Result of Stereotactic LINAC Radiosurgery of AVM

Son, Byung Chul; Kim, Moon Chan; Kang, Jun Ki; Suh, Tae Suk; Yoon, Sei Chul

Abstract

There are several treatment modalities of cerebral arteriovenous malformation, among these, stereotactic radiosurgery is effective means of obliteration of AVM with minimal morbidity. Since July 1988 to May 1998, our institute have conducted 98 radiosurgery for arteriovenous malformations. During this period, there were 2 major changes in delivering radiation, the first one is introduction of compute r dose planning system at september 1996(CUMC normal mode), the next was CUMC Stereotactic mode using non-flat filter delivery system since september 1997. Authors report the results of the first generation LINAC(linear Accelerator) radiosurgery before introduction of computer dose planning system for radiosurgery. During the first generation period(July 1988 to september 1996), 70 patients have had stereotactic radiosurgery for vascular malformation. Among these, 49 patients could be followed at least for 2 year with angiography or MRA(magnetic resonance angiography). Hemorrhage was the most common presenting causes of undergoing radiosurgery, the parietal lobe was the most prevalent area undergoing radiosurgery for vascular malformations. With grouping of volumes, vascular malformations we re grouped as 1) volumes less than 1 ml(n= 14 ), 2) 1- 4ml (n=21), 3) 4-10ml (n=4 ), 4 ) 10-14ml(n=4 ), 5 ) volumes over 14ml (n=6). the results are as follows ' for group 1 ), total occlusion rate were 85.7 F, for group 2 ) total occlusion rate were 71.4 %, for group 3 ),4 ),5 ), total occlusion rate were 50%. As a gross result of 49 patients, total occlusion rate was 69.4%, partial occlusion was 22.4%, no response 14.3 %. 4 cases of 70 radiosurgery for vascular malformation rebleeded (5.7 %) and 2 of these patients died of bleeding (2.9%). There was no symptomatic radiation complication, but 6 patients showed increased signal intensity on T2 weighted image on follow up magnetic resonance imaging.