Research

Translational breast cancer genomics: applications of molecular profiling in prognosis, prediction and novel therapeutics

The detailed description of the complex molecular taxonomy of breast cancer is the first step towards robustly identifying markers that have true clinical utility.

We have now characterised the genomic and transcriptomic landscapes of 2000 breast cancers using high-resolution SNP arrays (Affymetrix SNP6.0), gene expression arrays (Illumina) and focused mutation analysis (TP53). This has revealed a more complex taxonomy than that gleaned from expression analysis alone. By implementing joint copy number/expressionbased clustering we have identified at least 10 subtypes of breast cancer with markedly different clinical outcomes. Most of these subtypes robustly capture the copy number-driven events that dominate the landscape (ERBB2, Cyclin D, 1q, 8p12, 8q24, 20q13). Moreover the high resolution of the arrays used and the large number of tumours profiled allow precise mapping of copy number events, as illustrated with the ERBB2 amplicon (Figure 1). We have also identified subtypes with lower genomic instability where expression is driven by copy number‑independent events. These tumours are being prioritised for epigenetic and mutational analysis to help identify the driver events. We are also, in collaboration with Sam Aparicio in Vancouver, sequencing the exomes and transcriptomes of triple‑negative (ER-/PR-/Her2-) breast cancers, which are revealing significant heterogeneity in this important subtype. We continue to contribute to the International Cancer Genome Consortium (ICGC) effort to characterise breast cancer genomes to single nucleotide resolution by deep whole-genome sequencing. Importantly, the availability of 2000 breast cancers with a minimum of five years of clinical follow up constitutes a unique and valuable validation platform for the clinical significance of cancer genes identified by ICGC.

Figure 1
Array-CGH plots of breast cancer with ERBB2 amplification reveal a minimal amplicon containing only ERBB2 and C17orf37. The top-left panel shows a case with a small amplicon containing only ERBB2, C17orf37 and GRB7. The other three panels show cases where the distal breakpoint is located between C17orf37 and GRB7, excluding GRB7 from the minimal amplicon.

To date more than 8,500 paraffin-embedded samples from a population-based cohort and from four randomised clinical trials have been included in tissue microarrays (TMA). We have started to deploy this unique resource of clinically annotated tumours to robustly identify and validate prognostic and predictive markers. We have now shown that BCL2 is an independent indicator of favourable prognosis for all types of early-stage breast cancer and its addition to currently used prognostic models improves survival prediction. We have contributed the largest TMA to the study showing that in women with early breast cancer receiving adjuvant chemotherapy, the most powerful predictor of benefit from anthracyclines is Ch17CEP duplication, and not Her2 amplification as previously proposed. We have been major contributors to the study showing that six subtypes of breast cancer defined by expression of five markers show distinct behaviours with important differences in short term and long term prognosis. This study has definitively shown that triple-negative breast cancers are not a homogeneous entity and this has major implications to the design of clinical trials with agents that target defective homologous recombination.

We have followed our published observations on the importance of immunity in breast cancer outcomes by showing that immune response pathway modules antagonise or synergise to delineate novel prognostic subtypes.

Our efforts in pharmacogenetics, the study of germline genetic variation and its influence in drug disposition, have progressed. Using a large population-based cohort of breast cancer patients treated with adjuvant tamoxifen we have shown absence of association with survival in frequent variants of CYP2D6, including CYP2D6*4, questioning the validity of the reported association between CYP2D6 genotype and treatment response in breast cancer.

Collaborators: Sam Aparicio (University of British Columbia), Simon Tavaré (CRI), Paul Pharoah (Strangeways Research Laboratory), Helena Earl (Department of Oncology and Addenbrooke's Hospital), Gordon Wishart and Elena Provenzano (Addenbrooke's Hospital), Paul Edwards (Hutchison/MRC Research Centre), and the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC)

Functional breast cancer genomics: characterising tumour initiating/cancer stem cells in breast cancer subtypes

We have continued our work to identify and characterise which cells initiate different types of breast cancer and what is the nature of the pathways that are disrupted. We are particularly interested in the role of miRNAs, TGFbeta, a new oncogene on 8p12 (ZNF703), and NOTCH signalling in the regulation of normal and malignant breast stem cells. We are testing whether disruption of these pathways alters the self renewal of tumour initiating cells, which could lead to novel therapeutic approaches to eradicate tumours.

ZNF703 is a new breast cancer oncogene, and amplification occurs almost exclusively in more aggressive ER-positive tumours. We have shown that ZNF703 over-expression in epithelial progenitors in normal human mammary epithelium favours luminal differentiation and attenuates basal differentiation. This might explain why ZNF703 amplification/over‑expression predominates in Luminal B breast cancers.

Collaborators: John Stingl, Jason Carroll and Bruce Ponder (CRI), Eric Miska (Cancer Research UK/Wellcome Trust Gurdon Institute) and Shankar Balasubramanian (Department of Chemistry and CRI)