Less than half of patients with epithelial ovarian cancer (EOC) survive five years following diagnosis, underscoring the imperative need for improved treatment. Many patients, including those with advanced disease, initially respond to platinum agents, which constitute the backbone of therapy. However, tumors ultimately become resistant, rendering further treatment ineffective. Additionally, the poor tolerability of these agents warrants the exploration of more targeted treatments – one such strategy is exploiting synthetic lethal genetic relationships. Recent genomic sequencing efforts have revealed that as many of half of EOCs have homologous recombination (HR) alterations. HR is a critical pathway for the repair of platinum-induced ICLs, thus compromised HR is hypothesized to explain the initial response to chemotherapy in many patients. Accordingly, women whose tumors harbor mutations in the critical HR genes, BRCA1 or BRCA2 (BRCA1/2), demonstrate improved prognosis. BRCA1/2 mutations also confer exquisite sensitivity to inhibitors of the enzyme, poly(ADP-ribose) polymerase 1 (PARPis), hence loss of BRCA1/2 and PARP1 is synthetic lethal. A number of models have been proposed to explain this synthetic lethality, yet a consensus model that accounts for the diverse cellular roles of BRCA1/2 and PARP1 has yet to be established. Delineating the precise molecular underpinnings of PARPi action in BRCA1/2-deficient cells will aid clinicians in identifying the appropriate population of women with EOC likely to benefit from PARPi treatment and provide insight into resistance mechanisms that arise in these patients. Combining this approach with retrospective analysis of PARPi clinical trials will best define the proper indication for PARPi in EOC and other human cancers.

Major depressive disorder (MDD) is a prevalent and enervating mental illness affecting millions globally. Unfortunately, a significant proportion of patients do not receive clinical benefit from existing antidepressant medications. The limited effectiveness of currently available antidepressant drugs emphasizes the need to identify more effective medications for individuals who are treatment-resistant. We have previously reported abnormally elevated poly (ADP-ribose) polymerase-1 (PARP1) gene expression levels in the postmortem brain from MDD brain donors. PARP1 is a DNA damage repair enzyme that is also linked to neuroinflammation through multiple biochemical pathways. PARP1 upregulation in MDD could indicate a role for this enzyme in the etiopathology of MDD, particularly as it relates to neuroinflammation. In fact, we have shown that drugs that inhibit PARP1 produce antidepressant-like properties in two different rodent behavioral models that mimic depressed mood in humans. In the present study, we utilized a unique rodent behavioral model that produces depressive-like behavior by combining psychological stress with stimulation of inflammation. Depressive behavior produced by this experimental paradigm is not reversed by the prototypical antidepressant fluoxetine. This treatment-resistant depression was elicited by treating rats with injections of lipopolysaccharide (LPS; 0.1 ug/kg/day) and daily exposure to chronic unpredictable stress (CUS) for 28 days. Depressive behaviors were measured with sucrose preference and forced swim tests in 5 treatment groups (n=6-8 rats per group) including unstressed rats, CUS rats, CUS+LPS rats, and CUS+LPS rats treated with either the PARP1 inhibitor 3-aminobenzamide (3AB) or the antidepressant fluoxetine. We evaluated the role of neuroinflammation in this model by measuring the amount of microglial activation in several brain regions in rats from all treatment groups. Microglia activation was measured by quantifying the relative amount of expression of the microglia marker protein, IBA1, using an anti-IBA1 antibody. 3AB demonstrated robust and unique antidepressant activity superior to fluoxetine in the treatment-resistant rat model. IBA1-immunoreactivity levels were elevated in brains from CUS and CUS+LPS rats, although there was no evidence that LPS increased IBA1-immunoreactivity above levels found in CUS rats that did not receive LPS. Levels of IBA1-immunoreactivity in the brains from rats treated with either fluoxetine or 3AB trended lower as compared to the CUS and CUS+LPS groups, although this effect did not reach statistical significance. The lack of significant differences is likely related to small sample sizes; experiments are underway to increase the sample sizes of each group. The findings provide further support for the potential of PARP1 inhibitors in treating MDD and suggest that these drugs may be more effective, or more broadly effective than standard antidepressants.

