An analysis of the dose-response for arteriovenous malformation radiosurgery and other factors affecting obliteration.

Abstract

The aim of this study was to better understand arteriovenous malformation (AVM) obliteration rates after radiosurgery. We studied obliteration after Gamma knife radiosurgery in 351 AVM patients with 3-11 years of follow-up imaging. The median marginal dose was 20 Gy (range: 12-30) and median treatment volume was 5.7 cm(3) (range: 0.26-24). Stereotactic targeting was with angiography alone in 250 AVMs, and additional magnetic resonance (MR) imaging in 101 AVMs. We documented obliteration by angiography in 193/264 (73%) AVM, and by MR alone in 75/87 (86%) AVM for a 75% corrected obliteration rate. We identified persistent out-of-field nidus in 18% of embolized vs. 5% of non-embolized patients, (P = 0.006). Multivariate analysis correlated in-field obliteration with marginal dose (P < 0.0001) and sex (P < or = 0.026, but not for overall obliteration P = 0.19). A mathematical dose-response model for overall obliteration was constructed to generate a dose-response curve for AVM obliteration with a maximum overall obliteration rate of 88% and minimal improvement above 25 Gy. We could not define the value of alpha/beta for AVM obliteration to a level of statistical significance. The rate of AVM obliteration from radiosurgery depends on the marginal dose administered with a dose-response curve that reaches a maximum of approximately 88%. The dose-response plateau reflects problems with target definition which is made more difficult by prior embolization.