Impact on Seizure Control of Surgical Resection or Radiosurgery for Cerebral Arteriovenous Malformations
Wang, Joanna Y. BA; Yang, Wuyang MD; Ye, Xiaobu MD; Rigamonti, Daniele MD; Coon, Alexander L. MD; Tamargo, Rafael J. MD; Huang, Judy MD
Department of Neurosurgery, Johns Hopkins University School of Medicine, Baltimore, Maryland
Correspondence: Judy Huang, MD, Department of Neurosurgery, The Johns Hopkins University School of Medicine, Zayed Tower 6115F, 1800 Orleans Street, Baltimore, MD 21287. E-mail: email@example.com
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.neurosurgery-online.com).
Received December 08, 2012
Accepted June 24, 2013
BACKGROUND: Seizures are a common presenting symptom of arteriovenous malformations (AVMs). However, the impact of treatment modality on seizure control remains unclear.
OBJECTIVE: To compare seizure control after surgical resection or radiosurgery for AVMs.
METHODS: We analyzed retrospectively collected information for 378 patients with cerebral AVMs treated at our institution from 1990 to 2010. The application of strict inclusion criteria resulted in a study population of 164 patients.
RESULTS: In our cohort, 31 patients (20.7%) had Spetzler-Martin grade I AVMs, 51 (34.0%) grade II, 47 (31.3%) grade III, 20 (13.3%) grade IV, and 1 (0.7%) grade V. Of the 49 patients (30%) presenting with seizures, 60.4% experienced seizure persistence after treatment. For these patients, radiosurgery was associated with seizure recurrence (odds ratio: 4.32, 95% confidence interval: 1.24-15.02, P = .021). AVM obliteration was predictive of seizure freedom at last follow-up (P = .002). In contrast, for patients presenting without seizures, 18.4% experienced de novo seizures after treatment, for which surgical resection was identified as an independent risk factor (hazard ratio: 8.65, 95% confidence interval: 3.05-24.5, P < .001).
CONCLUSION: Although our data suggest that achieving seizure freedom should not be the primary goal of AVM treatment, surgical resection may result in improved seizure control compared with radiosurgery for patients who present with seizures. Conversely, in patients without presenting seizures, surgical resection increases the risk of new-onset seizures compared with radiosurgery, but primarily within the early posttreatment period. Surgical resection and radiosurgery result in divergent seizure control rates depending on seizure presentation.
ABBREVIATIONS: AVM, arteriovenous malformation
CI, confidence interval
DSA, digital subtraction angiography
SD, standard deviation
As the second most common manifestation of cerebral arteriovenous malformations (AVMs) after intracranial hemorrhage, seizures have a marked impact on quality of life.1-5 AVMs are hypothesized to create epileptic foci by causing local ischemia, gliosis, demyelination, or hemosiderin deposition.6-9 Although surgical resection or stereotactic radiosurgery, in addition to adjunctive embolization, are standard treatment options for selected AVMs, limitations in characterizing the natural history have challenged our understanding of treatment impact on patient outcomes.10-13 Few direct comparisons of treatment modalities with respect to seizure control have been performed. In this article, we compare seizure outcomes after surgical resection or radiosurgery.
PATIENTS AND METHODS
Retrospective examination of 378 patients diagnosed with a cerebral AVM at our institution over a 21-year period was conducted. All research protocols were approved by our Institutional Review Board for Human Research. Inclusion criteria consisted of AVM diagnosis after 1990 and adequate clinical data retrievable from institutional electronic medical records. A total of 156 patients were not included due to inadequate seizure history information (41.3%); untreated patients (n = 11, 2.9%), those with no follow-up information (n = 29, 7.7%), or those with other brain conditions (n = 18, 4.8%) were excluded, resulting in a study cohort of 164 patients. AVMs were graded with the Spetzler-Martin scale.14 Border zone arterial location was defined as the feeding of an AVM by 2 main cerebral arteries. Deep gray matter structures (basal ganglia, thalamus), brainstem, and functional cortical areas (sensorimotor, language, primary visual cortex) were considered eloquent locations. Hemorrhage was included as a possible predictor of seizure outcomes in our analyses.
