We review 29 children (presenting between 1985–1996) with dural arteriovenous shunts. By analysing the anatomical features from axial and angiographic imaging and examining the clinical history and pathophysiological characteristics, we hypothesize that different diseases can be distinguished and divided into three groups: dural sinus malformation (DSM), infantile type of dural arteriovenous shunts (IDAVS) and adult type of dural arteriovenous shunts (ADAVS). It was helpful to classify these diseases when assessing the treatment options and long-term prognosis. Our group of 29 children comprised 19 DSM, 7 IDAVS, 3 ADAVS. A slight male preponderance was noted in the DSM group. The range of symptoms encountered included mild cardiac failure and coagulopathies, macro-crania, developmental delay, mental retardation, seizures and focal neurological deficits (in the neonates and early infancy age group) with or without haemorrhagic venous infarctions secondary to venous outlet restriction. We found all types of lesion in the neonatal age group, but in general the different types of lesion correspond to the paediatric subgroups with DSM occurring in the neonatal age group, IDAVS in infancy and ADAVS in children. DSMs are revealed in the first few months of live and the prognosis is good if the torcular is not involved. Two types can be seen: 1) DSM involving the posterior sinus with or without the confluens sinusum, with giant dural lakes and slow flow mural AV shunting. Spontaneous thrombosis may further restrict cerebral venous drainage and subsequently lead to intraparenchymatous haemorrhagic infarction. 2) DSM involving the jugular bulb with otherwise normal sinuses but associated with a high flow sigmoïd sinus AVF. The prognosis is excellent with embolisation treament. IDAVS are high flow and low pressure. The sinuses are large and patent with no lakes. Clinical onset is seen in the first few years of life and the shunts are initially well tolerated. Progressive symptoms (symptoms of raised ICP and venous ischaemia) develop at a later age and initially respond to partial embolisation. The long term prognosis is poor with neurological deterioration in early adulthood. ADAVS present in all age groups and almost all of them are located in the cavernous venous plexus. Post embolisation outcome is excellent.

Operative ligation of portosystemic shunts is effective in controlling chronic portosystemic encephalopathy (CPSE) but is associated with significant mortality. Review of the records of five patients with CPSE treated with radiologic occlusion procedures showed that these are suitable alternatives to surgery. Three patients had alcoholic cirrhosis, one had hepatic fibrosis from schistosomiasis, and one had post-necrotic cirrhosis. All had CPSE with progressive, severe cerebral impairment refractory to clinical treatment. Four patients had a spontaneous splenorenal shunt, and one had a surgically created mesocaval shunt (MCS). Partial splenic embolization was performed in two patients, direct shunt embolization was performed via percutaneous transhepatic portography in two other patients, and an MCS embolization was performed in one patient through the inferior vena cava. In four patients embolization controlled the CPSE. In the remaining patient it could not be evaluated because of his premature death from intraabdominal bleeding, a late complication of the procedure. Interventional radiologic procedures are effective in the control of CPSE in selected patients.

We report a series of 101 patients with cerebral arteriovenous malformations (CAVM), in which 23 cases presented with one or several arterial aneurysm(s) (AA). Each AA could be classified into distal intra-lesional, proximal or remote. Patients with CAVM + AA tend to be older and more frequently present with epilepsy, haemorrhage events and neurological deficits. Of these 23 patients, 16 had their AVM treated partially or totally by embolization. In our series, the endovascular treatment of the arteriovenous shunt with a proximal AA on the same vessel has resulted in at least a regression, and sometimes a disappearance of the arterial ectasia. Although partial treatment of the AVM does no erase the risk of haemorrhage from the malformation itself, it may diminish the chance of developing a flow-related AA or any other expression of the high-flow angiopathy.

The pathogenesis of intracranial arterial aneurysms (AA) remains unclear, despite their clinical importance. An improved understanding of this disease is important in choosing therapeutic options. In addition to the "classical" berry-type aneurysm, there are various other types of intracranial AA such as infectious, dissecting or giant, partially-thrombosed aneurysms. From the clinician's perspective, the hypothesis that some of these intracranial AA might be due to abluminal factors has been proposed for several years. Indeed, this hypothesis and the empirical use of anti-inflammatory drugs in giant intracranial aneurysms have been confirmed by recent studies reporting that an enzyme involved in the inflammatory cascade (5-lipoxygenase or 5-LO) promotes the pathogenesis of specific aneurysms in humans. 5-LO generates different forms of leukotrienes which are potent mediators of inflammation. Adventitial inflammation leads to a weakening of the media from the abluminal part of the vessel wall due to the release of proinflammatory factors that invade the media, thereby degrading the extracellular matrix, the elastic lamina of the vascular wall, and, finally, the integrity of the vessel lumen. This in turn results in a dilation of the vessel and aneurysm formation. Moreover, neoangiogenesis of vasa vasorum is found in close proximity to 5-LO activated macrophages. In addition to this biological cascade, we argue that repeated subadventitial haemorrhages from the new vasa vasorum play an important role in aneurysm pathogenesis, due to a progressive increase in size mediated by the apposition of new layers of intramural haematoma within the vessel wall. Intracranial giant AA can therefore be regarded as a proliferative disease of the vessel wall induced by extravascular activity. Considering certain aneurysmal vasculopathies as an abluminal disease might alter current therapeutic strategies. Therapy should not only be aimed at the intraluminal repair of the artery, but also cross the vessel wall to reach the vasa vasorum. Drug-eluting stents placed proximal to the lesion and targeted to the origin of the vasa vasorum could be considered as a potential future option. "Intelligent" MRI contrast agents (i.e., macrophage marking) could be used to detect vasa vasorum proliferation and weakening of the vessel wall in vivo.

Onyx embolization of intracranial wide-neck aneurysms is safe and effective. Morbidity and mortality rates are similar to those of other current endovascular techniques. Larger samples and longer follow-up periods are necessary.