Seizure, second only to hemorrhage, is a common

Transcription

1 CLINICAL ARTICLE J Neurosurg 126: , 2017 Seizure outcomes after stereotactic radiosurgery for the treatment of cerebral arteriovenous malformations Benjamin J. Ditty, MD, Nidal B. Omar, MD, Paul M. Foreman, MD, Joseph H. Miller, MD, Kimberly P. Kicielinski, MD, Winfield S. Fisher III, MD, and Mark R. Harrigan, MD Department of Neurosurgery, University of Alabama at Birmingham, Alabama OBJECTIVE Patients with cerebral arteriovenous malformations (AVMs) commonly present with seizure. Seizure outcomes in patients treated with stereotactic radiosurgery (SRS) are poorly defined. A case series of patients with cerebral AVMs treated with SRS is presented to evaluate long-term seizure outcome. METHODS A retrospective review of the medical record was performed, identifying 204 consecutive patients with AVMs treated with SRS between January 1991 and June Clinical and radiographic data were evaluated. Seizure outcome was measured using the Engel Epilepsy Surgery Outcome Scale. Mean duration of follow-up was 37.1 months (SD 38.3 months) with a minimum follow-up period of 1 month. RESULTS Of the 204 patients with cerebral AVMs treated with SRS, 78 patients (38.2%) presented with seizures and 49 of those patients were treated with antiepileptic drugs (AEDs). Following SRS, 63 (80.8%) of the 78 patients who had had seizures prior to SRS were seizure-free at a mean follow-up time of 37.2 months (SD 41.3 months). Of the 49 patients who had been treated with AEDs, 17 were still taking AEDs at last follow-up. Of the 126 patients who did not present with seizures prior to treatment with SRS, only 5 patients (4.0%) had seizures in the post-srs period. There was no significant correlation between post-srs seizure status and patient demographic features, comorbidities, AVM characteristics, history of operative intervention, pre- or posttreatment hemorrhage, or radiographic degree of AVM resolution. CONCLUSIONS Stereotactic radiosurgery for treatment of cerebral AVMs is effective at providing long-term control of seizures. A substantial number of patients who were treated with SRS were not only seizure free at their last follow-up, but had been successfully weaned from antiepileptic medications. KEY WORDS seizure; stereotactic radiosurgery; arteriovenous malformation; AVM; vascular disorders Seizure, second only to hemorrhage, is a common presenting symptom of cerebral arteriovenous malformation (AVM). 3,13 It has been estimated that between 17% and 30% of patients with AVMs will have seizures as part of their symptomatology. 6 Seizures negatively affect patients quality of life and can render them dependent on antiepileptic medications. Additionally, the mortality rates for patients with epilepsy are 2 to 3 times greater than those of their peers. 18 The primary goal of stereotactic radiosurgery (SRS) for the treatment of AVMs is the prevention of hemorrhage through AVM obliteration. For patients with seizures due to an AVM, a secondary benefit of treatment may be resolution of seizures or a reduction in seizure frequency. The effect of SRS on seizures in patients with brain AVMs is not clear. A recent single-center retrospective study found that for patients with seizures secondary to an AVM, those treated with radiosurgery were more likely to experience seizure persistence than patients treated with open surgery. 20 However, the study also found that for patients without preexisting seizures, those treated with open surgery were more likely to experience posttreatment seizures than those treated with radiosurgery. 20 A case series of cerebral AVMs treated with SRS is presented to further elucidate seizure outcomes in this population. ABBREVIATIONS AED = antiepileptic drug; AVM = arteriovenous malformation; CTA = CT angiography; DSA = digital subtraction angiography; ICH = intracerebral hemorrhage; SRS = stereotactic radiosurgery. SUBMITTED October 22, ACCEPTED December 29, INCLUDE WHEN CITING Published online April 8, 2016; DOI: / JNS AANS, 2017 J Neurosurg Volume 126 March

