Small intestinal neuroendocrine tumor (SI-NET) is the most common small bowel tumor in Europe and USA, with an annual incidence of around 0.3-1.3/100000 persons. SI-NETs are the most common type of gastroenteropancreatic NETs (GEP-NETs), and they are known for their ability to produce hormones such as tachykinins and serotonin, as well as for their favorable long-term prognosis in comparison to gastrointestinal adenocarcinoma. The overall aim of the thesis was to investigate unknown or unclear aspects of SI-NET disease, in connection with prognosis, treatment and follow-up. Paper I confirmed several known negative prognostic factors and also showed, for the first time, that para-aortal lymph node metastases and peritoneal carcinomatosis were associated with worse survival by multivariable analyses. Locoregional surgery was associated with a low post-operative mortality, and a prolonged long-term survival by multivariable analysis. In Paper II we continued to investigate peritoneal carcinomatosis and found it be a risk factor not only for death, but also for emergency re-surgery. Furthermore, genetic analyses of samples from primary tumors in patients with and without peritoneal carcinomatosis showed a difference in the DNA between these two groups. In Paper III the outcome after liver surgery and/or radiofrequency ablation of liver metastases was investigated. To summarize, no difference in survival was seen in patients treated with surgery/radiofrequency ablation in comparison with matched controls. However, a superior radiological response of liver metasases and lower U-5-HIAA values were seen in patients subjected to liver surgery and/or radiofrequency ablation compared to matched controls. Paper IV compared ultrasonography, computed tomography and 11C-5HTP-PET in the follow-up after radiofrequency ablation of NET liver metastases. The study concluded that 11C-5HTP-PET depicted all residual tumors after RFA and that it, if used, should be combined with computed tomography for easier interpretation, as RFA areas are not clearly distinguishable with 11C-5HTP-PET alone. Paper V studied gallstone complications after somatostatin analog treatment in SI-NET patients, and concluded that there was a rather high risk to be subjected to a cholecystectomy due to biliary colic, cholecystitis, cholangitis or pancreatitis after primary surgery in somatostatin analog treated patients.

Ideally, tumors that might cause morbidity and mortality should be treated, preferably early, with proven, convincing, and effective therapy to prevent tumor progression or recurrence, while maintaining a favorable risk-benefit profile for the individual patient. For patients with colorectal cancer (CRC), this diagnostic, prognostic, and therapeutic precision is currently impossible. Despite significant improvements in diagnostic procedures, a sizable number of patients with CRC have liver metastases either at presentation or will subsequently develop it. And in many parts of the world, most cancer-related deaths are still due to metastases that are resistant to conventional therapy. Metastases to the liver occur in more than 50% of patients with CRC and represent the major determinant of outcome following curative treatment of the primary tumor. Liver resection offers the best chance of cure for metastases confined to the liver. However, due to a paucity of randomized controlled trials, its timing is controversial and a hotly debated topic. This article reviews some of the main controversies surrounding the surgical management of colorectal cancer liver metastases (CRLM).

Despite progress in our understanding of the molecular drivers that propagate the overall process of metastasis, the adaptation of specific organs upon these molecular interactions for metastatic entry remains poorly understood. This is particularly true for liver metastases, the liver being a common site for metastatic disease, and metastatic hepatic tumors are more prominent than primary hepatocellular or biliary tumors. Liver metastases most commonly arise from colorectal cancer than any other cancer and constitute one of the most detrimental outcomes of cancer, characterized by poor prognosis, high mortality, and no effective therapies available other than surgical interventions. Since interactions between tumour cells and the tumour microenvironment play an important part in the engraftment, survival, and progression of the metastases, the discovery of new drivers of liver metastasis with the potential to become therapeutic and preventive targets is required to advance the care of liver