Microsurgical resection was performed as described previously.15 Intraoperative or postoperative angiography was used to confirm complete AVM obliteration.16
Radiosurgery was performed using the Leskell Gamma Knife (Elekta, Stockholm, Sweden) or with linear accelerator-based hypofractionated stereotactic radiotherapy. Radiosurgery was performed stereotactically, using angiography, magnetic resonance imaging (MRI), and computed tomography studies for planning. The mean ± standard deviation (SD) volume treated was 11.1 ± 14.1 cm3 (range, 1.0-61.0 cm3), and the median target volume was 46.0 cm3. The mean ± SD dose was 1600.4 ± 421.4 cGy (range, 800-2500 cGy) and the median dose was 1800 cGy.
Endovascular embolization was used as described previously with a mixture of ethiodol and n-butylcyanoacrylate for 65 patients.17 32 patients (49.2%) underwent 1 procedure, 16 (24.6%) underwent 2, 9 (13.8%) underwent 3, 6 (9.2%) underwent 4, and 2 (3.1%) underwent 5.
Seizure outcomes were assessed at follow-up visits after the last treatment. Seizure control was defined as seizure freedom from the date of the last treatment to the last follow-up. For patients who presented with seizures with more than 1 year of follow-up, seizure frequency was characterized using the Engel seizure frequency scoring system.18 It was verified that no patient or AVM characteristics were significantly different between this subset and all patients with presenting seizures. Seizure frequency was assessed at 3 time points. To grade frequency at the time of seizure onset, the 12 months after the first seizure were examined. The 12 months after the date of the last treatment were examined to assess seizure frequency 1 year after treatment. To score frequency at the last follow-up, the 12 months preceding the date of the last follow-up were examined. Seizure freedom was defined as an Engel frequency score of 0, 1, or 2.
For radiosurgical patients, obliteration was determined by follow-up digital subtraction angiography (DSA) or by MRI in those without DSAs. Although the predictive value of a negative findings on MRI after radiosurgery is 91%, including only patients undergoing DSA may artificially reduce the percentage of patients with obliterated AVMs.19 Functional outcome was determined from follow-up clinic visits and defined as return to pre-AVM presentation activities (work, school, driving, or other responsibilities).
Analysis of factors associated with seizures was performed using the Pearson χ2 or Fisher exact test for categorical variables and the Student t test for continuous variables. Normality of data was verified before using parametric tests. Significant factors from univariate analysis were entered into a stepwise multivariate logistic regression analysis using backward elimination. Results are presented as odds ratios (OR) with a 95% confidence interval (CI). To compare Engel scores over time, paired t tests were used.
Kaplan-Meier survival analysis was used to determine the cumulative probability of first posttreatment seizure; the log-rank test of equality was used to analyze time to AVM obliteration. Cox proportional hazards regression analysis was used to identify risk factors for posttreatment seizures. To directly compare seizure risk after surgical resection and after radiosurgery, patients who were treated using both modalities and those who were only treated with embolization were excluded. AVM obliteration was coded as a categorical time-varying covariate, with the date of the first negative imaging study defining the date of obliteration. Hazard ratios (HRs) were calculated using a 95% CI.
Statistical significance was defined as P < .05. All statistical analyses were performed using Stata/IC 12 (StataCorp LP, College Station, Texas).
For the 164 patients with brain AVMs (Table 1), the mean ± SD age at diagnosis was 32.9 ± 15.2 years (range, 4.5-74 years). There were 60 male patients (36.6%) and 104 female patients (63.4%). The distribution of Spetzler-Martin AVM grades was as follows: grade I, 31 (20.7%); grade II, 51 (34.0%); grade III, 47 (31.3%); grade IV, 20 (13.3%); and grade V, 1 (0.7%). The mean ± SD duration of follow-up was 38.1 ± 41.6 months (range, 0.2-230.5 months).
