause of disability in our community, for which motorised prosthetic devices offer a return to function and independence. Advanced robotic limb technology utilises a range of mind-prosthetic interfacing strategies to intuitively drive the limb. These approaches include direct recording of peripheral nerves, brain-recording interfaces, or the transposition of transected nerves to remaining muscle groups for myoelectric recording. All of the current methods are hampered by delicate neural biology, either leading to premature device failure or introducing unnecessary surgical risk. As an alternative, this thesis proposes a new solution: to develop a bioelectrode using tissue engineered skeletal muscle as a signal amplifier of activity from residual nerves for intuitive prosthetic control. Conceptually, the fabrication of such a device would begin with a tissue biopsy from the patient from which a pool of myogenic stem cells would be derived and expanded. These autologous cells would be used for the tissue engineering of skeletal muscle fibres, primed with neurotrophic biofactors to optimise the tissue for innervation. Flexible recording wires could be incorporated into this fabrication step, thus eliminating the trauma of electrode insertion and also optimising its biocompatibility. The bioelectrode device could also be designed to patient or prosthetic anatomy as required. This thesis developed key elements of the above proposal. Firstly, a bioprinting technique was established to tissue engineer functional skeletal muscle using a gelatin methacryloyl (GelMA) bioink. Bioprinting enabled the rapid deposition of muscle progenitor cells (primary mouse myoblasts) in layered fibres, reminiscent of native muscle architecture. Fabrication parameters were optimised to produce fibres with high cell viability and print resolution. These bioprinted constructs were then able to support advanced maturation of cells into multinucleated muscle fibres, as evidenced by molecular analysis and functional testing. There was a significantly greater upregulation of genetic markers of myogenesis when compared to monolayer myotube cultures, and this result was complemented with functional testing that demonstrated mature patterns of calcium handling and electrical activity. The bioprinted muscle was then implanted in the nude rat to assess its capacity for innervation and vascularisation. The tissue construct was implanted in an in vivo chamber, which was supplied by a surgically formed arteriovenous loop and transected nerve. After only two weeks, independent bundles of mature muscle fibres had developed, with histological evidence of neural integration and vascularisation. In vivo electrophysiological studies confirmed the presence of innervation by demonstrating muscle activity in response to neural stimulation. Lastly, the triad of bioprinted muscle, vasculature and nerve was housed in a customised 3D printed chamber as a prototype for a bioelectrode that could be surgically grafted onto transected peripheral nerves after limb amputation. To conclude, this thesis developed the principle elements of designing a bioelectrode for neuroprosthetic interfacing and provides the foundation for further tissue optimisation and integration of electrodes. Although the work presented is in the frontier stages of development, it offers an exciting glimpse into the future of modern medicine and brings the dream of mind-controlled motorised prosthetic limbs closer to everyday reality.
Assessment of the anabolic effects of PTH drug treatment and mechanical loading on bone using high-resolution imaging and in silico modelling
Osteoporosis (OP) is a progressive bone disease characterised by significant reduction of bone mineral density (BMD) due to loss of bone matrix and changes in bone tissue properties. OP is regarded as a worldwide health issue and identifying novel treatments is of central clinical importance. Daily injections of parathyroid hormone (PTH) and exercise have been proven to have an anabolic effect on bone, i.e., are capable of restoring bone mass. In this thesis, the anabolic action of PTH drug treatment and mechanical loading was investigated using in silico modelling and high-resolution imaging techniques. Novel drugs are continuously developed to reduce the risk of bone fractures in osteoporotic patients. PTH peptides such as PTH(1-34) are the first anabolic agents approved to treat severe OP. Despite its success to restore bone mass, PTH mechanism of action on bone cells is still unclear. Recently, the understanding