metastasis patients as well as cancer patients at risk of metastatic spread to the liver. The alteration of the physical structure of the tissue is extremely important in the progression of malignant diseases, such as cancer metastasis, as it directly affects the extravasation and colonization of tumour cells. The major hurdles in liver metastasis research, stem not only from our insufficient understanding of the molecular mechanisms directing and mediating metastasis particularly to the liver but also from the limited number of pre-clinical models available that mimic human disease and enable the study of the complex interactions between tumor cells and the liver microenvironment. The liver metastatic process underlies the acquisition of key adaptations by tumor-derived factors and is determined by both tumour-intrinsic properties and the crosstalk between tumour cells and stromal cells in the liver. A normal functioning and structurally intact extracellular matrix (ECM) constitute a hostile “soil” for seeding tumor cells to colonize. Eventually, it is the ability of tumor cells to remodel the liver microenvironment and create a supportive niche for metastatic tumor cell survival and outgrowth that determines successful metastatic colonization. Among tumour-secreted factors, which are recognized as major contributors to the formation of pre-metastatic and metastatic niches, tumor-derived extracellular vesicles (EVs) have recently arisen as crucial players in cell-to-cell communication and in the remodeling of distant microenvironments that favor organ-specific metastasis. Therefore, we sought to determine the role of tumor-derived EVs in the modulation of the liver microenvironment and their specific contribution to supporting metastatic colonization of the liver. The preliminary step to this process was to establish a model system to identify EV-associated targets and their effect on the ECM remodelling. Immunohistochemical analyses of primary colon tumour (CRC) and secondary liver metastases (CRC liver MET) tissue samples from patients with CRC revealed higher stromal TIMP1 levels in CRC liver MET than in CRC. The elevated stromal TIMP1 signature in the invasive front was associated with poor progression-free survival in patients with CRC liver MET. Our characterisation of the CRC tumour-derived EVs showed TIMP1 enrichment in the EVs (TIMP1EV) compared to its parental cell. Using cultures of primary liver fibroblasts, we could demonstrate that TIMP1 enrichment in the CRC-EVs was associated with regulation of TIMP1 levels in the EV-conditioned liver fibroblasts. Using our optimized ex vivo 3D ECM remodelling assay, we observed that pre-conditioning the liver fibroblasts with EVs from CRC cells promotes ECM remodelling. In accordance with our cell line model, we showed that serum-derived TIMP1EV from CRC patients promotes ECM remodelling. Moreover, high serum TIMP1EV expression in CRC liver MET patients was significantly associated with poor overall survival. In addition, our data also indicated that the determination of EV-associated TIMP1 is superior for non-invasive diagnosis than the analysis of soluble TIMP1 from total serum. Finally, we showed that HSP90AA is constitutively bound to TIMP1EV and that targeting HSP90AA leads to TIMP1 downregulation and inhibits ECM-mediated remodelling. This study defining the contribution of extravesicular TIMP1 to liver metastasis brings a novel insight into the molecular mechanisms through which tumor-secreted factors packaged via EVs promote remodelling of the liver microenvironment. The clinical significance of overexpression of extravesicular TIMP1 in patients with colorectal liver metastases highlights its potential as a prognostic biomarker and therapeutic target. With further clinical studies, Heparin and HSP90 inhibitors targeting the EV mediated TIMP1 regulation could be a putative treatment strategy to treat colorectal liver metastases.:Table of Contents Abbreviations v 1. Introduction 1 1.1 Colorectal cancer 1 1.1.1. Incidence and mortality 1 1.1.1. Tumor staging 2 1.1.1. Pattern of distant metastases in colorectal cancer 5 1.2 Colorectal liver metastases 6 1.2.1 Current evaluation and treatment strategies for colorectal liver metastases 7 1.2.2 The liver metastasis cascade - a multi-step process 10 1.3 Tumor microenvironment 12 1.3.1 Tumour-stroma interactions 15 1.3.2 ECM remodelling and its role in CRC tumor progression 17 1.4 Extracellular vesicles 21 1.4.1 EV types 21 1.4.2 Biogenesis and secretion of EVs 22 1.4.3 Molecular composition of EVs 24 1.4.4 Biological functions of EVs 