Academic research community (GECIP)

Akin to the caps on the end of shoelaces telomeres are structures that cap the ends of human chromosomes and protect them from damage. When telomeres stop working the chromosomes can become damaged and this helps tumours evolve to become more aggressive and less responsive to treatment. This project aims to use genomic data from breast and blood cancers to understand the mechanism by which dysfunctional telomeres can lead to genetic changes across the cancer genome. Ultimately this may lead to the development of clinical tools that will aid prognosis and prediction of response to treatment

Patients with suspect solid tumours undergo biopsies to retrieve part of the tissue for analysis. Our growing understanding of the variants underpinning cancer has led to a more personalised approach to treatment planning and the use of therapies targeted to a patient's specific variants. As the patient's DNA is extracted from the biopsy for sequencing, it is important that this tissue is preserved in the highest quality. In the clinic, most biopsies are stored as formalin-fixed paraffin-embedded (FFPE) samples, which are cheaper, easier to use and more accessible. However, this process damages the DNA, making it much more difficult to sequence and interpret. Freezing the biopsy and storing it as a flash-frozen sample preserves the DNA without damage but this is much more expensive and impractical for clinics in the long run. In the 100,000 Genomes Project (UK100kGP), whole genome sequencing (WGS) is performed on flash-frozen tumour samples from NHS patients to study cancer driver variants and mutational signatures. In this project, we will explore whether the same information can be found in the UK100kGP FFPE samples. We will aim to answer the question: if we only have access to FFPE samples, what information can we extract reliably? We will also investigate how to distinguish between true variants and ones introduced by FFPE-related DNA damage. Through this project, we will learn how best to use FFPE samples for reliable mutational analysis with the aim of making targeted therapies more accessible for all NHS patients.

The majority of the genome is non-coding and therefore variants in these regions do not affect protein structure or function directly. However, there is a large fraction of the non-coding genome that is involved in the regulation of gene activity. We are studying genetic variation present in the regulatory elements of breast cancer risk genes in the NGRL and pwill use in depth computational analysis to understand their effect on breast cancer development.

Pathogenic variants in breast cancer associated genes confer a high risk for the patient, and there are available prophylactic treatment options. However, patients, who have variants of uncertain significance (VUS) in these genes, are not offered these treatments, as there is no evidence of their increased risk. Patients carrying VUS are offered treatment and prevention based on their personal and family histories until the VUS are reclassified. We aim to perform computational analyses and use prevalence and phenotype association data from the NGRL to reclassify VUS variants, found in 9 breast cancer risk genes in the Qatari hereditary cancer screening program, and validate our results in an in vitro system.

This research project is approved, but is not approved for publication. Homologous Recombination (HR) deficiency is a predictor of a poor outcome from cancer treatment and identifying patients with it is an increasingly important treatment prioritisation strategy. Several FDA approved methods have been implemented in PARP inhibition clinical trials (e.g. the Myriad Genetics BRACAnalysis CDx (BRCA1/2) assay, and the Myriad myChoice HRD FDA-approved tumour test ). Recent clinical trials of breast and ovarian cancer have shown that patients with HR-deficiency are sensitive to platinum-based chemotherapy or PARP inhibition. Clearly, better HR-deficiency biomarkers will not only benefit breast and ovarian cancer patients with approved PARPi, but also benefit patients with other cancer types by accelerating PARPi clinical development in other cancer types. Furthermore, recent genomic analyses have revealed several signatures indicative of defective HR across several mutation types (e.g. "Signature 3" of single base substitutions (SBS), micro-homology (MH) mediated deletion etc.). However, individual signatures do not capture complex aspects of the “genomic instability" intrinsic to HR-deficiency. To overcome this limitation combinatorial approaches have been evaluated. The advantage of combining multiple HRD signatures into a single classification model was clearly demonstrated in the WGS-based "HRDetect" were six genomic signatures were integrated into a weighted model to identify BRCA1/2 deficiency. We have recently developed WES-HRD, a WES-based HRD classifier and validated its performance in the independent breast cancer samples with a down-sampling; achieving 92% sensitivity at 98% specificity. We plan to develop a WGS-based HRD classifier using similar method using Genomic England WGS data. We have also developed HE-HRD, an H&E image-based HRD classifier using deep learning, in which the signature-based HR-deficiency labels were served as a training and a testing set. We aim to evaluate the performance of our HRD models in the Genomic England breast cancer cohort. We also plan to develop WGS-based HRD classifier using similar method as used for WES-HRD. If successful we will develop H&E image-based HRD deep learning models for ovarian, pancreas, prostate, endometrial, lung, and other solid tumour types.

