Contribution of technological progress, inter-operator difference and experience of operators in Gamma Knife radiosurgery for arteriovenous malformation

Background: Outcomes of microsurgical resection for cerebral arteriovenousmalformation (AVM) largely depend on the skill and experience of the operator, but it is still unknown whether such individual differences similarly exists in stereotactic radiosurgery (SRS) for AVM. The purpose of this study was to assess the influence of the inter-operator difference and technological progress in SRS for AVM. Methods: During the past 20 years, 514 patients with AVM were treated by SRS by four neurosurgeons. Until 1992, angiography was solely used for dose planning, and computed tomography (CT) or magnetic resonance imaging (MRI) was jointly used thereafter. In the early years, dose planning was calculated with the first-generation computer system, KULA, and manually superimposed on the radiographical images. After 1998, treatment planning was made on the computer monitor with sophisticated dose-planning software, GammaPlan. The influence of inter-operator difference, the operator's experience, and radiographical or radiosurgical technologies on the rates of obliteration and morbidity was assessed by multivariate analyses. Results: The factors associated with higher obliteration rates were higher margin dose (p=0.003) and the presence of hemorrhagic event before SRS (p=0.002). There was no significant difference in either obliteration rate or morbidity among the five operators. However, after introduction of CT and MRI on dose planning, the risk of adverse events was significantly decreased. Especially for AVM larger than 3 cm in maximum diameter, each operator's experience (p=0.040) and use of GammaPlan (p=0.015) reduced morbidity. Conclusions: Inter-operator difference was not a significant factor associated with the rates of obliteration and the risk of adverse events after SRS for AVM in the multivariate analyses. Progress of the sophisticated planning software and the experience of the operator were associated with lower morbidity for larger lesions.