26 1.4.5 EVs in Tumor microenvironment 28 1.4.6 EVs in Tumor-fibroblast communication 29 1.4.7 Role of EVs in colorectal cancer 31 1.5 Tissue inhibitor of metalloproteinases (TIMP1) 35 1.5.1 TIMP1 in cancer 37 2. Background and Research Aims 39 3. Material and Methods 40 3.1 Material 40 3.1.1 Devices 40 3.1.2 Additional material and equipment 42 3.1.3 Fine chemicals 43 3.1.4 Biochemicals 45 3.1.5 Primary antibodies 46 3.1.6 Secondary antibodies 47 3.1.7 Nucleic acids 47 3.1.8 Consumables 50 3.1.9 Softwares 51 3.2 Methods 52 3.2.1 Patients 52 3.2.2 Immunohistochemistry 52 3.2.3 Hematoxylin eosin staining 54 3.2.4 Cell lines 54 3.2.5 Primary liver fibroblast cell lines 54 3.2.6 Passaging and freezing of cells 55 3.2.7 Revival of frozen cells 55 3.2.8 Cell counting 56 3.2.9 EV Isolation from CRC cell lines 56 3.2.10 Isolation of serum-derived EVs from liquid biopsies 56 3.2.11 Characterisation of EVs 57 3.2.12 Treatment of Fibroblasts with EVs 58 3.2.13 Stimulation of PFs with recombinant TIMP1 59 3.2.14 RNA isolation 59 3.2.15 cDNA synthesis 59 3.2.16 Quantitative Real-Time PCR (qRT-PCR) 60 3.2.17 Protein quantification 61 3.2.18 Immunoblotting and co-immunoprecipitation 61 3.2.19 ELISA 62 3.2.20 TIMP1 Knock-Out (KO) and Over-Expression (OE) 62 3.2.21 17 AAG and HSP90AA antibody treatment 63 3.2.22 3D ECM-remodelling assay 63 3.2.23 PKH staining 65 3.2.24 In vivo experiments 65 3.2.25 DAPI staining 66 3.2.26 Tissue explant model 66 3.2.27 Statistical analysis and reproducibility 67 4. Results 68 4.1 Identification of TIMP1 as target molecule 68 4.1.1 Identification of TIMP1 as a target through data mining 68 4.1.2 Localization pattern of TIMP1 in CRC and CRC liver MET 70 4.1.3 Invasion front-specific overexpression of TIMP1 in the stroma of patients with CRC liver MET is associated with poor progression-free survival (PFS) 72 4.2 Model system to study CRC-EV mediated ECM remodelling 73 4.2.1 Investigating the role of CRC- derived EVs in the evolution of colorectal liver metastases 73 4.2.2 Characterizsation of isolated EVs from the CRC cell lines 74 4.2.3 TIMP1 enrichment in EVs derived from CRC cell lines 75 4.2.4 CRC-derived TIMP1EV regulates TIMP1 levels in recipient fibroblasts 76 4.2.5 TIMP1EV mediated TIMP1 upregulation in the recipient fibroblast is an EV-mediated effect 79 4.2.6 Recombinant TIMP-1 induces TIMP1 levels in recipient pFs in a time- and concentration-dependent manner 81 4.2.7 Alteration of TIMP1 levels in HCT 116 cells translates into EVs but does not affect EV packaging. 83 4.2.8 TIMP1EV levels in CRC EVs determine TIMP1 levels in recipient fibroblasts 85 4.2.9 EV-mediated TIMP1 upregulation in pFs induces ECM remodelling 86 4.2.10 TIMP1 levels in the PFs influence the extent of ECM remodelling 88 4.3 Clinical significance of TIMP1EV 89 4.3.1 TIMP1 enriched in serum-derived EVs of CRC patients compared to healthy controls 89 4.3.2 Serum derived TIMP1EV from CRC patients regulate TIMP1 levels in primary liver fibroblasts 91 4.3.3 Serum derived TIMP1EV from CRC patients promote ECM remodelling 93 4.3.4 TIMP1EV exhibits superior stratification power compared to soluble TIMP1 in liquid biopsies 93 4.3.5 TIMP1EV is a non-invasive independent prognostic marker in colorectal liver metastases 94 4.4 Targeting TIMP1EV mediated ECM remodelling 97 4.4.1 TIMP1EV binds to HSP90AA 97 4.4.2 HSP90 inhibition interferes with TIMP1 protein stabilisation 99 4.4.3 17AAG attenuates TIMP1EV-mediated ECM remodelling 101 4.5 EVs derived from murine CRC cell lines regulate TIMP1 levels in recipient fibroblasts 104 4.6 Increased homing of CRC EVs to the liver compared to other organs 106 4.7 TIMPEV regulates TIMP1 levels in liver tissues 108 5. Discussion 112 5.1 TIMP1 Localization and its significance in liver metastases 112 5.2 Model system to study the role of CRC-EVs in liver metastasis 113 5.3 In-vitro model to study the pro-metastatic effects of TIMP1EV 114 5.4 Serum-derived extravesicular TIMP1 and its pro-metastatic functions underlying remodeling of the extracellular matrix 116 5.5 Clinical significance of TIMP1EV in colorectal liver metastases 117 5.6 Scope of HSP90 inhibitors in the prevention and treatment of CRC liver metastases...……………………………………………………………………………………..118 6. Future perspectives and concluding remarks 120 7. Graphical summary of the findings 122 Zusammenfassung 123 Summary 125 List of figures 127 List of Tables 129 References 130 Acknowledgements 163 Appendix 165

