Browsing by Author "Junquera Mencia, Ada"

High-grade serous ovarian cancer (HGSC) presents a complex clinical challenge as it is often diagnosed at advanced stages. Neoadjuvant Chemotherapy (NACT) is commonly used to treat advanced stage (III and IV) patients who cannot undergo primary debulking surgery. However, 80% of them experience relapse and develop resistance to platinum-based chemo therapies. While NACT alters the tumor immune microenvironment in a treatment response specific manner, its underlying mechanisms remain unclear. Understanding the effect of NACT on tumor microenvironment (TME) is crucial to identify novel biomarkers and develop effective therapeutic strategies. Emerging spatially resolved methodologies, including highly multiplexed-imaging and spatial gene-expression profiling approaches provide novel information regarding spatial interactions and mechanisms operating at single-cell level. However, neither of these methods can offer a comprehensive view of the tumor immune landscape while also revealing the molecular mechanisms behind it. To overcome this limitation, this thesis proposes a novel method integrating tissue cyclic immunofluorescence (tCycIF) and GeoMx DSP spatial transcriptomics. tCycIF is a high-throughput multiplexed imaging method, while GeoMx DSP offers sequencing information at near to single-cell resolution in well-defined regions of interest (ROIs), enabling the dissection of the underlying molecular mechanisms. Here I set out to explore the effect of NACT on tumor stroma interface (TSI) and the surrounding tumor and immune cells, particularly IBA1+ macrophages, and CD8+ T cells. HGSC patient-derived pre- and post-NACT FFPE blocks were acquired from Helsinki University Biobank. To preserve morphological and spatial features, adjacent tissue sections underwent tCycIF and GeoMx experiments. Potential regions for spatial transcriptomics (Pre ROIs) were successfully delineated based on tCycIF staining patterns, enabling the identification of IBA1+ macrophage and CD8+ T cell rich areas within the TSI. By overlaying a crop of the tCycIF image onto the GeoMx scan, areas with distinct similarity degrees between tCycIF and GeoMx were generated. The GeoMx-ROIs were ultimately selected based on the Pre-ROIs and their similarity degree with tCycIF scanned images. Tumor and stroma segmentation was performed through a custom segmentation method guided by tumor specific marker pan-cytokeratin staining in GeoMx images, defining distinct spatial compartments for sequencing. Finally, the image-derived data from both techniques was integrated in a single file, enabling a combined subsequent analysis. The novel pipeline developed in this study opens promising research possibilities, as tCycIF-guided ROI selection allows for precise targeting of cell neighborhoods and structures after performing a thorough microenvironment exploration. This approach can potentially be adapted for other uses in a wide range of biomedical fields, beyond the focus of HGSC.