Raman Microspectroscopy and Fluorescence Lifetime Imaging Microscopy based Data-Driven Tissue Discrimination and Diagnostics

Abstract

The ultimate objective of any new interventional or surgical techniques is to achieve a balance of minimal invasiveness, optimal efficacy, and rapid treatment duration, all while minimizing the risk of complications. A pivotal component in the management of any disease is the distinction between the impacted target structures and the neighboring healthy tissue throughout all medical interventions, encompassing surgical procedures. Such differentiation is paramount in reducing harm to healthy tissue and augmenting the efficacy of the treatment. Within the current therapeutic procedures, innovations in the field of preoperative and postoperative diagnostics contribute to the improved differentiation of benign and malignant tissue structures. On the one hand, sophisticated imaging techniques support an improved surgical decision-making, while on the other hand, histopathological examination methods enable a precise classification of the tissue after surgery. In contrast, intraoperative tissue differentiation has been based on the time-consuming gold standard of frozen section diagnostics for many years. However, by incorporating the supplementary data provided by advanced imaging sensors during surgery, and integrating it with cutting-edge machine learning methodologies, it is feasible to augment the quality of information utilized for tissue differentiation, thereby increasing the precision of the overall process. Another important task of modern medicine is the patient-specific treatment of cancer, as it has been found that different patients do not respond in the same way to drug treatment due to developed resistance mechanisms. This thesis aimed to establish Raman microspectroscopy (RMS) as marker-independent, and non-destructive technique to monitor fibrotic and epigenetic modifications in malign and benign human tissue. Additionally, the potential of RMS and fluorescence lifetime imaging microscopy (FLIM) was evaluated to non-invasively monitor the drug efficacy on patient-derived organoids from cancer patients. Towards this aim, collagen type I (COL I) structures of formalin-fixed paraffin-embedded (FFPE) tissue sections of various fibrotic diseases were compared to respective control tissue sections. Incorporating Raman measurements into the experimental protocol, we conducted routine histological and immunofluorescence (IF) staining techniques to showcase the superior efficacy of RMS when paired with spectral deconvolution. This technique offers a time- and cost-efficient alternative to conventional procedures. For the differentiation of pathological tissue, the identification of epigenetic modes of action is a promising and potentially successful approach. In alignment with IF imaging of the most abundant epigenetic modification of 5-methylcytosine (5mC), Raman spectra of cell nuclei were evaluated using multivariate data analysis. Compared to invasive staining methods, non-invasive RMS showed promising results for the differentiation of pathological tissue changes in cardiac fibrosis and endometriosis. In addition, RMS and FLIM have been used on several bladder and colon cancer organoids to evaluate their potential to monitor patient-specific responses to drug treatment. The results showed both that organoids are generally suitable as a screening platform for drug treatments and that Raman and FLIM have the potential to assess drug sensitivity. This work highlights the potential of RMS for future applications in ex vivo tissue discrimination of fibrotic diseases and identification of epigenetic changes. In addition, this work demonstrates the proof of principle that RMS and FLIM are suitable for monitoring patient-specific responses to medications on organoid models.