Yes / There remains an unmet need for safe and cost-effective adjunctive treatment of advanced colorectal cancer (CRC). The omega-3 polyunsaturated fatty acid eicosapentaenoic acid (EPA) is safe, well-tolerated and has anti-inflammatory as well as antineoplastic properties. A phase 2 randomised trial of preoperative EPA free fatty acid 2 g daily in patients undergoing surgery for CRC liver metastasis showed no difference in the primary endpoint (histological tumour proliferation index) compared with placebo. However, the trial demonstrated possible benefit for the prespecified exploratory endpoint of postoperative disease-free survival. Therefore, we tested the hypothesis that EPA treatment, started before liver resection surgery (and continued postoperatively), improves CRC outcomes in patients with CRC liver metastasis. Methods and analysis: The EPA for Metastasis Trial 2 trial is a randomised, double-blind, placebo-controlled, phase 3 trial of 4 g EPA ethyl ester (icosapent ethyl (IPE; Vascepa)) daily in patients undergoing liver resection surgery for CRC liver metastasis with curative intent. Trial treatment continues for a minimum of 2 years and maximum of 4 years, with 6monthly assessments, including quality of life outcomes, as well as annual clinical record review after the trial intervention. The primary endpoint is CRC progression-free survival. Key secondary endpoints are overall survival, as well as the safety and tolerability of IPE. A minimum 388 participants are estimated to provide 247 CRC progression events during minimum 2-year follow-up, allowing detection of an HR of 0.7 in favour of IPE, with a power of 80% at the 5% (two sided) level of significance, assuming drop-out of 15%. Ethics and dissemination: Ethical and health research authority approval was obtained in January 2018. All data will be collected by 2025. Full trial results will be published in 2026. Secondary analyses of health economic data, biomarker studies and other translational work will be published subsequently. Trial registration number NCT03428477. / The EMT2 trial is funded by Yorkshire Cancer Research (L387) and is sponsored by the University of Leeds. The EMT2 biospecimen collection is funded by the National Institutes of Health (1R01CA243454-01A1) and is sponsored by the University of Leeds ( governance-ethics@ leeds. ac. uk). Both studies have been adopted to the NIHR Clinical Research Network (CRN) Portfolio (CPMS ID 34700 and 47372, respectively) and have benefited from CRN research staff support.

Les métastases hépatiques du cancer colorectal (MHCR) sont fréquentes et sont associées à une mortalité significative. Ces dernières décennies, des progrès thérapeutiques importants ont permis d'en améliorer le pronostic. Plus particulièrement, le rôle des résections hépatiques s'est considérablement élargi dans la maladie métastatique limitée au foie, conduisant à un changement radical dans la prise en charge. Ceci a naturellement eu des répercussions sur l'imagerie, qui doit être très performante au niveau lésionnel, nécessitant des sensibilité et spécificité très élevées. Si les techniques modernes ont permis une amélioration très nette en termes de sensibilité, en particulier grâce aux produits de contraste hépatospécifiques et de l'imagerie pondérée en diffusion, l'amélioration de la spécificité est moins claire et moins bien évaluée. Pourtant, la spécificité est tout aussi importante dans ce contexte, où les erreurs diagnostiques sont coûteuses, avec des chirurgies inutiles en cas de faux positifs, et des résections incomplètes en cas de faux négatifs. Ces deux situations sont accompagnées d'une morbi-mortalité très importante. Le présent travail de thèse va donc explorer de nouvelles pistes dont l'objectif ultime serait d'améliorer la spécificité de l'imagerie des MHCR. La première étude confronte l'apparence des métastases d'origine colorectale en IRM de haute résolution et leur histologie sous-jacente. Cette étude originale démontre que la fibrose tumorale apparait en hypersignal T2 et la nécrose tumorale en hyposignal T2 et hypersignal T1, ce qui va à l'encontre du dogme classique. La seconde étude explore la faisabilité de la fusion d'images IRM/échographie dans l'exploration de lésions hépatiques focales chez des patients ayant un cancer colorectal. Cette étude montre qu'un nombre significatif de lésions ne peuvent être visualisées à l'échographie qu'en utilisant la fusion, ouvrant la voie à une meilleure caractérisation lésionnelle en combinant les atouts de l'échographie et de l'IRM. Enfin, la troisième étude, complétée d'une revue iconographique, analyse le comportement IRM des lésions hépatiques après injection d'un produit de contraste intravasculaire. Elle montre une accumulation