Homologous Recombination (HR) is a major double-stranded break (DSB) DNA repair pathway. It utilizes the sister chromatid as a template, thus promoting error-free repair. The initial HR step is DNA end resection with the creation of RPA coated single-stranded DNA (ssDNA) overhangs. This is followed by RAD51 loading onto the resected DNA, triggering strand invasion, and homology search. The BRCA1 protein is fundamental for HR mediated DNA repair, and patients with germline BRCA1 mutations have a high risk of developing cancer. These BRCA1 mutant tumors are HR deficient, and as a result, sensitive to PARP inhibitors (PARPi). BRCA1 likely promotes DNA end resection by displacing 53BP1 from the DSBs and also promotes RAD51 loading by enabling the formation of the larger BRCA1-PALB2-BRCA2 complex trough its coiled-coil region. Although BRCA1 functions in both HR steps, its essential role in each step is unclear, with previous reports showing that RAD51 loading after DNA end resection occurred could be BRCA1 independent. Here, we demonstrated that full HR restoration after activation of end resection mediated by 53BP1 knockout (KO), requires the expression of a BRCA1 protein that retained the coiled-coil region and its protein interactions. We observed that although 53bp1 KO can restore the Mendelian frequencies of a Brca1 null mouse model, mice were still tumor prone, PARPi sensitive, and had low levels of RAD51 γ-irradiation-induced foci (IRIF), suggesting that HR levels were minimal. Moreover, in order to significantly increase PARPi resistance in human cancer cell lines, 53BP1 loss of function (LoF) needs to be associated with the expression of a hypomorphic BRCA1 protein that retained the PALB2 interaction. We also demonstrated that Brca1 compound heterozygosity could rescue the developmental defects observed in Brca1 mutant homozygous mice, as long as the alleles retained complementary HR-related functions. Additionally, these compound heterozygotes mice were not tumor prone and had a normal lifespan, suggesting that HR was restored. In contrast, when the allele combination did not exhibit complementary functions, the developmental defects persisted. Overall, we demonstrated that to fully function in mouse development, tumor suppression, and drug resistance, BRCA1 role is critical in both the DNA end resection and the RAD51 loading HR steps. / Biomedical Sciences

Le carcinome hépatocellulaire est l'un des cancers les plus fréquents et des plus sévères à travers le monde. Le diagnostic est souvent tardif et les traitements curatifs ne peuvent être proposés qu'à un nombre limité de patients. Les technologies modernes ont permis le développement de nouvelles méthodes de radiothérapie qui montrent aujourd'hui de bons résultats. Par ailleurs, bien que des déficiences dans les voies de réparation de l'ADN soient associées à une instabilité génomique et une susceptibilité au cancer, une inhibition de ces voies sensibilise les cellules cancéreuses à la chimiothérapie et à la radiothérapie. Dans ce contexte, les inhibiteurs de poly(ADP-ribose) polymérases (PARP) ont déjà montré des résultats prometteurs dans des études pré-cliniques et sont en cours d'évaluation clinique pour de nombreux cancers. Ce travail de thèse a consisté en l'évaluation du potentiel des inhibiteurs de PARP en combinaison avec la radiothérapie comme nouvelle option thérapeutique pour le carcinome hépatocellulaire. La première étape de ce travail a été de caractériser les profils d'expression et d'activité de plusieurs membres de la famille PARP dans des cellules cancéreuses du foie et des hépatocytes primaires humains ainsi que dans des tissus hépatiques. En second lieu, nous avons étudié le potentiel de l'inhibiteur de PARP ABT-888 seul et en combinaison à des radiations ionisantes in vitro. Le traitement par l'inhibiteur de PARP ABT-888 en agent seul a montré une sensibilité variable des différentes lignées cellulaires étudiées à cette drogue. Afin de comprendre la sensibilité variable des cellules cancéreuses hépatiques à l'ABT-888, nous avons analysé leur capacité de réparation des dommages à l'ADN et avons observé des capacités différentes entre les lignées cellulaires. Finalement, nous avons pu montrer que l'ABT-888 sensibilise les cellules cancéreuses hépatiques aux radiations ionisantes. Ce travail de recherche a permis de montrer que les inhibiteurs de PARP ont un fort potentiel pour améliorer les méthodes de radiothérapie utilisées dans la prise en charge du carcinome hépatocellulaire / Hepatocellular