49 patients (30%) presented with seizures. Of these patients, 9 (20.0%) had Spetzler-Martin grade I AVMs, 17 (37.8%) had grade II, 15 (33.3%) had grade III, and 4 (8.9%) had grade IV AVMs. The mean ± SD age of first seizure was 27.6 ± 14.2 years (range, 1.5-61.8 years). Factors independently associated with pretreatment seizures included anterior cerebral artery feeder (OR: 3.12, 95% CI: 1.22-7.97, P = .017) and middle cerebral artery feeder (OR: 3.87, 95% CI: 1.55-9.69, P = .004). Deep venous drainage and history of AVM hemorrhage were associated with nonseizure presentation (OR: 0.33, 95% CI: 0.13-0.83, P = .018; OR: 0.26, 95% CI: 0.10-0.71, P = .009; respectively) (Table 2). Seizure type was characterized in 45 patients: simple focal seizures in 8 (17.8%), complex focal seizures in 8 (17.8%), and generalized seizures (primarily and secondarily) in 29 (64.4%). Information on prescribed medications was available for 47 patients; 42 (89.4%) were prescribed antiepileptic drugs after seizure onset. During the year before treatment, medication information was available for 48 patients; 45 patients (93.8%) were taking antiepileptic drugs.
Treatment strategies included surgical resection alone (38 patients, 23.2%), embolization followed by surgical resection (18 patients, 11.0%), radiosurgery alone (57 patients, 34.8%), embolization followed by radiosurgery (42 patients, 25.6%), embolization alone (5 patients, 3.0%), and surgical resection and radiosurgery (4 patients, 2.4%). Thus, the treatment groups of surgical resection, radiosurgery, and embolization comprised 60, 103, and 65 patients, respectively (Table 1). The relationship between treatment modality and pretreatment or posttreatment seizure onset is depicted in Figures 1A and 2A, respectively. Superficial venous drainage and smaller AVMs were independently associated with surgical treatment (P = .028, P < .001, respectively). Only nonfrontal lobe location and greater AVM length were independently associated with radiosurgery (P = .007, P < .001, respectively). Radiographic studies for AVM obliteration were available for 158 patients at last follow-up, demonstrating that 86 (54.4%) achieved obliteration. The median time to obliteration in patients who underwent radiosurgery (94 patients) was 3.1 years (95% CI: 2.2-3.4). Among the 56 surgical patients, immediate obliteration was achieved in 54 (96.4%). Patients who underwent both surgery and radiosurgery were excluded from these analyses.
Seizure outcomes were characterized for 162 patients. A sensitivity analysis was performed to account for 2 patients with incomplete follow-up information; each was subsequently randomized to a different posttreatment seizure group. Using univariate analysis, pretreatment seizures, frontal lobe location, middle cerebral artery feeder, and duration of follow-up were associated with seizures after the last treatment. A significant interaction was detected between pretreatment seizures and treatment modality with respect to posttreatment seizure control. Multivariate analysis revealed independent associations of pretreatment seizures (OR: 22.3, 95% CI: 6.62-75.3, P < .001), duration of follow-up (OR: 1.01, 95% CI: 1.00-1.02, P = .009), and frontal lobe location (OR: 3.04, 95% CI: 1.25-7.34, P = .014) with posttreatment seizures (Table 1). A significant effect of treatment modality (OR: 7.07, 95% CI: 2.29-21.9, P = .001) and an interaction between pretreatment seizure status and treatment modality (OR: 1.01, 95% CI: 0.004-0.14, P < .001) were also found. An exploration of this interaction was performed through a subgroup analysis based on pretreatment seizure status.
Outcomes in Patients With Pretreatment Seizures
Of the 49 patients presenting with seizures, seizure outcomes were characterized for 48 patients, revealing that 19 patients (39.6%) became seizure free after treatment. Although only 41.2% of patients who underwent surgical resection continued to be treated for seizures, 73.3% of those treated with radiosurgery had seizure recurrence (Figure 1A). Using univariate analyses, follow-up duration, eloquence, and radiosurgery were predictive of seizure persistence. In contrast, surgical resection was associated with seizure freedom after treatment. Neither embolization nor previous hemorrhage predicted seizure outcomes. With multivariate analysis, only radiosurgery was independently associated with seizure recurrence (OR: 4.32, 95% CI: 1.24-15.02, P = .021). For the 97 patients who underwent radiosurgery, AVM obliteration at last follow-up was not significantly associated with seizure freedom. However, of the 33 radiosurgical patients experiencing seizures before treatment, only AVM obliteration predicted seizure control at last follow-up (OR: 0.05, 95% CI: 0.01-0.32, P = .002) (Table 3). Additionally, neither embolization itself nor the number of embolizations nor treatment order influenced seizure outcomes. No risk factors for seizure persistence, including treatment modality, were identified with the Cox proportional hazards model. However, the risk of seizure recurrence did not exceed 36.0% at 10.6 months after surgery, whereas it increased to 45.2% at 2 years and to 62.3% at 3 years after radiosurgery (Figure 1B).