of OP pathophysiology has considerably improved. Biomarkers reflecting bone physiology have been identified at cellular, tissue and organ levels. Cellular biomarkers reflect the dynamics of bone remodelling on a short time scale, whereas tissue and organ scale biomarkers show changes of BMD on a larger time scale. Computational modelling is a novel approach that allows to quantitatively characterise the effect of a drug treatment on the disease progression integrating physiology, disease progression, drug treatment and biomarker data in a comprehensive mechanism-based in silico model. In this context, part of this work was focused on the development of a full time-dependent mechanistic pharmacokinetic-pharmacodynamic (PK/PD) model of the action of PTH(1-34) on bone modelling and remodelling. This model was applied to rat models of OP to shed light on the inter-cellular and tissue scale mechanisms involved in the action of PTH(1-34) on bone cells. This in silico model has the potential to predict the long-term effects of drug treatments on clinical outcomes and provide a means for patient-specific estimation of bone fracture risk. Furthermore, it is well known that bone adapts its mass and structure in response to stresses and strains induced by an external mechanical load. The most extensively used animal model to test hypotheses related to mechanical loading is the in vivo axial compression of the mouse tibia. Common outcome measures of these models are bone geometric dimensions and bone mineral density using high-resolution imaging techniques, i.e., micro-computed tomography (micro-CT). In this thesis, end-point micro-CT imaging data were analysed to quantify the local adaptation response of bone to both mechanical loading and PTH(1-34) drug treatment in the mouse tibia loading model. An innovative image post-processing algorithm was developed to quantify the cortical thickness locally along the periosteum. Furthermore, an algorithm was developed to estimate stresses, strains and strain energy density (SED) on periosteal surfaces of the tibia, combining micro- finite element analysis and beam theory to compute animal-specifi c SED. Bone adaptation to mechanical loading was variable along the periosteum. Results suggest that bone adaptation is higher in regions with higher SED. Moreover, mechanical loading and PTH induce a combined anabolic adaptation effect on bone suggesting that the association of PTH(1-34) administration and exercise may be an effective treatment for OP.
Identifying functional drivers of epithelial-mesenchymal transition (EMT) in human breast cancer: the integrin/ILK axis
Breast cancer is the leading cause of cancer in women worldwide, and over 90% of deaths caused by breast cancer are due to metastases, many of which are not responsive to current therapies. The ability for cells to acquire a metastatic phenotype includes epithelial mesenchymal transition (EMT), invoked as a critical component of the metastatic cascade. During the process of EMT, epithelial cells undergo a temporary conversion acquiring molecular and phenotypic changes that facilitate the loss of epithelial features, and the gain of mesenchymal phenotype. Such transformation promotes cancer cell migration and invasion. EMT is typically characterized as a loss of the epithelial cell adhesion proteins E-cadherin and cytokeratins, coupled with the gain of mesenchymal-associated molecules N-cadherin and vimentin. However, these proteins may not always be present in cancer systems. For example, one of the limitations in the use of vimentin as a prognostic marker in breast carcinomas is the likelihood that vimentin-positive cells may have migrated away from the primary mass, and become buried in the surrounding stroma, which is also vimentin-positive. Therefore, the identification of new markers which better represent EMT in breast carcinomas, and allow for a more specific detection of EMT-derived or EMT-prone breast cancer cells in the tumour vicinity, could have a dramatic impact on breast cancer prognosis. The work presented in this thesis describes a comprehensive characterization of two human breast cancer EMT model systems: the in vitro PMC42 cell system and the in vivo EDW-01 patient derived xenograft system. Specifically, the focus of this project was to perform a sequence of studies to assess the regulation of α2 and β1 integrin (ITGα2, ITGβ1), and ILK. The functional role of the integrin/ILK axis in the mesenchymal state, and in the epithelial-to-mesenchymal transition is explored and assessed.
