Embolization For Supratentorial Brain Arteriovenous Malformation

SECTION I Study on correlation of angioarchitecture and hemorrhage in supratentorial brain arteriovenous malformationOBJECTIVE:Studies in patients with brain arteriovenous malformations (bAVMs) have yielded contradictory results about risk factors for hemorrhage, and further investigation should be focused on subgroups. The pattern of blood supply and drainage in brain are different between supratentorial and infratentorial region, furthermore, the angioarchitecture, manifestations and prognosis of supratentorial brain AVMs (sbAVMs) are also not identity with cerebellar and brain stem ones. We performed a retrospective study in patients with a group of supratentorial AVMs, examined the impact of demographic and angiographic features on rupture, established a new grading system to guild clinical treatment for sbAVM.METHODS:The clinical and angiographic data of 152 patients with supratentorial AVMs was retrospectively analyzed, dural arteriovenous fistula, cavernous malformation, brain non-plexus arteriovenous fistula and other vascular malformation were excluded. Baseline clinical characteristics include gender and age. Angioarchitecture of AVMs include size and anatomical location of nidus, pattern and number of feeding arteries and drainage veins, diameter of drainage vein and retrograde outflow, accompanied variceal dilatations, venous stenosis and associated aneurysm. The AVM was divided into 3 groups according to nidus diameter (<3cm,3-6cm,≥6cm). Anatomical location includes cortical functional region, cortical non-functional region, midline region (longitudinal cistern, corpus callosum and other structure located in midline) and deep location (basal ganglion, diencephalons and periventricular region). According their number, feeding artery was divided into 3 categories (1,2 and≥3). and drainage vein was divided into 3 categories (1,2 and≥3).Drainage vein was defined as superficial, deep and combined pattern, and pattern of feeding arteries include perforating and non-perforing blood supply. Associated aneurysm was divided into intranidal, prenidal aneurysm, and remote one was excluded. The data was managed and analyzed using SPSS software (version 13.0; SPSS Inc., Chicago, IL). We conducted analyses stratified by each variable (age, gender, nidus size, etc.) to assess the effect of the difference between initial hemorrhagic and nonhemorrhagic presentations. Furthermore, multivariate analyses were conducted after adjustment for some variables that demonstrated and associated with a significantly increased risk of hemorrhage in univariate anlyses, to assess the effect of modification and interaction among potential risk factors. A probability value of less than 0.05 was considered statistically significant in each analysis.RESULTS:In the 152 patients with supratentorial AVMs,92 were male and 60 female, age was 30.22±12.26 in average.86 cases presented with hemorrhage and 66 were none. According to the size,≤3cm in 40 cases,3-6cm in 89,≥6cm in 33;118 AVMs located in superficial part and 34 in deep part.18 cases with one branch of feeding arteries,37 with two,28 with three and 69 morez than 3; 67 cases with one branch of drainage veins,44 with two,41 with three and more than three. In blood supply and drainage pattern,44 cases presented with perforating blood supply, but other 108 were not; 91 were superficial venous drainage,24 were deep, and 37 were combined drainage. In the nidus of AVMs, fistula was in 11 cases; besides,34 and 49 drainage veins were demonstrated with variceal dilatation or venous stenosis respectively. Associated aneurysm was found in 43 cases,33 were intranidal aneurysms and 10 located in parent arteries. By univariate analyses, age, gender, size of nidus, the number of feeding artery and drainage vein, accompanied fistula, variceal dilatation, venous stenosis and associated aneurysm were independent with hemorrhage, but perforating blood supply, deep venous drainage, midline and deep location of AVM, The ratio of AVM volumn and venous diameter were demondtrated as risk factors of rupture of AVMs. By calculating risk index, the risk of rupture of deep location is 2.214 fold (χ2=4.038;P=0.044) than superficial location, perforating blood supply is 10.123 fold (χ2=25.904,p=0.000) than the cases with non-perforating like manifestation, deep venous drainage is 6.85 fold(x2=10.663, P=0.001)than superficial and 6.63 fold (χ2=8.427,P=0.004)than combined venous drainage respectively. Multivariate analyses demonstrated the dependent of hemorrhage with perforating blood supply, deep venous drainage, the ratio of AVM volumn and venous diameter, but midline and deep location of AVM was excluded. Hemorrhagic risk increased with the score of perforating blood supply and simple deep venous drainage.Conclusions:1.In the angioarchitecture, size of nidus, the number of feeding artery and drainage vein, accompanied fistula, variceal dilatation, venous stenosis and associated aneurysm were independent with hemorrhage;2. Perforating blood supply, deep venous drainage, midline and deep location of AVM were demondtrated as risk factors of rupture of AVMs. However, the correlation of AVM's location and hemorrhage could be a false result caused by perforating blood supply or deep venous drainage.3. The probability of rupture of supratentorial AVM can be calculated by the scale established on the risk factor of rupture, but its value should be demonstrated by prospective