Engel seizure frequency scores were significantly different between onset and at 1 year (P < .001) as well as between onset and last follow-up (P < .001) (see Figure and Table, Supplemental Digital Content 1 and 2, http://links.lww.com/NEU/A566, http://links.lww.com/NEU/A567). Of the 37 patients with adequate information to determine seizure frequency at last follow-up, 24 (64.9%) were seizure free and 8 (16.2%) had discontinuation of AEDs. AVM obliteration was achieved in 48.6% (n = 18), which was the only factor associated with seizure freedom (OR: 0.29, 95% CI: 0.003-0.28, P = .002). Of the 48 patients with information on AVM obliteration at last follow-up, functional outcomes could be assessed for 39 patients: 29 (74.4%) had returned to their previous activities without restrictions or returned to driving. Only obliteration was associated with restored functional capabilities (OR: 6.55, 95% CI: 1.17-36.6, P = .032).
Outcomes in Patients Without Pretreatment Seizures
Of the 115 patients without pretreatment seizures, seizure outcomes could be characterized in all but 1 patient, revealing a new-onset seizure incidence rate of 18.4% (n = 21). Of the patients who underwent surgery, 64.3% remained seizure free compared with 90.3% after radiosurgery (Figure 2A). Using univariate analysis, frontal lobe location and surgical resection were associated with de novo seizures. In contrast, radiosurgery was not associated with seizures. Embolization, follow-up duration, and previous hemorrhage were not predictive of seizure outcomes. By multivariate analysis, the final predictive model included frontal location (OR: 5.85, 95% CI: 1.94-17.6, P = .002) and surgical resection (OR: 6.79, 95% CI: 2.20-20.9, P = .001). Of the 15 patients who had more than 1 year of follow-up information (follow-up duration: 52.7 ± 29.0 months [mean ± SD; range, 18.6-123.2 months]), 8 (53.3%) experienced only early posttreatment seizures and were seizure free at last follow-up.
Frontal lobe location, treatment modality, and AVM obliteration were also demonstrated to be risk factors for posttreatment seizures with the Cox proportional hazards model. Due to the collinearity between the time to AVM obliteration and treatment modality, only treatment was included for consideration in the final multivariate model, which identified surgical resection rather than radiosurgery (HR: 8.65, 95% CI: 3.05-24.5, P < .001) and frontal lobe location (HR: 4.35, 95% CI: 1.70-11.2, P = .002) as independent risk factors for de novo seizures (Figure 2B).
Seizure control is an important factor to consider in the multidisciplinary management of AVMs. Pretreatment seizures, follow-up duration, and frontal lobe location were all independently associated with seizure recurrence. There was also an interaction between pretreatment seizure status and treatment modality, suggesting that seizure outcomes are dependent on seizure status before treatment. Some studies have not found an association between presenting seizures and posttreatment seizures, but this may be due to the fact that these series examined microsurgical resection, whereas our study included multiple treatment modalities, which may have distinct effects on seizure outcomes.5,20 Of 48 patients who experienced seizures before treatment, 60.4% had seizure recurrence, and we found that radiosurgery alone was associated with recurrence. However, for patients who had no seizures before treatment, de novo seizures after treatment were associated with surgical resection and frontal lobe location.
Seizure Outcomes in Patients With Pretreatment Seizures
In patients presenting with seizures, although only 39.6% of patients eventually experienced seizure freedom, surgical resection appears to result in better seizure outcomes compared with radiosurgery; 58.8% of surgical patients achieved seizure freedom compared with 26.7% of radiosurgical patients. Previously reported seizure freedom rates vary from 4% to 83% of patients after surgical resection9,20-22 and 19% to 85% after radiosurgery.23 A recent prospective study did not find a difference among treatment modalities in 5-year seizure risk, but this may be due to limitations in cohort size.24 Another multimodality study demonstrated that surgical patients were more likely to be seizure free compared with those who underwent radiosurgery.25 Additionally, in our cohort, the risk of recurrence remained stable at 36.0% less than a year after surgery, whereas risk continued to increase over time to 62.3% after radiosurgery.
