Seizure control after radiosurgery for cerebral arteriovenous malformations: a 25-year experience

Restricted access


Seizures are the second-most common presenting symptom in patients with lobar arteriovenous malformations (AVMs). However, few studies have assessed the long-term effect of stereotactic radiosurgery (SRS) on seizure control. The authors of this study assess the outcome of SRS for these patients to identify prognostic factors associated with seizure control.


Patients with AVM who presented with a history of seizure and underwent SRS at the authors’ institution between 1987 and 2012 were retrospectively assessed. The total cohort included 155 patients with a mean follow-up of 86 months (range 6–295 months). Primary outcomes assessed were seizure frequency, antiepileptic drug regimen, and seizure freedom for 6 months prior to last follow-up.


Seizure-free status was achieved in 108 patients (70%), with an additional 23 patients (15%) reporting improved seizure frequency as compared to their pre-SRS status. The median time to seizure-free status was estimated to be 12 months (95% CI 0–27 months) as evaluated via Kaplan-Meier survival analysis. The mean seizure frequency prior to SRS was 14.2 (95% CI 5.4–23.1) episodes per year. Although not all patients tried, the proportion of patients successfully weaned off all antiepileptic drugs was 18% (28/155 patients). On multivariate logistic regression, focal impaired awareness seizure type (also known as complex partial seizures) and superficial venous drainage were significantly associated with a decreased odds ratio for seizure-free status at last follow-up (OR 0.37 [95% CI 0.15–0.92] for focal impaired awareness seizures; OR 0.36 [95% CI 0.16–0.81] for superficial venous drainage). The effects of superficial venous drainage on seizure outcome were nonsignificant when excluding patients with < 2 years of follow-up. AVM obliteration did not correlate with long-term seizure freedom (p = 0.202, chi-square test).


This study suggests that SRS improves long-term seizure control and increases the likelihood of being medication free, independently of AVM obliteration. Patients with focal impaired awareness seizures were less likely to obtain long-term seizure relief.

ABBREVIATIONS AED = antiepileptic drug; ARUBA = A Randomized Trial of Unruptured Brain Arteriovenous Malformations; AVM = arteriovenous malformation; IQR = interquartile range; RCT = randomized controlled trial; SRS = stereotactic radiosurgery.

Article Information

Correspondence Ajay Niranjan: University of Pittsburgh, PA.

INCLUDE WHEN CITING Published online December 14, 2018; DOI: 10.3171/2018.7.JNS18304.

Disclosures Dr. Lunsford is an AB Elekta stockholder and a consultant for Insightec, DSMB.

© AANS, except where prohibited by US copyright law.



  • View in gallery

    Kaplan-Meier estimation for the proportion of population reaching seizure-free status (A) and medication freedom (B) with increasing follow-up. The number of patients reaching each major follow-up length is denoted below the x-axis.

  • View in gallery

    The frequency of seizures preoperatively and annually for up to 4 years postoperatively. The plus signs indicate outliers in the data. Statistical significance is indicated between the leftmost portion of the horizontal lines at the top of the graph and at each of the asterisks. Not shown are seizure frequencies > 50 episodes/year.

  • View in gallery

    Post-SRS Engel seizure classification and AED therapy of 128 patients with SRS-treated AVMs and > 2 years of clinical follow-up (9 patients either had missing data to make an Engel classification [3 patients] or were missing postoperative AED data [6 patients]). Percentages are displayed as a proportion of the total patient count within each Engel classification.

  • View in gallery

    Post-SRS AED regimen in patients with > 2 years of seizure freedom as a function of AVM location (A), obliteration status (B), and location of unobliterated AVMs (C). Patients with cortical AVMs (p = 0.028, chi-square test) and unobliterated cortical AVMs (p = 0.035, chi-square test) were maintained on an increasing number of AEDs. AVM obliteration alone was not associated with AED count (p = 0.317, chi-square test).





All Time Past Year Past 30 Days
Abstract Views 432 432 65
Full Text Views 143 143 6
PDF Downloads 76 76 1
EPUB Downloads 0 0 0


Google Scholar