Abstract
Endometrioid endometrial cancer (EEC), is rising in prevalence globally and there are limited treatment options aside from surgical removal of the uterus. This thesis focuses on early-stage EEC due to its high prevalence and ability to be treated non-surgically. The levonorgestrel-loaded intra-uterine system (LNG-IUS) is a common non-surgical treatment option but has highly variable success rates (43-81%) with no established biomarkers for predicting response. Research into alternative non-surgical treatments for EEC, such as cannabidiol (CBD), is limited but promising, showing potential in restricting EEC cell function in preliminary in vitro investigations, and should be investigated as a non-surgical treatment option alongside LNG.
The lack of biologically relevant research models for EEC has impeded scientific understanding of non-surgical treatments and their impact on the EEC tumour microenvironment. Patient-derived explant (PDE) models are biologically relevant models generated from patient tissue which retain native tissue architecture and cell-to-cell interactions. PDE models therefore provide more robust preclinical data compared to traditional cell monolayer culture, particularly for the investigation of novel therapeutics such as LNG and CBD. This thesis aimed to develop and utilise a PDE model of EEC to investigate the effect of LNG and CBD treatment on the tumour and immune compartments of the EEC microenvironment.
The development of EEC PDEs was carried out in Chapter 2. Tumour samples of early-stage EEC were cultured on gelatin sponges for 21 days. A long-term culture allowed for a more robust assessment of changes due to LNG treatment, which is generally used for 6-12 months. Histological assessment of explant viability, proliferation, and apoptosis demonstrated that explants retained their native architecture. Following validation, explants were treated with LNG and showed varying viability at 21 days. Tumours deemed resistant to LNG were assessed for their expression of five candidate biomarkers of LNG resistance, identified by previous work in cell lines. Three of these markers were present in all explants and warrant further investigation. Because of the intra- and inter-individual variability in explant viability, this chapter concluded that the PDE model may be best suited to assessing the effects of treatment at an earlier endpoint, which would also allow investigation of components of the tumour microenvironment which may not survive 21 days ex vivo, such as tumour-associated immune cells.
The PDE methodology was subsequently used to investigate the effect of treatment on the EEC immune compartment in Chapter 3. Firstly, T cells were identified as the most abundant immune cell population within uncultured EEC tissue by spectral flow cytometry. Multiplexed immunofluorescence was also used on uncultured EEC tissue to assess T cell infiltration into tumoural and stromal compartments in LNG-IUS naïve and LNG-IUS resistant tumours. There was an influx of CD8+ (cytotoxic) and regulatory T cells negative for the exhaustion marker, TIGIT in the stromal compartment of LNG-IUS resistant EEC tumours. A short-term (48 h) PDE model and a 13-colour spectral flow cytometry panel were used to probe the effect of LNG and CBD on EEC-associated T cells and results were compared to PBMC-derived T cells from healthy volunteers cultured in a monolayer. CBD demonstrated a highly cytotoxic role on PBMC-derived T cells in monolayer culture and suppressed the expression of key identification and activation markers by these cells. Comparatively, explant-derived T cell viability was unaffected by CBD, and its effect on the expression of activation markers was tumour-dependent. LNG was mildly immunosuppressive in PBMC-derived T cells but showed variability in its action on T cells derived from explants. These results demonstrated an anti-tumour efficacy of CBD in EEC explants with a limited reduction in T cell function, albeit with individual variability, which prompted further investigation into CBD as a novel treatment for EEC.
To investigate the role of CBD in EEC, cell line models including monolayer and spheroid cultures were utilised in Chapter 4. A non-lethal dose of CBD reduced proliferation, migration, and invasion of EEC but not non-cancerous endometrial epithelial cells. This suggests that CBD may have a greater effect on EEC cells. To investigate this further, the expression of known CBD receptors were compared between EEC and non-cancerous endometrial epithelial cells using qPCR and western blot. There was a downregulation of cannabinoid receptors in EEC at the transcript level, except for GPR55 which was upregulated in one EEC cell line. At the protein level, expression was similar across cell lines, although the same EEC cell line also expressed a high level of GPR55.
Together, findings from this thesis demonstrate the generation and utility of an ex vivo PDE model to investigate non-surgical treatments for EEC within different cellular compartments of tumour tissue. The model could be further utilised to study the mechanism of LNG response and its effect on the tumour microenvironment. Furthermore, these findings, coupled with a deeper investigation of CBD in cell lines, warrant the further investigation of CBD as a novel non-surgical treatment for EEC.