Of the factors considered in this study, only complete AVM obliteration was associated with seizure freedom and restoration of functional activities at last follow-up, which is consistent with previous radiosurgical series.26,27 The relationship between AVM obliteration and treatment modality appears critical when assessing the role of obliteration on seizure outcomes. Although complete surgical resection results in immediate obliteration, obliteration after radiosurgery requires several years.25-28 In our cohort and in previous studies, radiosurgery was associated with seizure recurrence, consistent with the result that obliteration predicted seizure freedom at last follow-up.26
These results suggest that AVM obliteration may be an important outcome to achieve for seizure control. However, it has also been demonstrated that obliteration is not necessary for improved seizure control after radiosurgery or surgical resection.20,29-34 This discrepancy may be reflective of treatment selection inherent in single-modality studies, which may lead to disparate findings. In studies of multimodality treatments, obliteration was found to be predictive of improved seizure outcomes.23,25 The effects of surgical resection and radiosurgery on surrounding parenchyma may also account for part of the complex relationship of treatment, obliteration, and seizure outcomes over time.5,23,26,31,35-39 Nevertheless, obliteration is critical in assessing the risk of future hemorrhage from an incompletely treated AVM.19,39,40
Seizure Outcomes in Patients Without Pretreatment Seizures
In patients who did not present with seizures, the rate of de novo seizures after treatment was 18.4%, with surgical patients experiencing greater risk than radiosurgical patients. Of the patients who underwent surgery, 35.7% experienced new-onset seizures compared with 9.7% after radiosurgery. In previous studies of surgical resection, this rate has ranged from 6% to 31.6% and as low as 3% in a contemporary series.5,20-23,41,42 By comparison, the incidence with radiosurgery has ranged from 0 to 6.5%.23,39 The differences in these seizure-control rates may reflect heterogeneity in the choice of anticonvulsant medications. Previous reports cited the use of phenobarbital and diphenylhydantoin. In contrast, the majority of our surgical patients received phenytoin or levetiracetam, and carbamazepine and valproic acid were also used.21,22,42 Crawford et al2 also identified surgical resection and frontal lobe AVM location as risk factors for new-onset seizures compared with conservative management and nonfrontal location. In our cohort, however, de novo seizures in surgical patients were limited to the early short-term postoperative period. As surgical resection resulted in immediate AVM obliteration in the majority of patients, posttreatment seizures in these subgroups may not be intrinsically related to the AVM, but rather due to the creation of new seizure foci through the manipulation of brain parenchyma or vasculature during resection.1,43
Limitations and Future Directions
This study has several limitations; as a single-institution retrospective study, selection bias may arise from treatment selection and referral patterns. Follow-up duration varied widely, which influenced the examination of seizure outcomes over time for the majority of patients. Another complication arises inherently from studying seizure outcomes; with longer follow-up time, more seizures can be detected. In addition, determination of the precise date of AVM obliteration was restricted by the availability of radiographic studies. Within this population, few patients achieved seizure freedom without AEDs by the last follow-up; it is unclear whether these patients needed the medication or whether there was ever an attempt to discontinue the medication. Information on the practice of initiating patients on seizure prophylaxis was also incomplete. In the future, multi-institutional prospective data comparing outcomes after surgical resection or radiosurgery will further clarify the role of both treatment modalities in seizure control. Despite limitations, our study is one of the few comparing different treatment modalities with respect to seizure outcomes.
Seizure control is an important outcome to consider in the multidisciplinary management of AVMs, but the impact of treatment on seizure rates has remained unclear. Our results suggest that seizure outcomes are dependent on seizure status before treatment. Only 39.6% of patients who presented with seizures became seizure free after treatment, demonstrating that the primary goal of treatment should not be seizure control. However, surgical resection appears superior to radiosurgery in achieving seizure freedom in these patients, perhaps due to the influence of effective AVM obliteration on seizure control after surgical resection. In patients without presenting seizures, the incidence of posttreatment de novo seizures was 18.4%. Conversely, surgical resection was identified as an independent risk factor for new-onset seizures compared with radiosurgery, but primarily within the short-term posttreatment period. AVM management for posttreatment seizures should include consideration of pretreatment seizure status and the divergent effects of surgical resection and radiosurgery on seizure control.
