Les Malformations Artérioveineuses sont des lésions vasculaires évolutives. Elles peuvent avoir des conséquences neurologiques lourdes liées au risque hémorragique élevé qui est leur mode de révélation le plus fréquent. Le développement de meilleurs outils d’imagerie a beaucoup amélioré le diagnostic de ces lésions et a permis une meilleure compréhension de leur interaction avec le tissu cérébral. Néanmoins, Leur 4eme dimension et leur évolution restent encore obscures.Par ailleurs, les traitements actuels (embolisation, microchirurgie, chirurgie stéréotaxique) sont associés à des risques de morbidité et mortalité importants.Pour ces raisons, nous avons travaillé sur un traitement potentiel peu invasif: des agents antiangiogéniques. Un modèle porcin simplifié a été élaboré consistant en l’occlusion unilatérale de l’artère carotide primitive et externe par voie endovasculaire. Ce modèle a présenté des preuves d’angiogénèse et des modifications anatomopathologiques proches de celles des MAVc humaines. Le Bevacizumab, un anticorps monoclonal pour le VEGF, a été administré in situ sur ce modèle. Les résultats ont montré que l’agent antiangiogénique altère l’angiogénèse sur les specimens histologiques. L’épaisseur de la paroi des vaissaux a été stabilisée après l’injection de Bevacizumab mais le volume des retia est demeuré identique montrant que le volume dépend de l’angiogénèse mais probablement également du flux / Brain arteriovenous malformations are dynamic, evolving vascular lesions. They present high morbidity rates due to hemorrhagic presentation that is the most frequent symptom at onset and potentially high mortality rates .Up to date imaging techniques has greatly facilitated the diagnosis of these lesions and the better understanding of their relation to adjacent brain tissue. However, it remains still a challenge to define their four dimensional nature and its consequences, a fact that can actually optimize their treatment. Embolization, surgery and stereotactic radiosurgery that are currently used in the treatment of AVMs carry also significant morbi-mortality risks. For this reason, a potential minimally invasive treatment with antiangiogenic agents was tested. A swine model was firstly created using the animal’s rete mirabile and performing an endovascular occlusion of one common and external carotid artery. This simplified model presented evidence of angiogenesis and histologic findings that are also observed in human AVMs compared to a control group. Secondly, Bevacizumab that is a monoclonal antibody to vascular endothelial growth factor, was in situ administered on this model. The results showed that the antiangiogenic agent tampered angiogenesis on histologic samples by stabilizing the wall thickness of the arteries but it did not have any effect on volume that is probably depending on flow and not only on angiogenesis.

Aspectos morfológicos do nido e embolização parcial neoadjuvante sem intenção de cura de malformações arteriovenosas encefálicas, precedendo a radiocirurgia, podem ter influência no resultado final do tratamento. Métodos: série consecutiva de 47 pacientes submetidos à radiocirurgia (1 a 5 sessões), precedida ou não por embolização com cianoacrilato. Acompanhamento clínico e radiológico mínimo de 36 meses. Resultados: a apresentação hemorrágica ocorreu em 68,1% dos pacientes tratados; destes, 62,5% portavam fístula arteriovenosa dentro da malformação arteriovenosa; 83,3% ectasia venosa e 90% restrição à drenagem venosa. A taxa de oclusão de embolização seguida de radiocirurgia foi de 46,1% e da radiocirurgia isolada foi de 52,4% (p=0,671). Foram identificados como fatores favoráveis à oclusão: baixo volume nidal, ausência de fístula arteriovenosa intranidal, maior dose de radiação e baixo grau na classificação das malformações arteriovenosas encefálicas baseadas na radiocirurgia (RBAS). Conclusões: o menor volume nidal (p < 0,001), o menor grau na escala RBAS (p=0,047), a ausência de fístula arteriovenosa intranidal (p=0,001) e a maior dose prescrita (p=0,001) tiveram correlação com resultado favorável no tratamento. Embolização seguida de radiocirurgia não foi superior à radiocirurgia isolada (p=0,772). A eliminação de fístulas arteriovenosas intranidais pela embolização pode aumentar a eficácia da radiocirurgia / Partial nidal embolization preceding radiosurgery of brain arteriovenous malformations (AVM) and some morphological nidal features may be related to final results. Methods: Analysis of a longitudinal cohort of 47 consecutive patients who underwent radiosurgery preceded or not by embolization. Embolizations were performed exclusively with n-butyl cyanoacrylate. Radiosurgery was delivered either as a single or divided up to 5 equal fractions. Clinical and radiological follow up of at least 36 months was obtained. Results: Hemorrhagic presentation was seen in 68.1% of the cases; 62.5% harbored intranidal arteriovenous fistulas (AVF), 83.3% had venous ectasias and 90% had venous outflow stenosis. The occlusion rate of embolization plus radiosurgery was 46.1% and radiosurgery alone was 52.4% (p = 0.671). Variables significantly associated with obliteration were lower nidus volume, lack of intranidal arteriovenous fistula, higher radiosurgical dose and lower grades in radiosurgical-based AVM scale (RBAS). Conclusions: a small nidus (p < 0.001), a lower RBAS grade (p = 0.047), no intranidal AVF (p = 0.001) and greater radiosurgical dose (p = 0.001) were associated to better results. Embolization followed by radiosurgery was not superior to radiosurgery alone (p = 0.772). Endovascular elimination of intranidal AVF\'s may help to promote radiosurgical occlusion

