This project will involve the identification of biomarkers or molecular methods that can help resolve the diagnosis of metastatic tumours of uncertain origin
Defined by the absence of confirmed cancer type diagnosis, cancer of unknown primary (CUP) is a devastating disease with an exceptionally poor outcome. Developing better diagnostic methods is therefore important to direct optimal therapy.
Gene-expression profiling has been used by our laboratory to develop tissue of origin diagnostic tests that can help resolve the origin of CUP tumours.
These tests have an accuracy of approximately 80-85% across major cancer types. However, one group of tumours that can be difficult to classify using gene-expression profiling are cholangiocarcinomas, a cancer that arises from the bile ducts.
Indeed, the distinction between adenocarcinomas of the pancreas, cholangiocarcinomas and other metastatic carcinomas are known diagnostic dilemmas for anatomical pathologists, where immunohistochemistry can be of limited value but where an accurate diagnosis can be critical, as treatment can be very different based on the cancer type.
The identification of biomarkers or molecular methods that can help resolve these tumours is therefore required.
This project will involve three aims:
- Aim 1: collation of published gene-expression datasets representing hepatocellular carcinomas, bile duct cancers (intra and extra-hepatic cholangiocarcinoma) and pancreatic adenocarcinomas. A systematic analysis will be conducted to harmonise the datasets and visualise the data in a multi-dimensional space
- Aim 2: Gene sets that can distinguish between these cancer types will be derived and machine learning will be employed to develop a classifier that will be tested on independent clinical samples analysed using RNA-seq
- Aim 3: gene sets will be surveyed to identify putative immunohistochemistry (protein) or in situ hybridisation (RNA) targets that can be used by anatomical pathologists in a routine histopathology setting.
Contact and more information
Dr Richard Tothill
Dr Andrew Pattison
Rare Disease Oncogenomics Group
University of Melbourne Centre for Cancer Research and Department of Clinical Pathology
Level 10, Victorian Comprehensive Cancer Centre