Clinical Applied Study Of Endovascular Interventional Treatment For Brain Arteriovenous Malformation

Objective The purpose of this part was to assess the demographic distribution and clinical and morphological characteristics of this group of patients with brain AVMs, to define each brain AVM angioarchitectural feature in this study, and to summarize and classify the angioarchitectural feature of all the cases in this group. This section of study was the base of section two and section three studies.Methods We reviewed and collected the data of 277 AVMs patients with complete past history and cerebral angiograms treated in our hospital from January 1994 to February 2005. As to the clinical presentation, the patients were divided into 3 groups according to their situations on the first admission: intracranial hemorrhage, epilepsy, and other presentations. Demographic We analyzed and recorded the basic clinical elements of each patient such as gender and age, carefully read each patient's angioarchitectural features, such as location, size, blood supply, and draining venous etc, and marked each angioarchitectural feature of all the malformed vessels following a certain standard. The data were synthesized and processed with SPSS10.0 statistical software package.Results (1)This group includes 277 brain AVMs patients aging from 6 to 69, with an average age at diagnosis of 28.34±12.99, among them, 230 (83%) were youngerthan 40, 175 (63.2%) were male and 102 (36.8%) were female. There were four multiple AVM cases, each including 2 lesions. Overall, there were 281 cases of AVMs, among which, 176 happened in male patients, and the rest 105 in female patients. (2)Among the 281 cases of AVMs, 167 (59.4%) were intracranial hemorrhage, 62 were epilepsy, and 52 had other clinical presentations. According to statistical examination, the significant differences in clinical presentations were attributed to the following factors: age, AVM location, AVM size, the pattern and amount of drainaged veins, the ratio of the diameter between main draining veins and main feeding artery, and the Spetzler scores, but were not associated with accompanying intracranial aneurysms and damages of drainaged veins. (3) Each angioarchitectural factor was interrelated: (D Deep perforating feeding arteries and venous drainage were commonly seen in deep malformations. (2) Small-sized AVM had fewer feeding arteries and fewer drainaged veins. (3) The smaller the ratio of V/F, the fewer the number of drainage was. ? Whether there were accompanying intracranial aneurysms was not associated with the AVM size and the presence of AVF.Section twoHigh risk angioarchitectural factors of hemorrhage in brain arteriovenousmalformationsObjective The purpose of this research was to study the angioarchitechitural factor related to the first hemorrhage in brain arteriovenous malformations (AVMs) , to seek the high risk angioarchitechitural factor in brain AVMs, and to guide the implementation of the management to brain AVMs.Methods We reviewed and studied the clinical presentations of the 281 cases of brain AVMs and concluded the angioarchitectural features between the hemorrhagic group and non-hemorrhagic group by comparison. All possible factors were analyzed by univariate analyses, then the significant factors from univariate analyses were used to construct a multivariate model relating the above features to the occurrence ofhemorrhage. The multivariate model was processed by Binary Logistic Regression with SPSSIO.O statistical software package. For all tests, a P value less than 0.05 was considered to be of statistical significance.Results By univariate analysis, AVM maxial size was <30 mm, infratentorial AVM, deep location, exclusively deep vein , one draining vein, and the ratio of the diameter between main draining vein and main feeding artery (V/F) was ^2 predicted hemorrhagic presentation. The presence of intracranial aneurysm, venous ectasia and venous stenosis was not associated with hemorrhage at presentation. When we used stepwise multiple logistic regression analysis, only AVM maximum diameter was <30 nun (OR:3.7; 95%CI:1.8-7.9), infratentorial AVM (OR:6.5; 95%CI:2.1-20.7) , and the ratio of V/F ^2 (OR:2.6; 95%CI: 1.3-4.9) were independent predictors (P<0.01) of hemorrhagic presentation; when the ratio between the diameter of main draining vein and the diameter of main feeding artery was ^1.5, the odds ratio was 7.1 (95%CI:2.9-17.3) ((P=0.000) .Conclusion Small AVM (<30 nun) ; infratentorial AVM and the ratio between the diameter of main draining vein and the diameter of main feeding artery ^2 were most powerful risk predictors for hemorrhagic AVM presentation. When the ratio of V/F was =￡;1.5, the bleeding risk was the highest. In contrast with many previous reports, intracranial aneurysm, venous ectasia and venous stenosis were not associated with hemorrhage in our study. But the angioarchitectural factors that predict the bleeding risk of brain AVMs should be tested in future prospective study.Section threeA proposed angioarchitecture grading system related to embolisation of brainarteriovenous malformationsObjective To analyse the risk of embolisation of brain AVMs, study the angioarchitectural factors affecting the difference of embolisation level, and to establish the angioarchitecture grading system for embolisation of brain AVMs.Methods We retrospectively analyzed the clinical data, angioarchitectural features, embolism complication and embolism level. We performed multivariate statistical analysis to determine if any of the variables was predictive of a poor outcome of embolisation (death or permanent neurological deficit), and if any of the variables was predictive of the percentage obliteration achieved by embolisation. Based on the above result, we established the angioarchitecture grading system.Results Endovascular procedures for embolisation were performed 324 times in 189 patients during an 11-year period. Embolisation was performed using N-butyl cyanoacrylate. Analyzed by patient, 3(1.6%) died and 5 (2.6%) had a permanent neurological deficit as a result of the embolisation. None of the demographic, angioarchitectural features, or procedure variables identified was predictive of a poor outcome. The size and the number of the first grade feeding artery are important factors which determine if fully embolised or determine embolisation percentage. The AVMs were given a score from 0 to 5 based on the size, number of the first grade feeding artery, and number of the second grade feeding artery. The assigned scores were as follows: nidus size (AVF=0, <30mm = l, ^30mm=2), number of the first grade feeding artery (l=0? 2—3 = 1> 554=2) and number of the second grade feeding artery (=^3 = 0, ^4=1) . Angiographic results based on percentage obliteration were grouped into four categories: complete, 80%-99%, 50%-79%, and 0-50%. In the AVMs with angioarchitecture scores 0-1, 24 (77.4%) of 31 were fully embolised, with scores of 4-5, none of the 101 were fully embolised, 82 (81.2%) were below 50% embolisation percentage. The correlation between the angioarchitecture grading system and the embolisation percentage was strong ( k =0.736> P=0.000) .Conclusion The angioarchitecture grading system predicts the embolisation percentage perfectly. The lower the scores, the higher the possibility of fully embolisation is; the higher the scores, the lower the embolisation percentage is. Combining the angioarchitecture grading system and the Spetzler grading system, we can make more rational treatment strategy for brain AVMs.