Using transcriptomics to understand cancer progression and predict response to therapy
Transcriptomics data provide useful information to better understand molecular phenotypes in cancer. Epithelial-to-mesenchymal transition (EMT) is one of these molecular phenotypes that is hijacked by cancer cells to obtain mobile mesenchymal characteristics which may assist cells to intravasate into blood stream, generate circulating tumour cells (CTCs) and metastasize to distant organs. CTCs also have heterogeneity in their molecular phenotypes and it is of utmost importance to understand these variations to be able to understand differences in their therapy response and use them to monitor treatment outcome. Using transcriptomics, we can also explore and predict molecular phenotypes associated with sensitivity to different therapeutic regimen. Although EMT is a single molecular phenotype, it can be regulated through different underlying molecular mechanisms, leading to differences in response to therapies. To identify samples with TGFβ-driven EMT, I derive a gene expression signature of EMT induced by TGFβ using metaanalysis and transcriptomics data integration. This signature is able to identify transcriptional profiles arising during TGFβ-driven EMT, and yields highly consistent results in multiple independent pan-cancer cell lines and patients data. Samples fitting this signature show lower number of mutations in elements of TGFβ signalling, poorer overall survival outcome and preferential response to certain drugs. Meta-analysis and data integration such as the above require careful attention to batch effects in datasets. I apply different batch correction methods in order to perform general normalisation or obtain differentially expressed genes (DEGs) in integrated transcriptomics data sets. Further, to classify the fit of individual samples to a gene signature, I apply existing single-sample scoring methods. However, these methods all use information borrowed from the whole set of samples, meaning they are not truly single sample scores. To address this, I developed a rank-based scoring method, called singscore, which generates more stable scores that are independent from sample size and composition in a dataset. CTCs are integral to cancer progression, but while these cells are extremely rare in blood, they have great potential to provide a real-time representation of cancer progression and treatment efficacy. I perform an assessment of current markers for enrichment and/or detection of CTCs, and then, introduce new CTC markers, including general, epithelial and mesenchymal markers obtained by analysing multiple breast cancer and blood data sets. I then assess their expression in publically available CTC data and a number of in-house patient samples. Finally, I use pharmacogenomics data in breast cancer cell lines and the singscore method to predict drug response outcome for 90 drugs based on gene expression data, which have been shown to be the most predictive molecular feature in breast cancer. I derive drug sensitivity signatures by quantifying associations between gene expression and drug response and evaluate the utility of these gene signatures using cell lines, PDX models and patient data and show consistent pattern of response across independent data sets. Further associations between drug sensitivity scores and EMT phenotype are assessed.
Analysis of C-peptide-specific CD4+ T cells in the peripheral blood of people with type 1 diabetes
Uncovering the primary antigen targets in type 1 diabetes (T1D) is essential to our understanding of disease pathophysiology. Despite the clear role of CD4+ T cells in orchestrating the immune destruction of the pancreatic cells, what they are targeting in human T1D has remained poorly defined. Most knowledge of in vivo T-cell responses in T1D derives from studies in mouse models, and translating results to humans has been limited to analysis of peripheral blood. However, only 3% of the total T cells in the body reside in the peripheral blood. Prior work at this institute by Mannering and colleagues on islet-infiltrating CD4+ T cells in humans, complemented by other similar studies, provided insight into the resident T-cell population of the target organ in subjects with T1D. These studies have concurred that a cleavage product of proinsulin, C-peptide, is a target antigen of islet-infiltrating CD4+ T cells. Because these studies were done in just a handful of deceased organ donors with T1D, they led to the question how relevant is C-peptide as an autoantigen in T1D more generally? Given the pancreas is not routinely accessible, to address this question, evidence of C-peptide as a target of CD4+ T cells was sought from the peripheral blood in subjects with T1D. The main obstacle to assessing T-cell targets in the peripheral blood is the lack of a sufficiently sensitive and reproducible T-cell assay. In Chapter 3, the CFSE-based proliferation assay was optimised for detection of C-peptide-specific CD4+ T-cell responses. The CFSE-based proliferation assay was demonstrated to have comparable reproducibility as compared to other currently available T-cell assays, and greater sensitivity than the commonly used ELISpot assay. In Chapter 4, using the CFSE-based proliferation assay, >60% of people with recent-onset T1D were shown to have a detectable peripheral C-peptide-specific CD4+ T-cell response. The response was disease specific because few control subjects were positive. Analysis of cloned C-peptide-specific CD4+ T cells revealed they were restricted by HLA alleles strongly associated with T1D risk, namely HLA-DQ8, -DQ2, -DQ8trans, -DQ2trans and HLA-DR4. This added further support to the notion that they were pathogenic. In Chapter 5, the hypothesis that autoantibodies to C-peptide may be detected in the serum of people with T1D, was tested. Using solid-phase ELISA, it was found, unlike C-peptide-specific CD4+ T cells, C-peptide autoantibodies are not detectable in the serum of subjects with T1D in a disease-specific manner. Together, these findings indicate that proinsulin C-peptide is commonly a target of autoreactive CD4+ T cells in newly-diagnosed T1D. Hence, C-peptide is a promising candidate for biomarker development and antigen-specific immunotherapy.