progressive du contraste au sein des angiomes, mais pas dans les métastases, conduisant à des apparences très différentes sur la phase tardive. Ceci était aussi observé dans les lésions de petites tailles, ce qui devrait permettre une meilleure spécificité dans les cas difficiles / Colorectal cancer liver metastases (CRCLM) are common and result in significant mortality. During the past decades, important therapeutic advances have improved the prognosis signficantly, especially through a marked expansion of the role of hepatic resections in liverlimited metastatic disease, leading to a radical change in management. This was naturally accompanied by an equally radical change in the imaging paradigm, now centered at the lesion level and not at the patient level, requiring very high sensitivity and specificity. While modern techniques have allowed a significant improvement in terms of sensitivity, especially through the use of hepatospecific contrast agents and diffusion imaging, the benefits in term of specificity are less clear, with only few studies focusing on and reporting the specificity of the techniques. However, specificity is equally important in this context, where diagnostic errors are costly, resulting either in unnecessary surgeries in case of false positives or in incomplete resections in case of false negatives. In this setting, our thesis will examine the results of three studies, which objective is to offer possible solutions to better understand the imaging of metastases and improve the specificity of liver imaging of CRCLM. The first study analyzes the association between high resolution MRI appearance of CRCLM and their underlying histology, showing that tumor fibrosis was in hypersignal on T2 Weighted Imaging while tumor necrosis was in hyposignal on T2 Weighted Imaging and hypersignal on T1 Weighted Imaging, which goes against the classical teaching about these lesions. The second study assesses the feasibility of using an MRI/Ultrasound fusion system in the exploration of liver lesions in patients with colorectal cancer. This study shows that more lesions were detected with ultrasound when using the fusion system, suggesting that a fusion system may allow a better characterization of lesions by combining the complementary information of MRI and ultrasound. Finally, the third study and its accompanying pictorial essay, explored the behavior of liver lesions after injection of an intravascular contrast agent. The main finding of this study was that hemangioma were accumulating the contrast over time while metastases were not, a key differentiating feature. This finding was found even in small lesions, often difficult to diagnose, suggesting that using such contrast in the exploration of liver lesions in patients with CRCLM would result in a higher specificity of the method

La résection hépatique est le seul traitement curatif des tumeurs malignes du foie et l’insuffisance hépatique est la première cause de morbi-mortalité après hépatectomie. L’amélioration du traitement des tumeurs du foie inclut le dévelopement de statégies capables de prévenir ou de traiter cette complication. Sur une série prospective récente de 232 hépatectomies (avec 0,8 % de mortalité à 3 mois) pour métastases hépatiques de cancer colorectal (MHCCR), une insuffisance hépatique était présente dans 7 % des cas d’hépatectomies majeures et était le facteur de plus mauvais pronostic pour la survie à 2 ans. Notre objectif a été d’évaluer de nouvelles approches thérapeutiques pour leur capacité à corriger l’insuffisance hépatique post-opératoire après hépatectomie élargie pour MHCCR. Un moyen mécanique de modulation pneumatique de l’hémodynamique portale et un moyen pharmacologique d’utilisation d’un facteur de survie des hépatocytes, la protéine recombinante HIP/PAP, ont été testés respectivement chez le porc et le rat. Nous montrons qu‘après hépatectomie (PHX) de 70 % sur foie normal, l’injection systémique de protéine HIP/PAP en péri-opératoire stimulait la régénération hépatique et améliorait la fonction hépatique chez le rat. En présence de MHCCR en place depuis 7 jours, la protéine HIP/PAP n’aggravait pas la maladie métastatique, et semblait au contraire diminuer la croissance tumorale après hépatectomie. In vitro, HIP/PAP n’augmentait pas la croissance tumorale de lignées cellulaires transformées dérivées de cancer du colon. Chez le porc, une sténose portale pneumatique pendant et après PHX majeure laissant en place moins de 0,5% du poids corporel entraînait une diminution de la pression portale intra-hépatique sans modifier le débit