study. SECTION II Study on correlation of angioarchitecture and post-embolic complication in supratentorial brain arteriovenous malformationOBJECTIVE:Endovascular embolization plays an important role in the treatment of brain arteriovenous malformation (AVM), however, as curative or palliative treatment, some complication such as brain hemorrhage and ischemia will be resulted. Currently, less post-embolic complication of AVM was reported, and it is difficult for evaluating the possibility of complication before embolization. We performed a retrospective study in patients with a group of supratentorial brain AVMs (sbAVMs), examined the impact of demographic, angiographic features and intra-operative events on rupture, and guild clinical treatment for sbAVM.METHODS:The clinical and angiographic data of 141 patients with supratentorial AVMs treated with liquid embolic agents was retrospectively analyzed. Baseline clinical characteristics include gender, age and primary presentation. Angioarchitecture of AVMs include size and anatomical location of nidus, pattern and number of feeding arteries and drainage veins, origins of feeding arteries, accompanied variceal dilatations, venous stenosis and retrograde outflow, associated aneurysm. Embolization index include embolization degree and embolization of drainage vein. The AVM was divided into 3 groups according to nidus diameter (<3cm,3-6cm,≥6cm). Anatomical location includes cortical functional region, cortical non-functional region, midline region (longitudinal cistern, corpus callosum and other structure located in midline) and deep location (basal ganglion, diencephalons and periventricular region). Both feeding artery and drainage vein can be divided into 3 categories according their number (1,2, and≥3). Origins of feeding arteries divided into simply and complex (blood supply from more than two trunk).Drainage vein was defined as superficial, deep and combined pattern, and pattern of feeding arteries include perforating and non-perforing blood supply. Associated aneurysm was divided into intranidal, prenidal aneurysm, and remote one was excluded. Post-embolic complication was classified as: non-neurologic, transient and permanent neurologic deficit; furthermore, neurologic deficit was divided into ischemic and hemorrhagic. The data was managed and analyzed using SPSS software (version 13.0; SPSS Inc., Chicago, IL) and the coefficient of correlation was achieved by univariate analyses. A probability value of less than 0.05 was considered statistically significant in each analysis.RESULTS:In the 141 patients with supratentorial AVMs were treated with embolization, there were 82 men and 59 women ranging from 5 to 64 years of age (mean 30.33±12.21 years).Initial symptoms included spontaneous hemorrhage (n=86), ischemic manifestation (n=12),headaches (n=14), seizure (n=20) and incidentally disclosed after head trauma (n=9).11 patients treated with NBC A and 130 treated with Onyx. According to the size,≤3cm in 38 cases,3-6cm in 84,≥6cm in 19; 46 AVMs located in functional cortex,68 in non-functional cortex,13 in midline region and 14 in deep location.17 cases with one branch of feeding arteries,33 with two and 91 with more than 3; 62 cases with one branch of drainage veins,40 with two and 39 with more than 3. In blood supply and drainage pattern,38 cases presented with perforating blood supply, but other 103 were not; 85 were superficial venous drainage,21 were deep, and 35 were combined drainage. In the nidus of AVMs, fistula was in 9 cases; besides,32 and 48 drainage veins were demonstrated with variceal dilatation or venous stenosis respectively. Associated aneurysm was found in 39 cases,31 were intranidal aneurysms and 8 located in parent arteries. According to embolization degree,29 were total,26 were subtotal, and 86 were partial embolization; volume of liquid embolic agents using in the operation was 0.25-9.5ml (mean 2.43±1.75ml). After embolization, non-neurologic complication occur in 5 cases, transient neurologic deficit in 15 cases(10 were ischemia and 5 were hemorrhage), permanent neurologic deficit in 10 cases (7 ischemia and 3 hemorrhage), and 2 death (one case was hemorrhage and the other died in post-operative fungal infection after decompression caused by venous infact). By univariate analyses, functional cortex location (OR 2.63, 95%CI 1.17-5.92), complex origins of feeding arteries (OR 2.81,95%CI 1.12-7.05) and simple deep venous drainage (OR 4.05,95%CI 1.53-10.72) was impact on occur of post-embolic complication (p<0.05), functional cortex location mainly caused permanent ischemic neurologic deficit (p<0.05), and occlusion of drainage vein was obvious correlated with post-embolic hemorrhage when the embolization degree was less than 99%. However, multivariate analyses excluded the relation between combined blood supply and post-embolic complicatoin (p=0.493)。