Development and validation of a novel marker tracking approach based on the low-cost Microsoft Kinect v2 sensor for assessing lower limb biomechanics during single-leg squat and treadmill gait
Pubescent females are twice more likely to suffer a non-contact ACL injury than their male counterparts. This disparity has been correlated with multiple concurrent factors, including biomechanical, anatomical and hormonal changes. ACL ruptures require serious and costly surgical interventions, which could be avoided if subjects at higher risk of injury were more carefully monitored and trained. Three-dimensional motion analysis is required to identify individuals at risk of ACL injury. Multi-camera optical systems are the gold standard for 3D motion capture, but they are very expensive and cumbersome. The aim of this thesis was to make motion analysis more accessible, developing an affordable and compact 3D motion tracking methodology, alternative to conventional multi-camera systems. A novel tracking approach was developed using Microsoft Kinect v2, employing custom-made coloured markers and computer vision techniques. This methodology was denoted as Kinect coloured marker tracking (KCMT). The accuracy of KCMT relative to a conventional Vicon motion analysis system was measured performing two Bland-Altman analyses of agreement, the first using single-leg squat (SLS) as benchmark task, the second using treadmill locomotion. The objective of the first study was to determine if KCMT-derived sagittal joint angles of the lower limb were accurate enough to allow discerning individuals at risk of ACL injury from those not at risk. Eleven healthy participants were asked to perform three SLS trials, while three-dimensional marker trajectories were simultaneously recorded using Vicon and KCMT respectively. Joint angles from the two systems were calculated via inverse kinematics using OpenSim. The limits of agreement (LOA) of the joint angles were −16°, 13° for hip flexion, −12°, 0° for knee flexion and −12°, 9° for ankle flexion. These results indicated that the agreement between KCMT and Vicon was joint dependent, and that further work was required for the novel methodology to replace conventional marker-based motion capture systems for the identification of ACL injury risk from SLS data. In the second study, an improved data collection protocol for the KCMT was used. Twenty participants were recruited, and markers placed on bony prominences near hip, knee and ankle. Three-dimensional coordinates of the markers were recorded during treadmill walking and running. The LOA of marker coordinates were narrower than −10 and 10 mm in most conditions, however a negative relationship between accuracy and treadmill speed was observed along Kinect depth direction. LOA of the knee angles measured in the global coordinate system were within −1.8°, 1.7° for flexion in all conditions and −2.9°, 1.7° for adduction during fast walking, suggesting that KCMT may be capable of discerning between subjects at risk of ACL injury and controls. The proposed methodology exhibited good agreement with a marker-based system over a range of gait speeds and, for this reason, may be useful as low-cost motion analysis tool for selected biomechanical applications.
Profiling desmoid tumours and exploring new strategies to prevent and treat desmoid tumours in familial adenomatous polyposis using a novel mouse model
Desmoid tumour is a benign growth that causes morbidity and mortality from local enlargement. Desmoids are thought to result from dysfunction in WNT signalling. Trauma, including that from surgical intervention, is implicated in the pathogenesis of desmoid tumour, with desmoids often occurring at surgical sites. Patients with familial adenomatous polyposis (FAP) possess a germline mutation in the APC gene on chromosome 5, resulting in numerous colorectal adenomas that inevitably develop into colorectal carcinoma. Patients with FAP undergo prophylactic colectomy to manage risk of future carcinoma. The germline mutation in APC, combined with surgical trauma from prophylactic colectomy, place patients with FAP at a significant risk of future desmoid tumour development. Desmoid tumour is the largest cause of mortality for patients with FAP following prophylactic colectomy. There is controversy regarding the utility of laparoscopic prophylactic colectomy in FAP, with some series suggesting that laparoscopic surgery results in a higher risk of future desmoid tumour. Desmoid tumours in FAP patients occur in the mesentery and abdominal wall. Management options are limited due to the precarious location, often in close relationship to the mesenteric vasculature. Desmoid tumours in this cohort pose significant clinical management challenges, with high recurrence rates after surgical excision, and no consensus on best medical management. Since desmoid tumours are rare tumours, a pre-clinical model would facilitate research in this area. This research thesis describes the development of a novel murine model for abdominal desmoid tumour that occurs in FAP. The Apcmin/+:p53-/- mouse develops numerous abdominal wall desmoid tumours. This model has been validated with histopathology and immunohistochemistry, and has facilitated the development of desmoid tumour cell lines, and xenograft studies in this area. The impact of surgical approach on future desmoid risk was investigated using the Apcmin/+:p53-/- mouse. Mice were subjected to laparotomy or laparoscopy and were then observed until they reached ethical endpoints, at which time an assessment of desmoid tumour burden was made. In the Apcmin/+:p53-/- mouse model, surgical approach had no impact on survival or the number of macroscopically identifiable desmoid tumours. Furthermore, the use of humidification/warming device for open and laparoscopic surgery was trialed and found to have no impact on survival or desmoid tumour burden. This research has also investigated the genomic landscape of human abdominal and abdominal wall desmoid tumours through RNA sequencing and immunohistochemistry. This study has identified that abdominal desmoid tumours share genomic similarity to wild-type gastrointestinal stromal tumours (negative for CD117 and PDGFRα). Sequencing identified a number of pathways (Such as VEGF, EGF and mTOR) involved in desmoid tumour formation that could be targets for therapy.