carcinoma is the third cause of cancer related death. Due its often late diagnosis and advanced stage, a limited number of patients can benefit from curative treatments. There is thus a constant need for new treatment strategies for patients with hepatocellular carcinoma. Targeting DNA repair pathways to sensitize tumor cells to chemoor radiotherapeutic treatments is now a common strategy under investigation for cancer treatment with inhibitors of poly(ADP-ribose) polymerases (PARP) showing great potential. The aim of this work was to evaluate the potential of PARP inhibitors alone and in combination with radiation therapy as a new strategy for the treatment of hepatocellular carcinoma. We first analyzed the expression and activity of different PARP genes in a panel of liver cancer cell lines and primary human hepatocytes as well as their DNA repair capacity and assess the impact of PARP inhibitors alone and in combination with ionizing radiation in these models on cell survival. A large range in expression of PARP family members, PARP activity and sensitivity to ABT-888 in the panel of liver cells was observed as well as differential excision/synthesis repair capacity. Finally, we showed that ABT-888 sensitizes liver cancer cells to the cell killing effects of ionizing radiation. PARP inhibitors show great potential for improving radiation therapy strategies used in the management of hepatocellular carcinoma

Over 16 million people in the US suffer from major depressive disorder (MDD) each year. Approximately 1/3rd of MDD patients (~5 million) obtain only partial remission or no benefit after trials with multiple drugs or drug combinations. Recently, Ordway and colleagues have reportedelevated levels of DNA oxidation and upregulated gene expression of the base excision repair enzyme poly(ADP-ribose) polymerase-1 (PARP1) in postmortem brain from donors who had MDD at the time of death, as compared to age-matched psychiatrically normal control donors. This study was designed to test whether an inhibitor of PARP, 3-aminobenzamide (3-AB), may be effective to alleviate depressive-like behaviors in a rodent model of treatment-resistant depression. Male rats were ip administered lipopolysaccharide (LPS;100ug/kg) daily for 28 days, and administered a chronic unpredictable stressor on each day. All rats were also administered saline, 3-AB (40 mg/kg), or the serotonin-reuptake inhibitor (SSRI) fluoxetine (trade name: Prozac; 10 mg/kg) on each day, approximately 30 min after LPS treatment. During the 28 day period of LPS treatment, animals were behaviorally tested 5 times on sucrose preference (a test of anhedonia). At the end of the 28 day period, rats were behaviorally tested on a test of acute stress, the Porsolt swim test. Results revealed that 3-AB alleviated anhedonia and the response to acute stress in the Porsolt swim test superior to the fluoxetine group, demonstrating the utility of a PARP inhibitor to alleviate depressive-like behavior in this model. In addition, fluoxetine produced a loss of weight which recovered over days, but not to control levels, and 3-AB did not produce this effect. This study shows that PARP inhibitors may be effective in treatment-resistant depression.

Charles University Faculty of Pharmacy in Hradec Králové Department of Pharmacology & Toxicology Student: Lucia Dziaková Supervisor: RNDr. Jakub Hofman, Ph.D. Title of diploma thesis: Study on interactions of PARP inhibitors with ABC drug efflux transporters. ATP-binding cassette (ABC) transporters are integral membrane proteins that use the energy obtained from ATP to carry transport of numerous endogenous substrances out of the cells, but attention is drawn primarily to the fact that they transfer also xenobiotics. Their overexpression in tumor tissue contributes to multidrug resistance (MDR), which in most cases leads to therapy failure. Poly(ADP-ribose)polymerase inhibitors (PARPi) represent a promising therapeutic approach in the treatment of cancers that exhibit defects in homologous recombination (HR). This work focuses on four selected PARPi (olaparib, rucaparib, niraparib, veliparib) and their interaction potential towards ABC drug efflux transporters (ABCB, ABCC1, ABCG2). In our work, we worked with MDCKII cells (parent, transduced by the transporters of interest) and utilized the principle of accumulation studies based on the measurement of fluorescence intensity of specific model substrates (hoechst33342, calcein AM, daunorubicin, mitoxantrone). We used established inhibitors of studied...

