Establishing 3D models for the study of intra-tumoural heterogeneity and response to chemotherapy in high grade serous ovarian cancer
File(s)
Author(s)
Ploski, Jennifer
Type
Thesis
Abstract
High grade serous ovarian cancer (HGSOC) accounts for most ovarian cancer cases. Initial platinum sensitivity is followed by relapse and development of resistance to therapeutic strategies. PARP inhibitors targeting the homologous recombination (HR) repair pathway have improved outcomes, however, resistance to PARPi is emerging.
Current methods of predicting treatment response rely on assessing the genomic mutational profile (namely for BRCA mutations) and testing for markers of genomic instability (GIS). However, previous work from the group has shown that the inherent heterogeneity of HGSOC extends to the genomic profile.
Early modelling of HGSOC relied on immortalized cell lines, which are imperfect replicas of the disease due to genetic drift and unclear origins. Cell lines do not replicate the complexity of in vivo tumour biology, which is better captured by models such as three-dimensional patient-derived organoids and ex vivo tumour explants. These models can replicate the genomic profile of the original tumours and preserve the tumour microenvironment.
This study aims to examine whether three-dimensional models derived from multisite sampling of chemo-naïve patients with HGSOC can capture phenotypic heterogeneity. Patient-derived organoids generated from metastatic deposits were examined for the epithelial-mesenchymal state and the cancer stem cell profile, and demonstrated intratumoural and intertumoural heterogeneity. Phenotypic heterogeneity was observed across sites in the response to cisplatin in paired organoid, explant, and monolayer cultures, with three-dimensional models exhibiting greater resistance.
Furthermore, this study aims to develop these models as a platform for drug testing. Digital pathology analysis was optimized for whole-slide imaging of treated ex vivo explants, enabling rapid generation of histology readouts and apoptosis scoring. Spatial heterogeneity was seen in the diverse responses to platinum across metastatic deposits.
Ultimately, patient-derived models can produce therapeutic sensitivity data within clinically relevant and actionable timeframes. These models can guide personalized treatment strategies, thus supporting precision oncology.
Current methods of predicting treatment response rely on assessing the genomic mutational profile (namely for BRCA mutations) and testing for markers of genomic instability (GIS). However, previous work from the group has shown that the inherent heterogeneity of HGSOC extends to the genomic profile.
Early modelling of HGSOC relied on immortalized cell lines, which are imperfect replicas of the disease due to genetic drift and unclear origins. Cell lines do not replicate the complexity of in vivo tumour biology, which is better captured by models such as three-dimensional patient-derived organoids and ex vivo tumour explants. These models can replicate the genomic profile of the original tumours and preserve the tumour microenvironment.
This study aims to examine whether three-dimensional models derived from multisite sampling of chemo-naïve patients with HGSOC can capture phenotypic heterogeneity. Patient-derived organoids generated from metastatic deposits were examined for the epithelial-mesenchymal state and the cancer stem cell profile, and demonstrated intratumoural and intertumoural heterogeneity. Phenotypic heterogeneity was observed across sites in the response to cisplatin in paired organoid, explant, and monolayer cultures, with three-dimensional models exhibiting greater resistance.
Furthermore, this study aims to develop these models as a platform for drug testing. Digital pathology analysis was optimized for whole-slide imaging of treated ex vivo explants, enabling rapid generation of histology readouts and apoptosis scoring. Spatial heterogeneity was seen in the diverse responses to platinum across metastatic deposits.
Ultimately, patient-derived models can produce therapeutic sensitivity data within clinically relevant and actionable timeframes. These models can guide personalized treatment strategies, thus supporting precision oncology.
Version
Open Access
Date Issued
2024-06-20
Date Awarded
01/09/2025
Advisor
Fotopoulou, Christina
Cunnea, Paula
Publisher Department
Department of Surgery & Cancer
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)