The Study Of Long-Term Hemorrhage Risk In Children With Brain Arteriovenous Malformations

ObjectiveIn order to clarify the risk between initial presentation in children with brain arteriovenous malformations (BAVM) and long-term hemorrhage risk was confounded, we compared the long-term hemorrhage risk between children and adults with BAVM. While providing guidance for the clinical treatment of children with BAVM.MethodsA total of391cases of BAVM patients were recruited in the study from July2009to March2014in Second Xiangya Hospital, the Central South University. Of which there were99cases of children and292cases of adults, ranging from5years old to76years old (28.6±12.4). All patients were diagnosed with the examination of cerebral angiography (DSA).Chi-square tests were performed to handle the count data of patient demographics, radiographic features of the BAVM; a survival analysis was performed to investigate the long-term risk of bleeding between the two groups:the period at risk began on the date of the first diagnosis of BAVM, and ended on the date of a subsequent intracranial hemorrhage (ICH). Univariate and multivariate Cox proportional hazards regression model were introduced to analyze the potential risk factors of the long-term hemorrhage.In multivariate analysis, by adding and excluding the" bleeding variable " to analysis the affect of the" bleeding variable " on the long-term hemorrhage. And to assess the extent of this effect, we compared the hazards ratio for childhood presentation with and without this predictor in the multivariate model described above.ResultDuring the follow-up, there were8subsequent ICH events within the99cases of children, and19subsequent ICH events within the292cases of adult patients with BAVM. The long-term annual average rate of bleeding was similar between the two groups:children2.74%VS adult2.35%(p=0.985, log-rank test).The long-term hemorrhage risk of children whose the initial BAVM presentation shown to be hemorrhage seemed to be a little higher than that of adults, but there was not statistically significant (p=0.057, log-rank test). In univariate Cox regression analysis, hemorrhagic presentation does increase the long-term hemorrhage risk (HR,2.84;95%CI,1.70-4.73; p<0.0001). When hemorrhagic presentation was excluded in the analysis of the multivariate model, this effect disappeared (HR,0.16;95%CI,0.02-1.26; p=0.082).Conclusion1. Compared with adults, the initial presentation of children with BAVM is more likely to present with hemorrhage.2. Hemorrhagic presentation does increase the long-term hemorrhage risk.3. The long-term hemorrhage risk of children is not higher than that of the adults. Without considering the cumulative risk of bleeding, children with BAVM do not need to more aggressive treatment than adults;4. Confounding by hemorrhagic presentation should be considered and excluded in any study comparing BAVM hemorrhage rates in children versus adults.