Les défauts dans les voies de réparation de l’ADN sont aujourd’hui largement exploités pour le traitement du cancer. En effet, la capacité des tumeurs à réparer les lésions induites par les traitements génotoxiques (chimio- et radiothérapie) leur confère une résistance intrinsèque ou acquise à ces traitements. Développer des inhibiteurs de réparation de l’ADN permettrait de contrecarrer cette résistance et de sensibiliser les tumeurs à ces thérapies conventionnelles. Les inhibiteurs de la Poly(ADP-ribose) polymérase (PARPi), premiers candidats de cette famille d’inhibiteurs de réparation de l’ADN, ont montré des résultats encourageants mais sont néanmoins restreints à une sous-population de tumeurs avec une déficience dans la voie de réparation par recombinaison homologue (DRH). De plus, des résistances à ces PARPi ont été constatées suite à la réactivation de la voie RH ou de voies alternatives. Il est donc urgent de développer des agents plus efficaces qui permettraient de limiter la problématique de résistance. Dans le laboratoire, nous avons identifié une nouvelle classe d’inhibiteurs de réparation de l’ADN, les Dbait, consistant en une petite molécule d’ADN double-brin qui miment une cassure double-brin (CDB). AsiDNA, une molécule de la famille Dbait, agit en séquestrant et hyper activant la protéine PARP et ses partenaires, ainsi que la protéine DNA-PK qui modifie la chromatine, inhibant ainsi le recrutement au niveau du site du dommage de plusieurs protéines de réparation des voies RH ou NHEJ. Dans ce manuscrit, nous avons étudié la question des mécanismes de sensibilité à AsiDNA, et nous avons identifié l’instabilité génétique, générée essentiellement par des défauts dans les voies de réparation des CDBs, comme caractéristique majeure pour être sensible à AsiDNA dans différents modèles de cellules et de xénogreffes. De façon intéressante, l’instabilité génétique ne corrélait pas avec la sensibilité aux PARPi, qui présentaient également un profil d’action différent d’AsiDNA. En se basant sur ces différences, et sur le mode d’action d’AsiDNA agissant en tant qu’inhibiteur de la voie RH, la combinaison de ces deux molécules permettrait de s’affranchir de la restriction génétique (DRH) essentielle pour l’efficacité des PARPi. Pour valider cette hypothèse, nous avons montré par des analyses moléculaires que l’olaparib, un PARPi, et AsiDNA préviennent le recrutement au niveau des sites des dommages de XRCC1 et de RAD51/53BP1, respectivement. La combinaison de ces deux inhibiteurs permettait l’accumulation des dommages non réparés résultant en une augmentation de la mort de cellules tumorales de différentes origines, et un retard significatif de la croissance des xénogreffes. Cependant, les cellules non tumorales ne présentaient ni une augmentation des dommages ni de la mort cellulaire. Ces résultats soulignent l’intérêt thérapeutique de la combinaison d’AsiDNA avec les PARPi qui permettrait de s’affranchir de la dépendance au statut DRH et d’élargir leur champ d’application. Dans cette thèse, nous avons également traité la question de la résistance acquise à AsiDNA. En effet, contrairement à l’imatinib et au 6-thioguanine, nous n’avons pas isolé de clones résistants à AsiDNA après des expériences de mutagénèse ou après des traitements répétés sur différents modèles cellulaires. Un tel comportement défie notre acceptation commune de la théorie Darwinienne pour expliquer la résistance des cellules tumorales aux traitements. / Defects in the DNA repair pathways are now widely exploited for the treatment of cancer. Indeed, the ability of tumors to repair the damage induced by genotoxic treatments (chemotherapy and radiotherapy) gives them an intrinsic or acquired resistance to these treatments. Developing DNA repair inhibitors would help to counteract this resistance and sensitize tumors to these conventional therapies. Poly(ADP-ribose) polymerase inhibitors (PARPi), first candidates for this family of DNA repair inhibitors, have shown encouraging results but are nevertheless restricted to a tumor subpopulation with Deficiencies in the