Dr. Alexander Coon is a proctor for the Pipeline embolization device (Covidien, Mansfield, Massachusetts) and a consultant for Covidien. The other authors have no financial or institutional interest in any of the drugs, materials, or devices described in this article.
1. Murphy MJ. Long-term follow-up of seizures associated with cerebral arteriovenous malformations: results of therapy. Arch Neurol. 1985;42(5):477–479.
2. Crawford P, West C, Chadwick D, Shaw M. Arteriovenous malformations of the brain: natural history in unoperated patients. J Neurol Neurosurg Psychiatry. 1986;49(1):1–10.
3. Crawford P, West C, Shaw M. Cerebral arteriovenous malformations and epilepsy: factors in the development of epilepsy. Epilepsia. 1986;27(3):270–275.
4. Hofmeister C, Stapf C, Hartmann A, et al.. Demographic, morphological, and clinical characteristics of 1289 patients with brain arteriovenous malformation. Stroke. 2000;31(6):1307–1310.
5. Thorpe M, Cordato D, Morgan M, Herkes G. Postoperative seizure outcome in a series of 114 patients with supratentorial arteriovenous malformations. J Clin Neurosci. 2000;7(2):107–111.
6. Leblanc R, Feindel W, Ethier R. Epilepsy from cerebral arteriovenous malformations. Can J Neurol Sci. 1983;10(2):91.
7. Kraemer DL, Awad IA. Vascular malformations and epilepsy: clinical considerations and basic mechanisms. Epilepsia. 1994;35(suppl 6):S30–S43.
8. Wolf HK, Roos D, Blümcke I, Pietsch T, Wiestler OD. Perilesional neurochemical changes in focal epilepsies. Acta Neuropathol. 1996;91(4):376–384.
9. Yeh HS, Kashiwagi S, Tew JM Jr, Berger TS. Surgical management of epilepsy associated with cerebral arteriovenous malformations. J Neurosurg. 1990;72(2):216–223.
10. Ondra SL, Troupp H, George ED, Schwab K. The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg. 1990;73(3):387–391.
11. Wedderburn CJ, van Beijnum J, Bhattacharya JJ, et al.. Outcome after interventional or conservative management of unruptured brain arteriovenous malformations: a prospective, population-based cohort study. Lancet Neurol. 2008;7(3):223–230.
12. van Beijnum J, van der Worp HB, Buis DR, et al.. Treatment of brain arteriovenous malformations: a systematic review and meta-analysis. JAMA. 2011;306(18):2011–2019.
13. Barr JC, Ogilvy CS. Selection of treatment modalities or observation of arteriovenous malformations. Neurosurg Clin N Am. 2012;23(1):63–75.
14. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg. 1986;65(4):476–483.
15. Pradilla G, Coon AL, Huang J, Tamargo RJ. Surgical treatment of cranial arteriovenous malformations and dural arteriovenous fistulas. Neurosurg Clin N Am. 2012;23(1):105–122.
16. Barrow DL, Boyer KL, Joseph GJ. Intraoperative angiography in the management of neurovascular disorders. Neurosurgery. 1992;30(2):153.
17. Radvany MG, Gregg L. Endovascular treatment of cranial arteriovenous malformations and dural arteriovenous fistulas. Neurosurg Clin N Am. 2012;23(1):123–131.
18. Engel J, Van Ness P, Rasmussen T, Ojemann L. Outcome with respect to epileptic seizures. In: Engel J, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York, NY: Raven Press; 1993, pp. 609–621.
19. Pollock BE, Kondziolka D, Lunsford LD, Bissonette D, Flickinger JC. Repeat stereotactic radiosurgery of arteriovenous malformations: factors associated with incomplete obliteration. Neurosurgery. 1996;38(2):318–324.
20. Englot DJ, Young WL, Han SJ, McCulloch CE, Chang EF, Lawton MT. Seizure predictors and control after microsurgical resection of supratentorial arteriovenous malformations in 440 patients. Neurosurgery. 2012;71(3):572–580.
21. Heros RC, Korosue K, Diebold PM. Surgical excision of cerebral arteriovenous malformations: late results. Neurosurgery. 1990;26(4):570–577.
