| dc.contributor.advisor | Beißbarth, Tim Prof. Dr. | |
| dc.contributor.author | Perera-Bel, Julia | |
| dc.date.accessioned | 2018-11-26T09:19:11Z | |
| dc.date.available | 2018-11-26T09:19:11Z | |
| dc.date.issued | 2018-11-26 | |
| dc.identifier.uri | http://hdl.handle.net/11858/00-1735-0000-002E-E511-6 | |
| dc.identifier.uri | http://dx.doi.org/10.53846/goediss-7161 | |
| dc.language.iso | eng | de |
| dc.rights.uri | http://creativecommons.org/licenses/by-nc-nd/4.0/ | |
| dc.subject.ddc | 610 | de |
| dc.title | Guiding Cancer Therapy: Evidence-driven Reporting of Genomic Data | de |
| dc.type | doctoralThesis | de |
| dc.contributor.referee | Beißbarth, Tim Prof. Dr. | |
| dc.date.examination | 2018-11-19 | |
| dc.description.abstracteng | Next Generation Sequencing (NGS) has been crucial for the breakthrough experienced by cancer genomics during the last decade. In turn, the knowledge gathered has fostered the development of targeted drugs and genomics-driven cancer treatment. Some university hospitals have built the infrastructure and invested in human resources for the implementation of NGS within precision medicine initiatives. However, the expertise required to integrate the data with available knowledge spans several disciplines; the information to decipher the clinical implications codified in the genome of a tumor is scattered across many resources; and the complexity of the data demands of computational support.
In the anticipation of a widespread use of clinical sequencing, this thesis describes an evidence-based workflow, the Molecular Tumor Board (MTB) Report, aimed at paving the way for genomics-driven oncology. Deciding whether or not a molecular alteration entails clinical action (i.e. if the variant is actionable) involves a wide-range of expertise and the need to keep up with the pace of new discoveries (e.g. clinical trials, conferences, preclinical studies). The workflow presented here uses public resources to narrow somatic variants from a tumor’s genomic profile down to actionable variants. Furthermore, actionable variants are classified into a six-level system based on the evidence that supports the actionability. The variables considered are cancer type in which the evidence exists and grade of predictive association between a variant and a drug. The classified variants and the evidence that supports their actionability are detailed in a concise report to support clinical discussions. To increase the usability and availability of the workflow, it has been implemented as a web-based application. The user can provide custom data as well as explore a public dataset. Actionable variants can be visualized in an interactive setting and downloaded in the aforementioned report format or in a tabular data file.
The MTB Report workflow was tested over two different large public datasets to evaluate its scope and strengths, The Cancer Genome Atlas (TCGA) and Genomics Evidence Neoplasia Information Exchange (GENIE). The results concerning variants currently used to guide treatment were in line with published numbers of patients receiving genomics-driven therapies. The results also suggested that these numbers could be increased to a large extent if low-evidence (clinical and preclinical evidence) and predictive associations that have not been established for the cancer type in the patient being tested (i.e. off-label) were considered.
A retrospective comparison study was performed for a subset of patients from the Molecularly Aided Stratification for Tumor Eradication Research (MASTER) precision medicine program. The variants identified by the MTB Report were compared to the variants suggested by the experts of the MASTER program. The results showed high concordance between both approaches, as the majority of expert suggestions were identified by the workflow. The workflow identified a plethora of other variants, that, though not yet actionable, depicted a comprehensive landscape of the actionability of the patient.
In all, this thesis work established a computational workflow aimed at enabling a widespread use of NGS for guiding clinical decisions. We envision that such efforts on standardizing genomic data interpretation and reporting will become useful resources in the field of precision medicine. | de |
| dc.contributor.coReferee | Sax, Ulrich Prof. Dr. | |
| dc.contributor.thirdReferee | Brors, Benedikt Prof. Dr. | |
| dc.subject.eng | Precision medicine | de |
| dc.subject.eng | Targeted therapies | de |
| dc.subject.eng | Genomic Report | de |
| dc.subject.eng | Predictive biomarkers | de |
| dc.subject.eng | Molecular tumor board | de |
| dc.subject.eng | Actionability | de |
| dc.subject.eng | Somatic variants | de |
| dc.subject.eng | Cancer genomics | de |
| dc.subject.eng | Bioinformatics | de |
| dc.subject.eng | Next generation sequencing | de |
| dc.identifier.urn | urn:nbn:de:gbv:7-11858/00-1735-0000-002E-E511-6-5 | |
| dc.affiliation.institute | Göttinger Graduiertenschule für Neurowissenschaften, Biophysik und molekulare Biowissenschaften (GGNB) | de |
| dc.subject.gokfull | Biologie (PPN619462639) | de |
| dc.identifier.ppn | 1041060394 | |