M.Sc. Qingli Guo defends her doctoral thesis "Developing Clinical Potential of Mutational Signatures: Insights from Formalin-Fixed Biopsies and MSI-High Tumours" on Tuesday the 12th of December 2023 at 13 o'clock in the University of Helsinki Main building, hall Karolina Eskelin (U3032, Fabianinkatu 33, 3rd floor). Her opponent is Professor Benjamin Schuster-Böckler (Ludwig Institute for Cancer Research, University of Oxford, UK) and custos Professor Ville Mustonen (University of Helsinki). The defence will be held in English.
The thesis of Qingli Guo is a part of research done in the Department of Computer Science and in the Bioinformatics and Evolution group at the University of Helsinki. Her supervisors have been Professor Ville Mustonen (University of Helsinki) and Professor Trevor A. Graham (The Institute of Cancer Research, UK).
Developing Clinical Potential of Mutational Signatures: Insights from Formalin-Fixed Biopsies and MSI-High Tumours
The advancement of cancer is driven by the accumulation of somatic mutations arising from various mutational processes within the individual's lifetime. A specific mutational process often generates a unique mutational signature, which holds valuable information about tumourigenesis in specific cancer types and facilitates the prediction of personalised treatment approaches. While numerous studies have focused on identifying novel mutational signatures, a crucial gap remains in translating this knowledge into clinical practice. The proof-of-concept studies incorporated within this thesis aim to bridge this gap.
The first aim of this thesis is to understand the single base substitution (SBS) mutational patterns caused by formalin fixation. Clinical archival biopsies are routinely preserved as formalin-fixed and paraffin-embedded (FFPE) tissues. However, the mutational profile within these samples is often riddled with formalin-induced DNA damage. Our findings indicated resemblances between these patterns and known signatures. Without chemical repair, the FFPE noise pattern closely mirrors SBS30 (caused by Base Excision Repair deficiency). When chemical repair is applied, formalin artefacts exhibit a pattern similar to signature 1 (ageing signature). We developed FFPEsig, a computational method that can accurately predict the biological mutation profiles in FFPE samples. This research provides insights into the mutational processes in archival cancer biopsies and offers a computational method to study mutational processes in such samples.
This thesis investigates also mutational signatures of small insertions and deletions (INDEL) in Microsatellite Instability (MSI) high tumours, which typically demonstrate a favourable prognosis and enhanced response to immunotherapy. However, the current MSI detection methods exhibit limited effectiveness in low-quality tumour samples. Leveraging machine learning techniques, we identified important mutational features, such as longer deletions, that enable sensitive MSI prediction in tumour samples with shallow coverage and limited cellularity. Building upon this finding, we developed MILO to predict Microsatellite Instability status in LOw-quality samples. Subsequently, we presented a case study in which MILO was applied to detect MSI cancer precursors in longitudinal clinical archival specimens. Our approach helps expand immunotherapy utilisation to a broader spectrum of patients.
Finally, we studied the biological process behind the formation of longer deletions during MSI cancer progression. Our results revealed a strong association between the number of longer deletions and advanced stages of MSI tumours. The longer deletions are more pronounced in metastatic MSI cancers compared to primary MSI tumours. The further evidence supports that these longer deletions are accumulated through slippage errors in a stepwise manner. Together, these findings revealed the dynamic genomic consequences associated with the progression of MSI cancers.
In summary, this thesis provides insights that contribute to our understanding of mutational processes in patient archivable materials and MSI tumours. Additionally, we introduced computational frameworks with the potential to aid the translation of mutational signatures into clinical assays in the future.
Availability of the dissertation
An electronic version of the doctoral dissertation will be available on the e-thesis site of the University of Helsinki at http://urn.fi/URN:ISBN:978-951-51-9989-8.
Printed copies will be available on request from Qingli Guo: email@example.com.