Thyroid carcinomas can be divided into different subtypes, with anaplastic thyroid carcinoma (ATC) having the most unfavourable prognosis. The exact tumor evolution of these carcinomas is largely unknown so far, therefore further knowledge is essential for a precise therapy decision. Therefore, the aim of the planned research project is to systematically elaborate clonal evolution via deep sequencing of a rare triphasic thyroid carcinoma collective. In order to identify heterogeneous tumor clones, an extensive thyroid carcinoma collective will be screened in preliminary tests using multigene analysis. Subsequently, suitable triplets are analysed using whole exome sequencing (WES). After subclonal reconstruction, clinical relevance of linear and non-linear tumor evolution will be classified.

One underlying mechanism in the development of cancer is the emergence of genomic instability caused by genetic mutations due to exogenous or endogenous DNA damage or failures in DNA damage repair. Healthy cells maintain genomic integrity by a variety of repair mechanisms, such as mismatch repair or nucleotide and base excision repair, each addressing unique forms of DNA damage. The deficiency of homologous recombination repair (HRD) resulting in DNA double-strand breaks is considered to be the most lethal of all DNA repair defects since cancer cells switch to the error-prone non-homologous end joining (NHEJ) pathway fostering genomic instability and cell death. HRD is therapeutically addressed by poly (ADP-ribose) polymerase (PARP) inhibition. PARP inhibition results in the accumulation of unrepaired DNA single-strand breaks which are converted into double-strand breaks during replication. The combination of homologous recombination repair deficiency and PARP inhibition leads to synthetic lethality of the tumor cells.

The most common reason for HRD are mutations in BRCA1 or BRCA2. Nevertheless, a major part of ovarian cancer patients shows HRD positivity without underlying BRCA mutation. Beyond BRCA mutations, alterations in other HR-related genes, as well as inactivation by promoter methylation can lead to HRD and the induced large genomic defects, so-called genomic scars, are used as biomarkers to stratify patients for PARP inhibitor therapies. These genomic scars comprise loss of heterozygosity (LOH), telomeric allelic imbalance (TAI) and large-scale state transitions (LST).

To avoid central testing without access to patient data, a local testing algorithm for patient stratification is needed. Testing for BRCA mutations by parallel sequencing has been integrated in clinical health care since several years. For genomic instability, different methodologies are in use, such as SNP arrays or different parallel sequencing approaches. To compare these different approaches with the central testing method (Myriad MyChoice CDx), a German Multi Center Technical Comparison Study in collaboration with the Institute of Pathology of the University Hospital Cologne was initiated. The results showed that for clinical care, targeted panels including a SNP backbone might be the best solution. They allow for simultaneous detection of BRCA mutations and genomic scars.

The primary aim of this project will be the analysis and definition of mechanisms involved in homologous recombination repair causing HRD positivity. Therefore, a next generation sequencing (NGS) panel analyzing the mutation status of 34 HRR associated genes among others will be implemented, validated and automated. The molecular mechanism and molecular effects (cell viability/ sensitivity to different therapy options) will be analyzed in vitro. For this purpose, identified mutations in HR-related genes will be mimicked in a cell line by a CRISPR/Cas9 induced knockout. To further investigate the correlation to the response to PARPi, expression profiling of HR relevant genes will be carried out on RNA level. Furthermore the methylation status of HR relevant genes will be assessed.

CGI Clinics is an EU-funded project that aims to develop a database for the interpretation of gene mutations in a biological and clinical context (Cancer Genome Interpreter - CGI). The project was initiated by Dr. Nuria Lopez-Bigas (Institute for Research in Biomedicine, Barcelona) and her team. 17 clinics and institutes from all over Europe have joined the project. The molecular pathology departments of the University Hospitals of Cologne and Aachen, members of the "CIO Aachen Bonn Cologne Düsseldorf", are the only project partners from Germany

The biggest problem with the interpretation of mutation profiles is currently that the existing platforms and databases are not based on a standardized system. One consequence of this is that it takes a lot of time to interpret the variants. Another challenge is that there are many mutation variants with unknown significance, i.e. the exact influence of the variants on the organism is not known.

CGI-Clinics aims to systematize the interpretation of gene mutations. It has set itself the goal of developing and optimizing the Cancer Genome Interpreter as a powerful "one-stop store" tool and establishing it in everyday clinical practice. The tool is intended to support doctors in their decision-making when selecting the right therapy. The aim is also to continuously improve personalized therapies. In its current version, the algorithms of the Cancer Genome Interpreter were developed based on the analysis of the genomes of 28,000 patients, covering more than 66 types of cancer and available to the scientific community in public sources. As new sequenced tumors are added over the course of the project, machine learning methods will improve predictions and thus also the interpretation of tumor mutations for new patients.

The data and results from the project will also be made available to patients. Patients are thus given the opportunity to better understand their disease - and also develop an understanding of the importance of making personal data available for research purposes. www.cgiclinics.eu