Homologous Recombination repair pathway (HRD). In addition, resistances to these PARPi were observed following the reactivation of the HR pathway or alternative pathways. It is therefore urgent to develop more effective agents to limit the resistance problem. In the laboratory, we have identified a new class of DNA repair inhibitors, Dbait, consisting of a small double-stranded DNA molecule that mimics a double-strand break (DSB). AsiDNA, a molecule of the Dbait family, acts by hijacking and hyper activating the PARP protein and its partners, as well as DNA-PK protein that modifies chromatin, thereby inhibiting recruitment at the damage site of several DNA repair proteins. In this manuscript, we studied the issue of mechanisms of sensitivity to AsiDNA, and we identified the genetic instability, generated mainly by defects in the DSBs’ repair, as major feature to be sensitive to AsiDNA in different models of tumor cells and xenografts. Interestingly, genetic instability does not correlate with sensitivity to PARPi, which also had a different action profile than AsiDNA. Based on these differences, and on the mode of action of AsiDNA acting as an inhibitor of the HR pathway, the combination of these two molecules would allow bypassing the genetic restriction (HRD) essential for PARPi efficiency. To validate this hypothesis, we have shown by molecular analyzes that olaparib, a PARPi, and AsiDNA prevent the recruitment at damage sites of the repair proteins XRCC1 and RAD51 / 53BP1, respectively. The combination of these two inhibitors allowed the accumulation of unrepaired damage resulting in an increase of tumor cells’ death, and a significant delay in the growth of xenografts. However, non-tumor cells were not sensitive to this combined treatment. These results highlight the therapeutic interest of combining AsiDNA with PARPi to recapitulate synthetic lethality in all tumors independently of their HR status. In this thesis, we also addressed the issue of acquired resistance to AsiDNA. Indeed, contrary to imatinib and 6-thioguanine, we didn’t recover resistant clones to AsiDNA after mutagenesis or after repeated cycles of treatment on different cell models. Such behavior challenges our common acceptation of a Darwin evolution theory to explain tumor cells resistance to treatment.

Co-mentorship / Le cancer de l’ovaire est le cancer gynécologique le plus létal, et cinquième cause de mort attribuable au cancer chez les femmes nord-américaines. La létalité de cette maladie est due notamment à sa détection tardive, à sa forte hétérogénéité, et à sa résistance au traitement. L’amélioration du sort des patientes atteintes du cancer de l’ovaire passera donc par une meilleure caractérisation de la maladie, ce qui inclut la découverte de biomarqueurs et le développement de voies de traitement efficaces. Pour ce faire, des modèles d’étude représentatifs et bien caractérisés sont nécessaires, et la nature hétérogène du cancer de l’ovaire souligne le besoin d’un grand nombre de modèles diversifiés sous plusieurs aspects. Les modèles cellulaires sont une option peu dispendieuse et facile à entretenir, avec lesquels il est possible d’effectuer des expériences à haut débit. La fiabilité des modèles cellulaires se base sur l’abondance et la diversité, où un grand nombre de lignées cellulaires issues de contextes cliniques et moléculaires variés sont requises pour dresser un portrait représentatif de la maladie. Dans le cadre de cette thèse, dix nouveaux modèles de cancer épithélial de l’ovaire, incluant des lignées de deux sous-types rares, sont décrits et caractérisés rigoureusement. Ces modèles sont diversifiés en phénotypes et en génotypes, et sont démontrés comme représentatifs de la maladie de laquelle ils sont issus. Ces nouvelles lignées, s’ajoutant à plusieurs autres dérivées dans notre laboratoire, pourront servir à étudier le cancer de l’ovaire sous plusieurs aspects. Notamment, un nombre de ces lignées ont servi à l’étude de la résistance au traitement aux inhibiteurs de la polymérase de poly(ADP-ribose) (PARP), une voie thérapeutique émergente