Aspectos morfológicos do nido e embolização parcial neoadjuvante sem intenção de cura de malformações arteriovenosas encefálicas, precedendo a radiocirurgia, podem ter influência no resultado final do tratamento. Métodos: série consecutiva de 47 pacientes submetidos à radiocirurgia (1 a 5 sessões), precedida ou não por embolização com cianoacrilato. Acompanhamento clínico e radiológico mínimo de 36 meses. Resultados: a apresentação hemorrágica ocorreu em 68,1% dos pacientes tratados; destes, 62,5% portavam fístula arteriovenosa dentro da malformação arteriovenosa; 83,3% ectasia venosa e 90% restrição à drenagem venosa. A taxa de oclusão de embolização seguida de radiocirurgia foi de 46,1% e da radiocirurgia isolada foi de 52,4% (p=0,671). Foram identificados como fatores favoráveis à oclusão: baixo volume nidal, ausência de fístula arteriovenosa intranidal, maior dose de radiação e baixo grau na classificação das malformações arteriovenosas encefálicas baseadas na radiocirurgia (RBAS). Conclusões: o menor volume nidal (p < 0,001), o menor grau na escala RBAS (p=0,047), a ausência de fístula arteriovenosa intranidal (p=0,001) e a maior dose prescrita (p=0,001) tiveram correlação com resultado favorável no tratamento. Embolização seguida de radiocirurgia não foi superior à radiocirurgia isolada (p=0,772). A eliminação de fístulas arteriovenosas intranidais pela embolização pode aumentar a eficácia da radiocirurgia / Partial nidal embolization preceding radiosurgery of brain arteriovenous malformations (AVM) and some morphological nidal features may be related to final results. Methods: Analysis of a longitudinal cohort of 47 consecutive patients who underwent radiosurgery preceded or not by embolization. Embolizations were performed exclusively with n-butyl cyanoacrylate. Radiosurgery was delivered either as a single or divided up to 5 equal fractions. Clinical and radiological follow up of at least 36 months was obtained. Results: Hemorrhagic presentation was seen in 68.1% of the cases; 62.5% harbored intranidal arteriovenous fistulas (AVF), 83.3% had venous ectasias and 90% had venous outflow stenosis. The occlusion rate of embolization plus radiosurgery was 46.1% and radiosurgery alone was 52.4% (p = 0.671). Variables significantly associated with obliteration were lower nidus volume, lack of intranidal arteriovenous fistula, higher radiosurgical dose and lower grades in radiosurgical-based AVM scale (RBAS). Conclusions: a small nidus (p < 0.001), a lower RBAS grade (p = 0.047), no intranidal AVF (p = 0.001) and greater radiosurgical dose (p = 0.001) were associated to better results. Embolization followed by radiosurgery was not superior to radiosurgery alone (p = 0.772). Endovascular elimination of intranidal AVF\'s may help to promote radiosurgical occlusion

No description available.