comparé au groupe sans anneau. Cette modulation de la pression était associée à une diminution significative de la bilirubine sérique et des lésions histologiques du foie restant au 7ème jour après chirurgie. Au total, il serait possible grâce à la protéine HIP/PAP et à l’anneau portal de corriger l’insuffisance hépatique post-hépatectomie en protégeant les cellules du foie de la mort et du stress secondaires aux modifications hémodynamiques et biochimiques. La question de leur association pour diminuer l’incidence de l’insuffisance hépatique reste entière. / Liver resection is the only curative treatment of tumoral liver malignancies and post-operative liver insufficiency is the 1st cause of post-operative mortality. Tumor liver treatment improvements must be associated to innovatives methods to prevent and cure this complication. On a recent prospective study including 232 hepatectomies (with a 3-month post-operative mortality of 0.8 %) for colorectal liver metastases (CRLM), post-operative liver insufficiency was present after 7 % of major hepatectomies and was the worst 2-year survival prognostic factor. To prevent this complication, we have evaluated a perioperative systemic injection of a recombinant anti-oxydant protein (HIP/PAP) before a major hepatectomy for CRLM in rats. In vitro then in model of implanted CLRM since 7 days, we have showed that HIP/PAP did not increased tumoral growth. After 70 % hepatectomy, perioperative systemic HIP/PAP injection improved liver function. After 70 % hepatectomy, HIP/PAP seemed decrease tumoral growth of implanted CRLM. On pig, we have assessed the consequences of a portal vein stenosis with pneumatic ring during and after a major hepatectomy that conserved a liver volume < 0.5 % of body weight. With this method, we have showed that the portal stenosis decreased intra-hepatic portal pressure wihout modify the portal flow by comparison with pigs without ring. This device was associated with a significant diminution of bilirubin plasmatic concentration and liver remnant histological lesions at sacrifice on post-operative day 7. Overall, chemical and physical methods and moreover their combination should allowed to decrease post-operative liver insufficicency.

No description available.

No description available.

The purpose of this work was to further understand nutritional status, especially body weight and composition, during colorectal cancer progression. Population-based studies of colorectal cancer patients were conducted using administrative health data (primary and co-morbid diseases, demographics), and computed tomography (CT) imaging (body composition). In cohort 1, administrative health data was used to study comorbidities and nutritional status in 574 colorectal cancer patients referred for chemotherapy. Multivariate Cox regression revealed several comorbidities, performance status and weight loss 20% predicted survival. In cohort 2, a serial CT image analysis assessed longitudinal body composition changes during the last 12 months preceding death from colorectal cancer (n=34). Body composition changes were typified by exponential increases in liver metastases with concurrent accelerations of muscle and fat loss. These results have the potential to make a difference in how colorectal cancer patients are treated and researched by dietitians, oncologists, and health services researchers. / Nutrition and Metabolism

Ideally, tumors that might cause morbidity and mortality should be treated, preferably early, with proven, convincing, and effective therapy to prevent tumor progression or recurrence, while maintaining a favorable risk-benefit profile for the individual patient. For patients with colorectal cancer (CRC), this diagnostic, prognostic, and therapeutic precision is currently impossible. Despite significant improvements in diagnostic procedures, a sizable number of patients with CRC have liver metastases either at presentation or will subsequently develop it. And in many parts of the world, most cancer-related deaths are still due to metastases that are resistant to conventional therapy. Metastases to the liver occur in more than 50% of patients with CRC and represent the major determinant of outcome following curative treatment of the primary tumor. Liver resection offers the best chance of cure for metastases confined to the liver. However, due to a paucity of randomized controlled trials, its timing is controversial and a hotly debated topic. This article reviews some of the main controversies surrounding the surgical management of colorectal cancer liver metastases (CRLM).