Research

The model systems used are generally preclinical models of pancreatic cancer. These include cell line models, human tumour xenografts and, through collaboration with David Tuveson (CRI), the KPC GEM (genetically engineered mouse) model. We developed a novel LC-MS assay for use in that model to assess the PK and activation of gemcitabine in pancreatic tumour tissue. This assay is suitable for clinical use and collaborative clinical studies are being planned. We will also modify the method to quantify gemcitabine metabolites incorporated into tumour tissue DNA. In view of the poor drug delivery to pancreatic cancer tissue, a joint PhD student (with the Tuveson laboratory) is investigating the biology of the hypovascular phenotype in KPC tumours, in order to manipulate tumour/vasculature/stromal interactions, to improve drug efficacy.

Our work on the PK and PD of the fluoropyrimidine capecitabine continues. The metabolism and delivery of active capecitabine metabolites to tumour tissue has been identified in pancreatic cancer allografts and the KPC autochthonous mouse model. Capecitabine inhibited tumour growth in the allograft model and efficacy in the KPC model is currently being tested. Combination strategies, using capecitabine, are being evaluated in in vitro models, prior to commencing in vivo studies.

In collaboration with John Griffiths (CRI), a study in patients is assessing the feasibility of non-invasive magnetic resonance fluorine spectroscopy, to assess the metabolism and accumulation of capecitabine and its metabolites in normal liver and breast tumour tissue. Clinical studies are in progress to assess the PK of capecitabine in specific sub-groups of patients following surgery: gastrectomy in gastric cancer patients and Whipple's procedure in pancreatic cancer patients. Each of these studies incorporates a pharmacogenetic assessment of patients' capecitabine metabolising capacity, which will be linked to the PK data. In addition, laboratory studies are underway to characterise the impact of specific genetic polymorphisms in fluoropyrimidine metabolising enzymes.

In the last year, the EPCTT has completed a Phase II trial, assessing a novel vaccine approach for colorectal cancer. Nine studies are currently recruiting patients. In addition to our capecitabine studies, the following studies are enrolling patients: three phase I trials of novel drug combinations (in patients with breast, CNS and pancreatic cancers), a phase I trial of a novel growth factor receptor inhibitor, a phase I trial of an immunotherapeutic in patients with melanoma, and a biomarker study of an anti-angiogenic therapy for renal cancer. This latter study includes both PET and MR based PD studies, highlighting our close collaboration with researchers developing these emerging technologies. Six further protocols (four phase I) are in development, for initiation in 2011. We are collaborating with Tim Eisen (Department of Oncology), hosting a Hales Clinical Fellow, who is identifying functional defects in the VHL tumour suppressor protein caused by mutations detected in renal cancer, and linking these to patient outcome.

Figure 1
The effect of different drugs on the cell cycle of cancer cells are tested alone and in combination. To achieve this, the oscillation of proteins and potential biomarkers are measured during the cell cycle. These oscillations are incorporated into a mathematical model of the cell cycle which acts as a knowledge repository and allows the identification of potentially synergistic drug combinations. Here, HeLa cells are stained for Cyclin E, Cyclin B1, and their DNA content. The measured oscillation of the two proteins relative to each other is used to parameterise the mathematical model (along with data for other proteins), allowing the subsequent investigation of the effect of drugs.

Aurora kinases (AK) have been identified previously as potential targets for anticancer therapeutics. The aurora family of serine/threonine protein kinases plays a critical role in cell division, with key roles in the mitotic spindle checkpoint. Deregulation of AK expression or function is believed to provoke genetic instability and AK-A has been identified as a cancer susceptibility gene. Elevated expression levels of AK are detected in a high proportion of melanoma, colon, breast, ovarian, gastric and pancreatic tumours and our collaborators have shown over-expression of AK-A in urothelial tumours (Veerakumarasivam et al., Cell Cycle 2008; 7: 3525).

In the lab, we have been comparing a number of AK inhibitors with different potency/specificity and are now proceeding to test rational combinations of similar agents in vitro, in selected cancer cell lines. Ashok Venkitaraman (Hutchison/MRC Research Centre) has shown that AK-A over-expression (at levels found in tumour cells) overrides the spindle assembly checkpoint signal, inducing resistance to taxanes in vitro (Anand et al., Cancer Cell 2003; 3: 51). We are investigating the interaction surfaces of AK inhibitors and paclitaxel combinations in pancreatic and urothelial cell lines and have preliminary data that confirm that these interactions are complicated, with regions of synergy and regions of antagonism. It will be critical to understand these in order to optimise the dose scheduling for clinical trials. Assessments of cell sensitivity are being linked to measurements of cytokinetics and gene expression, developing assays that can be used as potential PD markers. Effective drug combinations will be tested in appropriate xenograft and/or transgenic models of cancer.

In general, it is assumed that combinations of agents have similar effects on normal and tumour cells, but this is not always the case. The optimal combination would show synergy in cancer cells and antagonism in normal cells, reducing the toxic side effects that often limit dosing. We are investigating which of the available normal diploid cell types are most relevant for these comparisons. This project also incorporates mathematical modelling approaches, with a model of the cell cycle and spindle assembly checkpoint being used to predict the effects of the drugs. The mathematical models are also being linked with PKPD modelling approaches, to allow the prediction of the efficacy of certain drugs/combinations/dosing schedules in vivo (a collaboration with Bob Jackson, Pharmacometrics). We will extend our pre‑clinical findings into clinical trials, incorporating the biomarkers we identify.

We also have collaborations with Steve Ley and Rebecca Myers (Department of Chemistry) and Fanni Gergely (CRI). We are co-supervising a Cancer Research UK Medicinal Chemistry Programme PhD student, who is synthesising and evaluating selective allosteric inhibitors of the kinesin motor protein HSET, a promising new anticancer target. We have also established a mouse xenograft model of disseminated ovarian cancer, in which tumour growth is monitored by bioluminescent detection of the luciferase-expressing tumour cells in the abdomen, to support a collaboration with Gillian Murphy (CRI), investigating TNF-alpha converting enzymes (TACE) as therapeutic targets.