Les malformations artérioveineuses (MAV) cérébrales sont une pathologie rare et hétérogène dont l’imagerie est complexe du fait de leur caractère dynamique et la prise en charge controversée notamment du fait des risques liés aux traitements proposés.La première partie de cette thèse porte sur l’imagerie des MAVs, axée sur une nouvelle technique, l’angiographie en 4 dimensions. Initialement, nous avons étudié sa faisabilité et validé cette technique en la comparant à l’angiographie cérébrale conventionnelle. Puis nous l’avons comparé à une autre modalité dynamique, l’angiographie par résonance magnétique en 4 dimensions, dans l’analyse des caractéristiques angio-architecturales des MAVs. Enfin, nous avons utilisé cette technique comme aide au planning pré et peropératoire dans la chirurgie des micro-MAVs.La seconde partie de cette thèse porte sur la prise en charge des patients porteurs d’une MAV dans le cadre d’études cliniques. Face à l’absence de consensus et aux hétérogénéités des pratiques sur la prise en charge de cette pathologie, la réflexion de ce travail a été conduite en plusieurs temps : un état des lieux basé sur une revue systématique des critiques faites à un essai randomisé sur les MAVs non rompues ; puis la conception d’une nouvelle étude pragmatique en soins courant, randomisée et intégrant le jugement clinique du praticien et de l’équipe multidisciplinaire ; l’évaluation de l’applicabilité de cette conception aux différentes pratiques via l’interrogation des praticiens par questionnaires ; la mise en place de cette étude multicentrique internationale TOBAS (Treatment of Brain Arteriovenous Malformation) ; et enfin, l’analyse des résultats de la phase pilote montrant la faisabilité de l’étude. / Brain arteriovenous malformations (AVMs) are a rare and heterogeneous pathology. Imaging anAVM is complex because of their temporal dynamic feature. The management is controversial in particular due to the risks associated with the proposed treatments.The first part of this work is about imaging AVMs. We focused on a new technique, 4-Dimensional digital substraction angiography (4D DSA). Initially, we studied its feasibility and validated this technique in comparison with conventional cerebral angiography. Then we compared 4D DSA with another dynamic modality, 4-dimensional magnetic resonance angiography (4D MRA), in the analysis of angio-architectural characteristics of AVMs. Finally, we used this technique as a supplementary tool in the pre- and intraoperative planning of micro-AVM surgery.The second part deals with the management of AVMs patients in clinical studies. Given the lack of consensus, and the heterogeneity of practices in the management of this pathology, the conduct of this work was done in several stages: a systematic review and a critical analysis of a randomized trial recently published on unruptured brain AVMs; the design of a new pragmatic randomized trial incorporating the clinical judgment of the multidisciplinary team, the evaluation of the applicability of this design to the different practices using questionnaires; the implementation of this international multi-center study called TOBAS (Treatment of Brain Arteriovenous Malformation); and finally, the analysis of the pilot phase of the study.

Introduction: The surgical and endovascular results of the treatment of pial AVM provided at our Neurosurgical centre are presented. These results are supported by neuropsychological outcomes of subgroup of treated patients. Going by these results and by an overview of literary data on the efficacy and complications of each therapeutic modality, the optimal algorithm of indications is presented Cohort of patients: The main series comprises 222 patients aged 9 to 87 years treated in the years 1998 - 2013. The surgical group consists of 85 patients, 55 patients received solely endovascular treatment. Thirty-four patients were consulted and referred directly to the Radiosurgical unit. The remaining 48 were recommended to abide by the strategy of "watch and wait". A subgroup of 66 patients, who underwent treatment of AVM was neuro-psychologically tested at least two years after treatment using a battery of tests constructed specifically for this study. A control group consisted of 10 subjects without any neurological disease. Results: In the surgical group, serious complications were 3.5% at a 96.5% therapeutic efficacy. As for AVM treated with purely endovascular methods, serious procedural complications were seen in 5.5% of patients, with efficacy totalling 36.4%. One observed patient suffered...