The impact of sex differences on host and tumour prognostic factors in patients with non-small cell lung cancer
Epidemiological studies demonstrate that women live longer than men following diagnosis of non-small cell lung cancer (NSCLC) even after controlling for prognostic factors. This study examines the one biological trait all patients have, sex, and seeks to understand and generate new knowledge with respect to the impact of sex on survival. Patient and tumour characteristics of women and men are examined from presentation and diagnosis, to staging, with specific focus on the impact of these factors on outcomes in NSCLC treated surgically with curative intent. The study also explores how to include sex differences in medical research more generally. Aim and research question The aim is to examine the impact on survival of the association of sex with the following recognised prognostic patient and tumour characteristics: • age • performance status • smoking history • positron emission tomography maximum standardised uptake value • tumour, node, metastasis (TNM) staging The research question is: “What is the impact of sex on validated and putative prognostic factors in non-small cell lung cancer and how can we translate understandings in sex differences in lung cancer to facilitate a more targeted research and therapeutic approach to improve patient outcomes?” This is answered by sex disaggregated analysis of the prognostic value of host and tumour factors from a detailed clinical dataset, including patient outcomes, and compared to an independent population level validation dataset. Finally, I examine international examples of success in sex differences medical research more broadly, and the policy landscape that is preventing translation of both the findings in lung cancer, and sex differences in health and disease across the health care continuum in Australia. Method A detailed surgical database was developed from patients treated surgically with curative intent from the Peter MacCallum Cancer Centre and St Vincent’s Hospital Melbourne from 2000-2010. An additional cohort of patients was identified and analysed from the Surveillance, Epidemiology and End Results (SEER) database, the US population level database. The SEER database was matched to the Melbourne cohort with respect to surgical treatment with curative intent and date of surgery, to ensure continuity with clinical protocols. Extensive clinical data were collected to allow analyses of the impact of sex differences on prognostic variables in three key areas: 1. Host factors (sex, age, smoking and performance status) 2. The maximum standardised uptake value on Positron Emission Tomography 3. TNM staging The patient outcome was disease specific five-year survival. During this research, a new edition of TNM staging was developed resulting in two different editions of TNM staging being used, with the 7th edition in chapter 3 and 4 and 8th edition in chapter 5. Data were analysed with IBM SPSS Statistics software (SPSS) version 21 (Chapter 4), version 22 (Chapter 3) and versions 25 and R (Chapter 5). Findings There are biological differences between women and men in the disease process in early NSCLC. These differences span elements from patient characteristics to survival outcomes. Irrespective of the parameter examined, male sex was a consistent negative prognostic factor. However, the prognostic value of previously identified tumour and host characteristics was equally valid for men and women. Whilst the finding of poorer survival in men with NSCLC is not new, researchers and clinicians have assumed that this was because women are less likely to smoke; due to ethnicity (there is a distinct variant of NSCLC which is less aggressive and occurs more commonly in women of Asian descent); because women are more likely than men to get adenocarcinoma, with its better survival profile compared with squamous cell cancer; and that women are likely to seek treatment at an earlier stage. The research presented in this thesis demonstrates that these assumptions do not fully explain the observed survival differences. Whilst the observations are correct superficially, the causality is false. Survival differences between men and women persist irrespective of ethnicity, histology and disease stage. These findings have important implications for research design, translation, clinical guidelines and practice. Significance A better understanding of the impact of patient and tumour sex on tumour characteristics has the potential to improve understanding of the biology of lung cancer and may lead to different staging and treatment approaches. Understanding the impact of sex on patient response to treatment may improve patient outcomes for men and women by improving selection of therapies tailored to each sex. In addition, novel targets for therapeutics may be identified. This thesis presents a comprehensive delineation of the differences in survival between men and women with NSCLC. These differences raise important questions with respect to the accuracy and efficacy of TNM staging when applied to clinical decision-making for both sexes and indicate that current therapeutic decision-making driven by stage should be reviewed. There is a significant possibility that the oncological community may be over-treating women and under-treating men. Publications from this study may provide direction for future clinical trials, redefine staging, assist in changing clinical practice by more accurate prediction of tumour behaviour, and support the move to individualised treatment.
