Targeted treatment for ovarian cancer: A multidisciplinary approach using a replenishable peritoneal implant and natural killer cell immunotherapy

A mouse model inoculated with D-Luciferase expressing human ovarian cancer cell line (D+Luc-OVCAR-8), I demonstrated minimally invasive IP delivery of expanded (e)NK cell immunotherapy (once weekly) and Interleukin (IL)-15 (3 times per week). This regimen significantly reduced tumour burden at Days 35 and 42 compared to standard IP injection. I demonstrated the potential of the implant for longitudinal peritoneal fluid sampling through a transcutaneous port using negative pressure, allowing minimally invasive monitoring of therapeutic response over time. Overall, this chapter presented a promising implant for personalised cancer treatments, enabling repeated, localised therapy delivery and real-time monitoring. Primary NK cells isolated directly from peripheral blood have shown potent anti-tumour activity. However, variability in NK cell function between donors remains a significant limitation in their clinical application. B7-H3, a tumour-associated antigen overexpressed in ovarian cancer represents a promising therapeutic target. Chapter 4 explores the therapeutic potential of eNK cells in combination with B7-H3 Tri-Specific Killer Engager (TriKEs) to treat ovarian cancer in vitro and in vivo. The B7-H3 TriKE comprised of three components; an anti-CD16 that targets NK cell, and anti-B7-H3 that targets cancer cells, which are joined by an IL-15 linker. This TriKE enables NK cell persistence and specificity, further enhancing the therapy. I hypothesised that the addition of B7-H3 TriKE can rescue donors with poor natural cytotoxicity, enabling the NK cells to exert a higher and consistent therapeutic effect. When assessing the in vitro cytotoxic effect, significant donor variability was observed across six donors, with baseline cytotoxicity ranging from 10%-60% at 4 hours. I evaluated cytokine production and found that B7-H3 TriKE significantly enhanced cytokine production across all donors compared to IL-15 or media alone in vitro. However, in vivo studies using a D+Luc-OVCAR-8 ovarian cancer NSG mouse model, revealed that therapeutic efficacy of eNK cells with B7-H3 TriKE was donor-dependant, with donor 11 exhibiting robust tumour control while donor 41 providing no therapeutic benefit. I determined that B7-H3 TriKE could not rescue donors with poor baseline cytotoxicity. This work highlights the need for pre-screening donors to optimise therapeutic outcomes. Leading on from this work, Chapter 5 evaluates the expansion and therapeutic potential of adaptive NK (aNK) cells compared to eNK cells. Cytomegalovirus (CMV), a ubiquitous herpesvirus, is increasingly recognised for its ability to modulate the immune system, particularly NK cells. CMV drives the differentiation of aNK cells, which exhibit enhanced functionality, persistence, and tumour-targeting capabilities. Notably, ovarian cancer patients with higher CMV antibody levels, a marker of robust immune response, exhibit improved survival, underscoring aNK’s therapeutic potential. In this chapter, I demonstrate the effective expansion of NK cells with an adaptive phenotype (Natural Killer Group 2 member C (NKG2C)+, NKG2A-) from two donors, exhibiting variability in yield, phenotype and cytotoxicity. I assessed the in vitro cytotoxicity and identified optimal stimulation to be IL-15, which significantly enhanced cytotoxicity in both aNK and eNK cells compared to baseline or B7-H3 TriKE stimulation, although donor variability in cytolysis persisted. Next, I investigated the in vivo therapeutic efficacy of eNK and aNK cells combined with IL-15. Both significantly reduced tumour burden in the D+Luc-OVCAR-8 human ovarian cancer mouse model by Day 21. Despite significant differences in in vitro cytotoxicity favouring eNK cells, in vivo tumour control was comparable between the two cell types. I analysed post-euthanasia IP fluid and found no significant differences in antigen expression between the groups, except for higher NKG2C expression in aNK cells, a marker of activation. This chapter highlights the therapeutic potential of aNK cells, which exhibit memory-like features and highlights the impact of donor variability on NK cell expansion, phenotype, and cytotoxicity, and again emphasising the need for donor pre-screening to optimise therapeutic outcomes. Due to the variability observed with primary NK cells, I sought to investigate whether an NK cell line could reduce this inconsistency and enhance targeted cytotoxicity through genetic modifications, offering a more uniform and reliable therapeutic approach. In Chapter 6, I investigated the cytotoxic potential of a human NK cell line modified to express a potent death receptor 5 (DR5) specific Tumour Necrosis Factor (TNF) - related apoptosis-inducing ligand (TRAIL) variant. The TRAIL/DR5 pathway, often overexpressed on tumour cells, is a promising therapeutic target. I hypothesised that modifying NK cells to express a DR5-specific TRAIL variant would enhance their cytotoxicity against TRAIL-sensitive (D+Luc-OVCAR-3) and TRAIL-resistant (D+Luc-SKOV-3) human ovarian cancer cell lines in vitro. I compared the cytotoxic effect of both modified and non-modified NK cells. Both significantly reduced D+Luc-OVCAR-3 cell viability as compared to non-treated controls at both 4 hours and 16 hours, demonstrating effective cytotoxicity. Importantly, TRAILv-KHYG-1 cells significantly outperformed non-modified NK cells, indicating enhanced killing capabilities. In contrary, neither modified and non-modified NK cells significantly reduced D+Luc-SKOV-3 cell viability. Through investigating the role of DR5 surface level expression, I found that D+Luc-SKOV-3 cells exhibited higher DR5 surface expression compared to D+Luc-OVCAR-3 cells, indicating DR5 expression alone is insufficient to drive cytotoxicity. These findings are important as they highlight the complexity of NK cell mediated killing in ovarian cancer and underscore the disease’s heterogeneity. This work highlights the importance of combination therapies to enhance NK cell-based immunotherapy efficacy. Finally, Chapter 7 explores the development of an in ovo model as a bridge between in vitro and in vivo. I hypothesised that the chick chorioallantoic membrane (CAM) assay can serve as a reliable and cost-effective 3D model to evaluate the efficacy of therapeutic regimens for treating human ovarian cancer. In this chapter, I demonstrated that cisplatin treatment exhibited a clear dose-dependant cytotoxicity in D+Luc-OVCAR-3 cells. I evaluated inhibitory concentration (IC)20, TRAILv-KHYG-1, and its combination, and found that the combination achieved the highest cytotoxicity, significantly surpassing the effects of either treatment alone in 2D in vitro culture. I optimised cell density and growth conditions for D+Luc-OVCAR-3 cells within CAM model to ensure consistent tumour development and reliable therapeutic response evaluation. However, attempts to establish a robust and reproducible in ovo CAM model faced significant challenges. Despite extensive optimisation efforts, the model was not reliable and could not be utilised to investigate the in vitro therapeutic findings. These findings are important as it highlights the challenges in establishing reliable 3D models. In summary, this thesis describes the development of a novel intraperitoneal delivery system enabling localised, repeatable therapy administration and real-time monitoring. I also evaluate the efficacy of TRAIL modified KHYG-1 cells, B7-H3-targeted TriKEs, and compare the therapeutic potential of adaptive and expanded NK cells. Collectively, these findings provide a foundation for advancing personalised, adaptable immunotherapy strategies for ovarian cancer. The work carried out in this thesis highlights the importance of interdisciplinary approaches that integrate innovative delivery systems, robust preclinical models, and enhanced cellular to work towards an innovative solution for the treatment of ovarian cancer.

Funder

Science Foundation Ireland

Publisher

University of Galway

Rights

CC BY-NC-ND

cbn

Collections

University of Galway Theses (PhD Theses)