des dix dernières années. Malgré des études prometteuses, plusieurs patientes ne répondent pas au traitement initial, ou cessent de répondre après une durée de traitement, menant à une rechute. Ces deux phénomènes, dits résistance intrinsèque et résistance acquise respectivement, nuisent grandement à la survie des patientes. À l’aide de lignées cellulaires initialement sensibles, nous avons développés des modèles de résistance acquise aux inhibiteurs de PARP par exposition prolongée. Ces lignées dérivées, en combinaison avec des lignées intrinsèquement résistantes, ont mené à la découverte d’un traitement combiné synergique entre les inhibiteurs de PARP et une autre classe de molécule, les inhibiteurs de la nicotinamide phosphoribosyltransférase (NAMPT). Cette enzyme étant responsable de la synthèse du substrat de PARP, la combinaison de ces deux thérapies éradique de façon efficace les modèles résistants, autant à résistance acquise qu’intrinsèque, et ralentit la croissance tumorale en modèle murin. Les inhibiteurs de PARP étant déjà approuvés, et les inhibiteurs de NAMPT étant déjà en essais cliniques, la combinaison de ces voies thérapeutiques serait facilement envisageable en clinique, et l’universalité de son efficacité pourrait drastiquement améliorer le sort de patientes atteintes du cancer de l’ovaire, n’ayant aucun autre recours efficace. / Ovarian cancer is the most lethal gynecological cancer, and fifth leading cause of cancer-related deaths in North American women. The lethality of this disease is notably due to its late detection, its strong heterogeneity, and its resistance to treatment. Improving the fate of ovarian cancer patients will require a better characterization of the disease, which includes discovering biomarkers and effective treatment options. To this effect, representative and well characterized study models are required, and the heterogeneous nature of ovarian cancer underlines the need for a vast number of models with diverse characteristics. Cellular models are an inexpensive and easy to maintain option, with which high throughput experiments are possible. The reliability of cellular models depends on their abundance and diversity, where a large number of cell lines from various clinical and molecular contexts are required to accurately represent the disease. In the context of this thesis, ten new models of epithelial ovarian cancer, including cell lines of two rare subtypes, are described and rigorously characterized. These models are diverse in phenotype and genotype, and are shown to accurately represent the disease from which they are derived. These new cell lines, in addition to many others described in our laboratory, will serve to better study ovarian cancer and its many facets. Notably, a number of these cell lines were used to study resistance to poly(ADP-ribose) polymerase (PARP) inhibitor treatment, an emerging therapy of the past ten years. Despite encouraging studies, a number of patients do not respond initially, or cease to respond after a length of treatment, leading to relapse. These two phenomena, dubbed intrinsic and acquired resistance respectively, greatly hinder patient survival. Using initially sensitive cell lines, we developed models of acquired PARP inhibitor resistance by prolonged exposure. These derived cell lines, along with intrinsically resistant cell lines, served to discover a synergistic combination treatment between PARP inhibitors and another class of drugs, nicotinamide phosphoribosyltransferase (NAMPT) inhibitors. NAMPT as an enzyme is mainly responsible for synthesizing PARP’s substrate, and the combination of these two therapies effectively inhibits models of both intrinsic and acquired resistance, and slows tumor growth in mice. PARP inhibitors are already routinely used in the clinic, and NAMPT inhibitors are currently undergoing clinical trials, making this combination of therapeutic options easily conceivable for clinical use in the future, and the universality of its efficacy could drastically improve the fate of ovarian cancer patients who are out of options.