22. Piepgras DG, Sundt TM Jr, Ragoonwansi AT, Stevens L. Seizure outcome in patients with surgically treated cerebral arteriovenous malformations. J Neurosurg. 1993;78(1):5–11.
23. Hoh BL, Chapman PH, Loeffler JS, Carter BS, Ogilvy CS. Results of multimodality treatment for 141 patients with brain arteriovenous malformations and seizures: factors associated with seizure incidence and seizure outcomes. Neurosurgery. 2002;51(2):303–309.
24. Josephson CB, Bhattacharya JJ, Counsell CE, et al.. Seizure risk with AVM treatment or conservative management prospective, population-based study. Neurology. 2012;79(6):500–507.
25. Hyun SJ, Kong DS, Lee JI, Kim JS, Hong SC. Cerebral arteriovenous malformations and seizures: differential impact on the time to seizure-free state according to the treatment modalities. Acta Neurochir (Wien). 2012:154(6):1003–1010.
26. Yang SY, Kim DG, Chung HT, Paek SH. Radiosurgery for unruptured cerebral arteriovenous malformations long-term seizure outcome. Neurology. 2012;78(17):1292–1298.
27. Yang SY, Paek S, Kim D, Chung HT. Quality of life after radiosurgery for cerebral arteriovenous malformation patients who present with seizure. Eur J Neurol. 2012;19(7):984–991.
28. Shorvon SD, Andermann F, Guerrini R. The Causes of Epilepsy: Common and Uncommon Causes in Adults and Children. New York, NY: Cambridge University Press; 2011.
29. Steiner L, Lindquist C, Adler JR, Torner JC, Alves W, Steiner M. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg. 1992;77(1):1–8.
30. Gerszten P, Adelson P, Kondziolka D, Flickinger J, Lunsford L. Seizure outcome in children treated for arteriovenous malformations using gamma knife radiosurgery. Pediatr Neurosurg. 1996;24(3):139–144.
31. Eisenschenk S, Gilmore RL, Friedman WA, Henchey RA. The effect of LINAC stereotactic radiosurgery on epilepsy associated with arteriovenous malformations. Stereotact Funct Neurosurg. 1998;71(2):51–61.
32. Heikkinen E, Konnov B, Melnikov L, et al.. Relief of epilepsy by radiosurgery of cerebral arteriovenous malformations. Stereotact Funct Neurosurg. 1989;53(3):157–166.
33. Kida Y, Kobayashi T, Tanaka T, Mori Y, Hasegawa T, Kondoh T. Seizure control after radiosurgery on cerebral arteriovenous malformations. J Clin Neurosci. 2000;7(suppl 1):6–9.
34. Schäuble B, Cascino G, Pollock B, et al.. Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology. 2004;63(4):683–687.
35. Régis J, Kerkerian-Legoff L, Rey M, et al.. First biochemical evidence of differential functional effects following gamma knife surgery. Stereotact Funct Neurosurg. 1996;66(1):29–38.
36. Kurita H, Kawamoto S, Suzuki I, et al.. Control of epilepsy associated with cerebral arteriovenous malformations after radiosurgery. J Neurol Neurosurg Psychiatry. 1998;65(5):648–655.
37. Régis J, Bartolomei F, Hayashi M, Chauvel P. Gamma knife surgery, a neuromodulation therapy in epilepsy surgery. Acta Neurochir Suppl. 2002;84:37–47.
38. Lim Y, Lee C, Koh J, et al.. Seizure control of gamma knife radiosurgery for non-hemorrhagic arteriovenous malformations. In: Chang TW, Katayama Y, Yamamoto T, eds. Advances in Functional and Reparative Neurosurgery. New York, NY: Springer; 2006:97–101.
39. Guo W, Karlsson B, Ericson K, Lindqvist M. Even the smallest remnant of an AVM constitutes a risk of further bleeding. Acta Neurochir (Wien). 1993;121(3-4):212–215.
40. Hoh B, Carter B, Ogilvy C. Incidence of residual intracranial AVMs after surgical resection and efficacy of immediate surgical re-exploration. Acta Neurochir (Wien). 2004;146(1):1–7.