Conclusions:1. Embolization of supratentorial arteriovenous malformation can achieve excellent or good outcomes, the complication is few but should be considerated.2. Age, gender of the patients and primary presentation were not correlated with embolic complication.3. Embolization degree and occlusion of drainage vein were independent with embolic complication, but occlusion of drainge vein was correlated with post-embolic hemorrhage when the embolization degree was less than 99%.4. In the angioarchitecture, size of nidus, the number of feeding artery and drainage vein, perforating blood supply, complex origins of feeding arteries, accompanied variceal dilatations, venous stenosis, retrograde outflow, associated aneurysm and Spetzler-Matin grade were independent with complication, but the risk of functional cortical location and simple deep venous drainage was 2.63 and 4.05 times than non-functional cortical AVMs with superficial drainage vein. Furthermore, functional cortical AVMs was correlated with permanent ischemic neurologic deficit. SECTION III Clinical Study on Treatment of Supratentorial Brain Arteriovenous Malformation Using Non-adhesive Liquid Embolization AgentsObjective:Onyx is a new liquid polymerizing embolic material theoretically permiting slower filling, achieving better penetration and obliteration of the nidus, lacking of adherence and seems advantageous over n-BCA for embolization of brain arteriovenous malformation (AVM). Onyx is increasingly used in endovascular therapy of AVMs, however, the embolic effect and post-embolic complication are still under discussion. We report our experience in the treatment of supratentorial brain arteriovenous malformations (SBAVMs) with Onyx, statistically analyses the effect and complication.Methods:From June 2006 to July 2008,130 patients with SBAVM were embolized with Onyx. There were 78 men and 52 women ranging from 5 to 64 years of age (mean 30.27±12.45 years). Initial symptoms included spontaneous hemorrhage (n=79), ischemic manifestation (n=11), headaches (n=13), seizure (n=20) and incidentally disclosed after head trauma (n=7). Angiomorphological factors of AVMs include size and anatomical location of nidus, pattern and number of feeding arteries and drainage veins, origins of feeding arteries were achieved and the correlation of these factors and embolization degree, furthermore, post-operative complication was analyzed. The AVM was divided into 3 groups according to nidus diameter (<3cm,3-6cm,≥6cm). Anatomical location includes cortical functional region, cortical non-functional region, midline region (longitudinal cistern, corpus callosum and other structure located in midline) and deep location (basal ganglion, diencephalons and periventricular region). Both feeding artery and drainage vein can be divided into 3 categories according their number (1,2, and≥3). Origins of feeding arteries divided into simply and complex (blood supply from more than two trunk).Drainage vein was defined as superficial, deep and combined pattern, and pattern of feeding arteries include perforating and non-perforing blood supply. Embolization degree was categorized as total (100%), subtotal (≥90%) and partial (<90%) embolization. Post-embolic complication was classified as:non-neurologic,transient and permanent neurologic deficit; furthermore, neurologic deficit was divided into ischemic and hemorrhagic. The data was managed and analyzed using SPSS software (version 13.0; SPSS Inc.,Chicago, IL),and the coefficient of correlation was achieved by Spearman rank correlation analysis. A probability value of less than 0.05 was considered statistically significant in each analysis.RESULTS:In the 130 patients with supratentorial AVMs were treated with embolization,≤3cm in 32 cases,3-6cm in 80,≥6cm in 18;44 AVMs located in functional cortex,62 in non-functional cortex,12 in midline region and 12 in deep location.17 cases with one branch of feeding arteries,28 with two or three, and 57 with more than 3; 55 cases with one branch of drainage veins,37 with two and 38 with more than 3. In blood supply and drainage pattern,35 cases presented with perforating blood supply, but other 95 were not; 79 were superficial venous drainage,16 were deep, and 35 were combined drainage.In the nidus of AVMs, fistula was in 6 cases; besides, 29 and 46 drainage veins were demonstrated with variceal dilatation or venous stenosis respectively. Associated aneurysm was found in 37 cases,30 were intranidal aneurysms and 7 located in parent arteries. According to embolization degree,28 were total,26 were subtotal, and 76 were partial embolization; volume of liquid embolic agents using in the operation was 0.3-9.5ml (mean 2.5±1.77ml). After embolization, post-operative hemorrhage occur in 5 cases, transient neurologic deficit in 8 cases, permanent neurologic deficit in 6 cases, and 1 death. By univariate analyses, D≤3cm (OR11.31,95%CI4.35-29.46),single feeding artery (OR 134.67,95%CI 16.38-1107.50),double feeding artery (OR12.95,95%CI3.20-52.44). single arterial origination (OR 25.67, 95%CI 7.14-92.29),terminal blood supply (OR3.81,95%CI 1.07-13.55),single drainage vein(OR10.06,95%CI 3.51-28.86),superficial venous drainage (OR2.90,95%CI 1.08-7.74)及Spetzler-MatinⅠ-Ⅱ(OR6.84,95%CI 2.41-19.46) was impact on embolization degree (p<0.05); however, multivariate analyses found single drainage vein (p=0.003) and 1-2 feeding arteries (p=0.009) were correlated with embolization degree. Morbidity and mortality of complete embolization were 6.92% and 0.77%, and no difference with all cases.Conclusions:(1) Onyx is suitable for embolization of SBAVMs because of its diffuse controllable properties and improves the embolization effect of supratentorial brain arteriovenous malformation obviously. (2) number of feeding artery and drainage vein were correlated with embolization degree; the sbAVM with single or double feeding arteries and single drainage vein could be complete embolized more easily. (3)The embolization degree was independent with post-operative complication, complete embolization would not increase the occurrence of complication.