[ES] El curso natural de los tumores va acompañado de la acumulación progresiva de alteraciones genómicas, propiciando una cadena de eventos que resultan en inestabilidad genómica (IG). Éste fenómeno, caracterizado por alteraciones en el número de copias, constituye un hallmark genómico con impacto pronóstico más allá de la histología y otras características moleculares del tumor. En el ámbito de la investigación en oncología ginecológica, la IG ha ganado fuerza en los últimos años, permitiendo la estratificación de pacientes de acuerdo al pronóstico y la respuesta a agentes que dañan el ADN, como las terapias basadas en platinos y los inhibidores de PARP. En el cáncer de ovario, en particular, se ha descrito un subgrupo molecular caracterizado por alta incidencia de alteraciones en el número de copias relacionado con un mejor pronóstico y respuesta a quimioterapia. Esta correlación presenta la IG como un buen marcador predictivo y pronóstico. Así, un modelo basado en la IG trasladable a la práctica clínica constituirá una herramienta útil para la optimización de la toma de decisiones. La era de la medicina personalizada llegó de la mano de los estudios integrativos, donde las técnicas de alto rendimiento se aplican de manera combinada para obtener una visión molecular global de los tumores, completando y complementando la caracterización clásica a nivel anatómico e histológico. Esta tesis propone un estudio global de la IG como biomarcador pronóstico y predictivo de respuesta en cáncer ginecológico, haciendo hincapié en el cáncer de ovario seroso de alto grado y cáncer de endometrio. A través de la aplicación de estrategias basadas en NGS con la adaptación de pipelines de análisis disponibles obtuvimos los perfiles de IG de muestras de tejido fijadas en formol y embebidas en parafina, de una manera fiable, portable y coste efectiva, combinando herramientas de machine learning para ajustar modelos predictivos y pronósticos. Partiendo de esta premisa, ajustamos y validamos, en cohortes clínicas bien caracterizadas, tres modelos a partir de los datos ómicos individuales y un modelo integrativo (Scarface Score) que demostró la capacidad de predecir la respuesta a agentes que dañan el ADN en un escenario clínico concreto de pacientes con cáncer de ovario seroso de alto grado. Paralelamente, desarrollamos y validamos un algoritmo basado en el perfil de mutaciones, con impacto pronóstico, en cáncer de endometrio. Este algoritmo consiguió una estratificación que respondía al perfil de IG de los pacientes. Finalmente, se caracterizó un panel de líneas celulares de cáncer de ovario a nivel de respuesta, genético y genómico. Se interrogó el estatus de la vía de recombinación homóloga y su asociación a patrones de IG, completando el perfil molecular y estableciendo las bases para futuros estudios preclínicos y clínicos. Los resultados obtenidos en esta tesis doctoral presentan herramientas de gran valor para el manejo clínico en cuanto a la búsqueda de una medicina personalizada. Adicionalmente, diferentes estudios para trasladar el modelo predictivo a otros escenarios clínicos pueden ser explorados, usando como base el planteado, pero restableciendo puntos de corte nuevos y específicos. / [CA] El curs natural dels tumors va acompanyat de l'acumulació progressiva d'alteracions genòmiques, propiciant una cadena d'esdeveniments que resulten en inestabilitat genòmica (IG). Aquest fenomen, caracteritzat per la presencia de alteracions en el nombre de cópies, constitueix un hallmark genòmic amb impacte pronòstic més enllà de la histologia i altres característiques moleculars del tumor. En l'àmbit de la recerca en oncologia ginecològica, la IG ha guanyat força en els últims anys, permetent l'estratificació de pacients d'acord amb el pronòstic i la resposta d'agents que danyen l'ADN, com les teràpies basades en platins i els inhibidors de PARP. En el càncer d'ovari en particular, s'ha descrit un subgrup molecular caracteritzat per una alta incidència d'alteracions en el nombre de còpies relacionat amb un millor pronòstic i resposta a quimioteràpia. Aquesta correlació presenta la IG com un marcador predictiu i pronòstic adeqüat. Així, un model basat en la IG traslladable a la pràctica clínica constituirà una eina útil per a l'optimització de la presa de decisions. L'era de la medicina personalitzada va arribar de la mà dels estudis integratius, on les tècniques d'alt rendiment s'apliquen de manera combinada per a obtenir una visió molecular global dels tumors, completant i complementant la caracterització clàssica a nivell anatòmic i histològic. Aquesta tesi proposa un estudi global de la IG com a biomarcador pronòstic i predictiu de resposta en càncer ginecològic, posant l'accent en el càncer d'ovari serós d'alt grau i càncer d'endometri. A través de la aplicación d'estratègies basades en NGS amb l'adaptació de pipelines d'anàlisis disponibles, vam obtenir els perfils de IG de mostres de teixit fixades en formol i embegudes en parafina d'una manera fiable, portable i cost efectiva, combinant eines de machine learning per a ajustar models predictius i