41. Abad J, Alvarez F, Manrique M, Garcia-Blazquez M. Cerebral arteriovenous malformations: comparative results of surgical vs. conservative treatment in 112 cases. J Neurosurg Sci. 1983;27(3):203–210.
42. Yasargil MG. AVM of the Brain: Clinical Considerations, General and Special Operative Techniques, Surgical Results, Nonoperated Cases, Cavernous and Venous Angiomas, Neuroanesthesia. New York, NY: Thieme Medical Publishers; 1988.
43. Morgan MK, Johnston IH, Hallinan JM, Weber NC. Complications of surgery for arteriovenous malformations of the brain. J Neurosurg. 1993;78(2):176–182.
The authors of this paper present the seizure as an outcome measure in a large cohort of patients treated for cerebral arteriovenous malformations (AVMs) at a single center. Long-term seizure outcome was evaluated with respect to AVM architecture, location, clinical presentation, and treatment modality. In the entire group, analysis factors that independently predict seizure outcomes were pretreatment seizure, duration of follow-up, and frontal lobe location. The pretreatment seizure appear to be most predictive of posttreatment seizure, yet the duration of follow-up appears to increase the risk the seizure rather than the opposite. It would be expected that the risk of seizure would decrease with time as more AVMs are cured by radiosurgery.
Although the authors compare patients treated with radiosurgery with those treated surgically, a large group of patients in the radiosurgery group were treated with previous embolization (25.6% of all patients). The effect of radiosurgery on AVM embolization is well-known to delay the AVM obliteration. It could be that previous embolization affects the AVM cure rate and thus have an impact on seizure outcome in this category. The seizure outcome was also presented as a function of presence of pretreatment seizures. Patients with pretreatment seizures had better seizure control in the surgical resection group (persistence of seizures in 41.2%) than the radiosurgery group (73.3%). This is the opposite in patients with no previous seizures. New-onset seizures developed in 8.7% of radiosurgery patients compared with 35.7% in the surgical group. The authors did discuss the seizure outcome, yet they did not present the risk of rebleeding post-treatment. It has been shown before that the highest risk of seizure recurrence post-treatment. Data from the literature are suggestive of higher seizure control in the radiosurgery treatment group up to 80%, but the authors were able to demonstrate that in patients presenting with seizures and treated with radiosurgery, the seizure incidence is only 18.2% if the AVMs are obliterated, and seizures persisted in 81.8% if there is AVM residual. Although the authors do not make that very clear, but it seems that one of the most important predictors of being seizure free after radiosurgery is complete obliteration of the AVM and would suggest that longer follow-up would be needed to generate final comments about the results.
The authors examined the association of posttreatment seizures based on treatment modality in 164 patients with AVMs. They have found that in patients with pretreatment seizures, radiosurgery was associated with much higher seizure recurrence (odds ratio: 4.32) compared with surgery. In contrast, they found that surgery results in high de novo seizures post-treatment in those patients without pretreatment seizures (hazard ratio: 8.65). Although seizure control has not been the primary goal of AVM treatment in the past, in light of emerging new findings on the impact of AVM treatment on overall outcome and hemorrhage risk compared with natural history, functional outcome, including seizure control, will become an increasingly important consideration. This study is therefore timely and addresses an important question. A prospective study with a larger cohort will be needed, however, to further validate the findings.
In patients with intracranial arteriovenous malformations (AVMs) without a history of seizures, what factor is associated with the lowest incidence of future seizures?
1. Surgical resection
2. Frontal lobe location
3. Stereotactic radiosurgery
5. History of hemorrhage
In patients with intracranial arteriovenous malformations (AVMs) presenting with seizures, which of the following factors may contribute to improved seizure control?
1. Surgical resection of the AVM
2. Stereotactic radiosurgery
3. Eloquent location of the AVM
4. Adjunctive embolization
5. History of hemorrhage
In patients being treated for intracranial arteriovenous malformations (AVMs), which of the following variables has the strongest impact on the functional outcome?
1. Complete AVM obliteration
2. Spetzler Martin Grade
3. Treatment modality
4. History of hemorrhage
5. Adjunctive embolization
Arteriovenous malformation; Radiosurgery; Seizure; Surgery
Supplemental Digital Content
Copyright © by the Congress of Neurological Surgeons
Highlight selected keywords in the article text.