TOWARDS PATIENT-SPECIFIC MODELS IN HIGH GRADE SEROUS OVARIAN CANCER (HGSOC): LINKING EPIGENETIC TRACING OF CELL OF ORIGIN WITH ACTIONABLE ORGANOID MODELS

Ovarian Cancer (OC) is a major cause of cancer-related mortality, due to the late-stage diagnosis and failure of surgery and chemotherapy to fully eradicate the disease, that is reflected in a high rate of tumor relapse after treatments. Patients with high-grade serous ovarian cancer (HGSOC), representing the largest majority of OC (~70%), have not experienced significant improvement in overall survival in last decades (1), pointing to the acute need to identify new predictive biomarkers and therapeutic targets for clinical settings. This unresolved emergency derives from our poor understanding of HGSOC biology combined with the recognized lack of suitable clinically relevant models for this disease, as the available models fail to relate molecular aberrations to clinical histories (2). Moreover, one major challenge that sets HGSOC apart from all other solid tumors, that has witnessed significant progress in recent years, is the persistent uncertainty about its cell of origin and the consequent lack of molecular signatures for the unequivocal assignment of specific samples to either an ovarian surface epithelium (OSE) or a fallopian tube fimbrial epithelium (FI) origin, the two candidate tissues still debated in literature (3, 4). Altogether, these aspects have hampered the dissection of pathogenic mechanisms and the identification of targets for improved clinical care of patients, driving us towards the embracement of new paradigms for effective advancements in the clinical setting. To overcome these issues, in the past few years, in the lab, we have devoted our efforts to the identification of the cell of origin of the tumor and the development of a clinical model allowing the study of this disease. Our applications aimed at brought a valid boost towards those precision oncology approaches that promises to transform cancer care by delivering personalized treatments tailored to the genetic and epigenetic specificities of patients’ tumors, whose remarkable heterogeneity characterizes even seemingly homogeneous histotypes. This level of complexity urged the need to develop and validate patient-derived cancer models that recapitulate to a meaningful extent the features of primary or metastatic tumors, both for efficient drug development and for guiding therapeutic strategies. Among new technologies, patient-derived organoid cultures are emerging as powerful models able to recapitulate in a meaningful extent the features of primary tumors (5, 6). Their value was proved both in guiding therapeutic decision for broadly curable tumors (7) and in gaining new insights for intractable tumors such as pancreatic adenocarcinoma (8). OC studies have started to benefit from these advanced models only recently (9–11), and a huge part of this PhD work regards the generation of organoids models from both HGSOC and normal samples of FI and OSE, in order to encompass all kind of tissues useful for the study of the pathology. Indeed, while tumor organoids represent an in vitro proxy of OC at an advanced stage, normal organoids from healthy tissues can be used as control but also as an indefinitely expandable cellular model in which investigate tumor alterations in the context of a normal tissue. This is really significant also in light of our recent publications on the cell of origin of HGSOC (12), where we showed the establishment of a method that is able to stratify HGSOC tumors through an approach based on methylomics and transcriptomics that: i) allows to distinguish OSE- from FI- derived tumors; ii) allows to identify for each tumor subtype the specific transcriptomic alterations that could drive the tumorigenic process; iii) reveals a prognostic value for the tissue of origin of this disease. Taken together, these two sets of information could help in define the molecular mechanisms governing the disease at an unprecedented resolution, setting a paradigm for harnessing patient-specific cancer models in the complementary goal of dissecting pathogenic mechanisms and extracting their actionable features. The extrapolation of information from organoids taking in consideration the cell of origin, through both the increase of our knowledge of the pathology and the identification of new markers and targets to be used as novel drug therapies, will prospectively pave the way to the improve the clinical care of patients with HGSOC.