pronòstics. Partint d'aquesta premissa, vam ajustar i validar, en cohortes clíniques ben caracteritzades, tres models a partir de les dades omiques individuals i un model integratiu (Scarface Score) que va demostrar la capacitat de predir la resposta a agents que danyen l'ADN en un escenari clínic concret de pacients amb càncer d'ovari serós d'alt grau. Paral·lelament, desenvoluparem i validarem un algoritme basat en el perfil de mutacions amb impacte pronòstic en càncer d'endometri. Aquest algoritme va aconseguir una estratificació que responia al perfil de IG dels pacients. Finalment, es va caracteritzar un panell de línies cel·lulars de càncer d'ovari a nivell de resposta, genètic i genòmic. Es varen interrogar l'estatus de la via de recombinació homòloga i la seua associació a patrons de IG, completant el perfil molecular i establint les bases per a futurs estudis preclínics i clínics. Els resultats obtinguts en aquesta tesi doctoral presenten eines de gran valor per al maneig clínic en quant a la cerca d'una medicina personalitzada. Addicionalment, diferents estudis per a traslladar el model predictiu a altres escenaris clínics poden ser plantejats, usant com a base el propost però restablint punts de tall nous i específics. / [EN] The natural course of tumors matches the progressive accumulation of genomic alterations, triggering a cascade of events that results in genomic instability (GI). This phenomenon includes copy number alterations and constitutes a genomic hallmark that defines specific outcomes beyond histology and other molecular features of the tumor. In the context of gynaecologic oncology research, GI has gained strength in the last years allowing the stratification of patients according to prognosis and response to certain DNA-damaging agents, such as platinum-based therapies and PARP inhibitors. Particularly in ovarian and endometrial cancers, it has been described a molecular subgroup characterized by high copy number alterations (CNA) related to good prognosis and better response to chemotherapy. This relationship highlights GI as a predictive and prognostic biomarker. Hence, a GI-based model translated into clinical practice would constitute a tool for optimizing clinical decision-making. The era of personalised medicine arrived together with the coming of integrative studies, where results of high-throughput techniques are combined to obtain a comprehensive molecular landscape of the diseases, bringing a new paradigm to characterize the tumors beyond classical anatomic and histological characteristics. This thesis proposes a global study of the phenomenon of GI as a prognostic and predictive biomarker of treatment response in gynaecological cancers, mainly focused on high-grade ovarian cancer and endometrial cancer. Through the development of an NGS-based strategy with the adaptation of available pipelines of analysis, we obtained GI profiles on formalin-fixed paraffin-embedded samples in a reliable, portable, and cost-effective approach, with the combination of Machine Learning tools to fit prognostic and predictive models based on the integration of omic data. Based on that premise, we fit and validated, in well-characterized clinical cohorts, three single-source models and an integrative ensemble model (Scarface Score) that proved to be able to predict response to DNA-damaging agents in a clinical scenario of High-Grade Serous Ovarian Cancer. In addition, a mutational-based algorithm (12g algorithm) with prognostic impact was developed and validated for endometrial cancer patients. This algorithm achieved a GI-based stratification of patients. Finally, a panel of ovarian cancer cell lines was characterized at the response, genetic and genomic level, interrogating homologous recombination repair pathway status and its associated GI profiles, completing the molecular landscape, and establishing the basis and breeding ground of future preclinical and clinical studies. The results reported in this Doctoral Thesis provide valuable clinical management tools in the accomplishment of a reliable tailored therapy. Additionally, future studies in different tumor types and drugs for implementation of the predictive model can be planned, using as a base the defined one but re-establishing new and specific cut-offs. / The present doctoral thesis was partially funded by GVA Grants “Subvencions per a la realització de projectes d’i+d+i desenvolupats per grups d’investigació emergents (GV/2020/158)” and “Ayudas para la contratación de personal investigador en formación de carácter predoctoral” (ACIF/2016/008), “Beca de investigación traslacional Andrés Poveda 2020” from GEICO group and Phase II clinical trial (POLA: NCT02684318, EudraCT 2015-001141-08, 03.10.2015). This study was awarded the Prize “Antonio Llombart Rodriguez-FINCIVO 2020” from the Royal Academy of Medicine of the Valencian Community / López Reig, R. (2023). Genomic instability as a predictive biomarker for the application of DNA-damaging therapies in gynecological cancer patients [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/199026

No description available.