There are differences in variants causing breast cancer in different populations. In many individuals of Pakistani and Colombian origin with strong family history of breast cancer no known cancer mutation can be identified, suggesting that other currently unknown variants cause the familial clustering. We have identified rare, possibly pathogenic variants in patients from Pakistan and Colombia. The aim of this project is to assess their effect in individuals from the UK.

Breast Tumors are characterized by several histopathological subtypes which respond differentially to chemotherapy, radiation, and investigational therapies. Through the investigation of mutational status associated with clinical phenotypic data driven by artificial intelligence analysis, clinically relevant signatures can be identified of therapeutic importance. This can identify hotspot variants, which impact key biological pathways effectuating therapeutic response.

Genetic screening has become a routine part of medical care for breast cancer patients. The information can be used for cancer prevention in relatives and to guide therapy choices in patients. However, there are still many genetic variants identified via screening for which the impact on the risk of breast cancer is unknown, leaving patients and families with a lot of uncertainty. Our goal is to develop a novel strategy that uses the genetic information contained in the patient’s tumour to determine if a genetic variant is disease-causing or not. This would bypass the need to study each variant using lengthy molecular laboratory tests. We will analyse the genomes of multiple tumours from breast cancer patients who carry variants of unknown significance in several breast cancer predisposition genes. We will study the same variants in a traditional molecular laboratory setting using validated functional assays and compare the results. If successfully implemented, our new classification strategy will resolve the uncertainty for many patients whose genetic tests reported a variant of unknown significance and will improve the efficacy of genetic testing for future breast cancer patients.

Breast cancer patient's mortality is mainly due to its spread to other parts of the body than primary tumour itself (metastasis). By understanding and preventing this process, breast cancer could be cured. Breast cancer spreads by invading into surrounding tissues and finding their way into the lymphatic and vascular system (lymphovascular invasion). Cancer cells can then travel to other parts of the body (metastasis). Identification of the key mechanisms involved in lymphovascular invasion has so far remained mysterious in beast cancer.

Breast cancer is the most frequently diagnosed cancer amongst women in the UK. Improvements in genetic sequencing technology and a reduction in cost are facilitating access to cancer genetic testing in mainstream breast cancer care. This technology has the potential to benefit patients by identifying individuals that have an increased risk of developing breast cancer. It may also cause harm through the identification of genetic variants where the association with breast cancer risk is less clear. These are called Variants of Uncertain Clinical Significance (VUS). This project aims to determine whether the variants that are present in a breast cancer and also the appearance of that cancer under a microscope can be used to better interpret these VUS.

Histopathologists check that the seemingly normal tissues around the edges of the surgically-removed tumours are healthy. However, early cancer changes do not always affect the appearance of breast tissue, so histopathology can miss early markers of cancer risk. We propose to expand upon previous work in which we examined tumour tissue and normal-looking tissues in the affected breast and identify signatures in normal-looking tissues able to predict outcome and guide follow-up screening to transform the care of breast cancer.

Lung and breast cancer are two of the most common cancer types world-wide with a very high death rate. Previous studies have already uncovered the complex genomic landscape of this disease, but little is known about the epigenetic landscape and how those two interact with each other. Analysis of differences in the mutation and copy number distribution relative to replication timing changes between normal and cancer could reveal new gene targets and new biological processes that are active during tumour development. It could also allow us to investigate replication timing as a potential biomarker.

Cancer is an evolutionary disease that changes and adapts over time. Cancer cells that are more aggressive/resistant to treatment that develop within tumours can survive and reproduce more easily than the original cancer. High levels of genetic diversity in tumours drives this evolution and is linked to poor patient prognosis and the development of drug resistance, but this diversity also leads to unique vulnerabilities in cancer cells that we can target with drugs.

In this project we will use the tumour sequencing data in the Genomics England dataset to understand what kinds of DNA sequences evolve repeatedly in different tumours from different patients and consistently drive drug resistance. Once we have identified these sequences we can then look in databases of known cancer vulnerabilities to find ways to target them and develop new drugs that prevent drug resistance developing.

The NHS Long-Term Plan for cancer aims to have 55,000 more people each year survive their cancer for five years or more, so the development of therapies that specifically tackle cancer evolution and drug resistance are urgently required.