Introduction: The surgical and endovascular results of the treatment of pial AVM provided at our Neurosurgical centre are presented. These results are supported by neuropsychological outcomes of subgroup of treated patients. Going by these results and by an overview of literary data on the efficacy and complications of each therapeutic modality, the optimal algorithm of indications is presented Cohort of patients: The main series comprises 222 patients aged 9 to 87 years treated in the years 1998 - 2013. The surgical group consists of 85 patients, 55 patients received solely endovascular treatment. Thirty-four patients were consulted and referred directly to the Radiosurgical unit. The remaining 48 were recommended to abide by the strategy of "watch and wait". A subgroup of 66 patients, who underwent treatment of AVM was neuro-psychologically tested at least two years after treatment using a battery of tests constructed specifically for this study. A control group consisted of 10 subjects without any neurological disease. Results: In the surgical group, serious complications were 3.5% at a 96.5% therapeutic efficacy. As for AVM treated with purely endovascular methods, serious procedural complications were seen in 5.5% of patients, with efficacy totalling 36.4%. One observed patient suffered...

Les malformations artério-­‐veineuses cérébrales (MAVc) sont des pathologies vasculaires agressives présentant un risque hémorragique lourd de conséquence en terme de morbi-­‐mortalité. Les outils d’imagerie disponibles actuellement ne permettent de comprendre que difficilement leur angio-­‐ architecture. Nous avons développé dans notre travail deux outils d’imagerie permettant d’affiner la compréhension de l’anatomie des ces malformations : un algorithme de segmentation semi-­‐automatisé et un algorithme d’anamorphose sphérique convexe. Ces algorithmes ont été élaborés pour être utilisés sur les acquisitions d’angiographie rotationnelle 3D ; ils permettent de mieux visualiser la veine de drainage principale des MAVc, notamment d’identifier une sténose ou une ectasie focale sur cette veine, et également de déceler de façon plus fiable la présence d’un anévrysme intra-­‐nidal. Ces améliorations dans l’analyse de l’angio-­‐architecture des MAVc permettront vraisemblablement de réduire le risque thérapeutique pour ces malformations. En vue de tester le potentiel des agents anti-­‐angiogéniques pour le traitement des MAVc, nous avons élaboré un modèle porcin simplifié de MAVc consistant en une occlusion unilatérale d’artère carotide primitive par voie endovasculaire. La comparaison entre le volume de rete mirabile à J0 et à 3 mois et les valeurs obtenues pour un groupe témoin a montré une augmentation significative du volume du rete mirabile chez les porcs ayant eu l’occlusion carotidienne. D’autre part, une tendance nette à l’augmentation des taux de VEGF (vascular endothelial growth factor) à proximité du rete mirabile était observée dans le groupe occlusion. Enfin, des modifications anatomopathologiques proches de celles des MAVc humaines étaient visualisées sur les pièces autopsiques des rete mirabile dans le groupe occlusion. / Brain arteriovenous malformations (bAVMs) are aggressive vascular malformations presenting a haemorrhagic complication risk that may lead to severe consequences in terms of morbi-­‐mortality. Available imaging tools poorly help in understanding their angio-­‐architecture. We have developed two imaging tools improving our understanding of the anatomy of these malformations: a semi-­‐automated segmentation algorithm and a convex spherical anamorphosis algorithm. These algorithms have been elaborated for use on 3D rotational angiography acquisitions; they provide a better visualisation of the bAVMs’ main draining vein, especially for venous stenosis or for focal ectasia. They also help in depicting accurately intranidal aneurysms. These improvements in the analysis of the bAVMs’ angioarchitecure may help in reducing the therapeutic risk for these malformations. For a further testing of the potential of anti-­‐angiogenic agents for the treatment of bAVMs, we have elaborated a simplified swine AVM model consisting in the occlusion of a common carotid artery by endovascular means. The comparison between the volume of the rete mirabile at D0 and 3 months and those measured in a control group showed a significant increasing of the retia in the occlusion group. Moreover, a tendency was observed concerning an increase in VEGF (vascular endothelial growth factor) serum levels close to the rete mirabile in the occlusion group. Finally, pathological changes close to those seen in human bAVMs were observed on autopsy samples in the occlusion group.