Identification of driver mutations and tumour evolution in HER2 positive breast cancer

This thesis describes the results of a variant calling project using data from the TCHL (Trastuzumab, Carboplatin, and Docetaxel, with Lapatinib) breast cancer whole exome sequencing project. Using a data processing pipeline based on the GATK (Genome Analysis Tool Kit) Best Practices Pipeline, variants were called on a total of 34 tumour samples from 25 patients. This included 20 patients where only a pre-treatment sample was taken from that patient, and 5 patients where samples were taken from multiple timepoints across the course of therapy. The whole exome sequencing data were analysed for the presence of known or predicted driver mutations. The subclonal architecture and mutational signatures of the genomes of these samples were also analysed.

The cohort showed mutational signature patterns typical of a breast cancer cohort. The genomic landscape of the samples taken from the same patient across multiple timepoints was used to analyse the evolutionary history of the tumours in those patients and how they had evolved in response to therapy. The genomic landscape of samples from patients who showed complete response to therapy is compared to the genomic landscape of patients who did show complete response to therapy.

The results of this study highlight the following subjects as promising areas for future study with larger scale cohorts:

-The impact of the subclonal complexity of a tumour on the probability it will show complete response to therapy

-The impact of the presence of SNVs and indels in RAD50, ARID1B, DHX9, IZF3, TNPO2, UBR5 and TAOK1 on the probability that a tumour will show complete response to therapy