Radiation-associated breast, thyroid and solid malignancies in patients attending the Peter MacCallum Cancer Centre Late Effects service
Background: Survivors of childhood, adolescent and young adulthood (CAYA) malignancies have an increased risk of subsequent primary malignancies, particularly after exposure to therapeutic radiation. The Peter MacCallum Cancer Centre Late Effects (PMCC LE) service provides individualize, multidisciplinary care and surveillance advice for survivors of malignancies, especially CAYA malignancies. Methods: A retrospective review was performed of patients exposed to therapeutic radiation attending the PMCC LE service from 1st January 2000 to 20th February 2013. All invasive malignancies, in-situ malignancies, benign tumours and deaths were evaluated. Separate time-to-event analyses was performed for radiation-associated breast, thyroid and solid malignancies in patients exposed to chest, thyroid and any therapeutic radiation respectively, measured from the date of first attendance to the PMCC LE service and stratified by the interval from completion of radiation to the first attendance. The incidence of breast and thyroid malignancies was compared to the Australian general population. Compliance with breast and thyroid surveillance recommendations was determined by assessing the number of screen events over the period of attendance to the service. Clinicopathological features and management of radiation-associated breast and thyroid and other solid malignancies was examined. Ultrasound and cytological workup of radiation exposed thyroid nodules was assessed. Results: After excluding 187 patients, 534 included patients developed 194 invasive malignancies; 147 were radiation-associated and 47 non-radiation associated. The most common malignancies were non-melanoma skin (37.1%), thyroid (17.0%) and breast (12.9%) malignancies. Patients whose first attendance was ≥15+ years after radiation exposure experienced the highest incidence of radiation-associated breast, thyroid and solid malignancies, with 23%, 8% and 27% affected after 10 years of subsequent follow-up respectively. The incidence of breast and thyroid malignancy was elevated 11.2 and 57.6 times respectively compared to the Australian general population (both p<0.001). Compliance with breast surveillance using mammography or any screening modality was observed in 18.4% and 28.6% of women at risk respectively. Twenty-eight radiation-associated breast malignancies occurred in 24 women (16.7% bilaterality). Breast malignancies diagnosed after the first attendance to the PMCC LE service were more likely screen-detected (p=0.002). Most were hormone receptor positive (87.5%), invasive ductal carcinomas (82.1%) managed with mastectomy (89.3%). Compliance with thyroid surveillance was observed in 76.9% of patients at risk. Ultrasound features of microcalcification and increased internal vascularity had a low sensitivity (62.5%) for predicting a malignant nodule, which improved when used in conjunction with a Bethesda IV-VI result (91.7%), although cytological assessment was not performed in 45.6% of operative cases. Thirty-three patients had a radiation-associated thyroid malignancy; 45.4% (n=15) were incidental. The majority were papillary thyroid cancers (88.9%); of which 12.5% were node positive and 34.4% were multifocal. Node positive thyroid cancers were more likely to present symptomatically (p=0.03). There were 36 deaths in the cohort (6.7%), most commonly attributable to radiation-associated malignancies (41.9%), especially brain, breast and sarcomatous malignancies. Conclusions: Patients attending the PMCC LE service have a high burden of subsequent malignancies that typically occur after a long latency. Ongoing long-term surveillance is essential and judicious management with adherence to guidelines is advocated in this unique population of patients.