Systems Pathology Group

Introduction

Pathology and molecular pathology are crucial for correct diagnosis of cancer and delineation of subtypes. This is essential for prognosis, but is more applicable to groups of patients than an individual. Newer techniques, such as Her2 testing, give some hope that we will also be able to predict the response to treatment in individual cases. Our approach is to harness and maximise the use of pathology in complementing basic science and translating to the clinical setting.

My particular interest is in developing image analysis and the development of better quantification techniques in histopathology, combined with dynamical experiments in cell culture. By combining both approaches it is possible to model some of the key biochemical and signalling pathways involved in breast cancer treatment and drug resistance. Dependent upon mathematical techniques employed this may be both a discovery tool and also a predictive tool for a single cancer specimen.

Overview

In order to understand how genetic and epigenetic abnormalities in cancer impact on therapeutic response, we are using a combination of in vitro experimentation, studies on cancer specimens and advanced computational modelling to make predictions of clinical responses to therapy in breast and ovarian cancer. New mathematical models exploit what is known about signalling networks and therapeutic mechanisms to produce a dynamical system, which can be personalised and perturbed to allow simulation of the natural history of a tumour and responses to therapy.

We work closely with two collaborating groups: Igor Goryanin's team in Edinburgh Centre for Systems Biology, and John Crawford's team in Simbios, Dundee. We have developed a kinetic model of receptor tyrosine kinase signalling (in collaboration with the Centre for Systems Biology, Edinburgh), which makes predictions about how breast and ovarian cancers respond to their microenvironment, and how PI3K activation and PTEN loss modify RTK-inhibitor responses. We are performing detailed pathway analysis using high-throughput reverse phase protein arrays to generate kinetic data from in vitro models, in vivo models, and tissues, and our aim is to validate new hypotheses within breast and ovarian cancers using multiplex immunofluorescence. These predictions are providing new approaches to understanding how treatments can be combined, and how molecular pathology can be improved in order to individualise therapy.