2 B. J. Ditty et al. Methods Two hundred four consecutive patients with AVMs treated with SRS between January 1991 and June 2012 were prospectively entered into a radiosurgical database. Institutional review board approval was obtained, and the medical records of these patients were retrospectively reviewed. Patient demographic characteristics, comorbidities, presentation, and adverse events; AVM size, location, and drainage pattern; and seizure history were assessed. Patients were seen in the clinic at 1 month posttreatment to assess their clinical status. Following the initial visit, they were seen in the clinic for MRI at 3 months, then at 6 12 month intervals for MRI for 3 years. At 3 years, either CT angiography (CTA) or digital subtraction angiography (DSA) was performed to assess AVM involution. Seizure outcomes were measured using the Engel Epilepsy Surgery Outcome Scale, in which Class I corresponds with a lack of disabling seizures, Class II denotes rare disabling seizures, Class III indicates worthwhile improvement in seizure frequency, and Class IV signifies no worthwhile improvement in seizure frequency. 5 The Engel class at last follow-up was recorded. Radiosurgery All radiosurgical cases were performed using the Leksell Gamma Knife with Gamma Plan software (Elekta AB). The AVM nidus was defined using MRI and CTA, and supplemented with DSA in select cases. The margin SRS dose was constructed to include the entire volume of the AVM. A dose of 17.5 Gy (range Gy) to the 50% (range 40% 70%) isodose line was the most commonly used treatment plan. A second radiosurgical treatment was performed in 23 patients (11.3%) of patients overall; 10 patients (12.8%) presented with seizure. Statistics All statistical analysis was carried out in Excel 2010 (Microsoft) and JMP Pro 12 (SAS). A t-test was used to compare continuous variables across groups. Pearson s chi square or Fisher s exact test were used for comparisons of categorical variables, depending on expected cell count. The significance level was set as 0.05, and a 2-sided probability testing was used. Results Of the 204 patients with AVMs treated with SRS, 78 (38.2%) presented with seizures. With regard to sex, the cohort was evenly divided (39 males and 39 females). Mean age at time of the first SRS treatment (± SD) was 38.5 ± 16.9 years (range years). Spetzler-Martin grades were available for 77 of 78 patients: 5 Grade I AVMs (6.5%), 18 Grade II (23.4%), 36 Grade III (46.8%), 17 Grade IV (22.1%), and 1 Grade V (1.3%). Mean duration of follow-up was 36.4 ± 37.6 months (range months). Nine (11.5%) of the 78 patients who presented with seizures also had radiographic evidence of acute hemorrhage at the time of presentation. Six patients (7.7%) experienced hemorrhage following SRS; 1 of these patients had a history of hemorrhage prior to treatment. Posttreatment hemorrhage occurred at a mean of 618 ± 646 days (range days) after SRS. Five (6.4%) of the 78 patients had previously undergone craniotomy for attempted AVM resection or clipping of an aneurysm associated with an AVM. No patients underwent endovascular embolization prior to SRS. Ten patients were retreated with SRS at a mean of 1153 ± 458 days (range days) following the initial SRS. Fifty-nine (75.6%) of the 78 patients had undergone imaging to determine response to SRS in the form of either DSA, MRI, or CTA (Fig. 1). Thirty-four (57.6%) of the 59 patients achieved complete obliteration of the malformation. Seventeen patients (28.8%) demonstrated partial obliteration of the AVM. Eight patients (13.6%) demonstrated no radiographic response to treatment. These 8 patients with stable lesions on imaging had a mean follow-up of 20.5 ± 20 months, with only 1 patient requiring ongoing antiepileptic drug (AED) therapy following SRS. Pretreatment medication lists were available for 54 (69.2%) of the 78 patients, of whom 49 had been treated with at least 1 AED. Of these 49 patients, 17 (34.7%) remained on AEDs at a mean follow-up of 29.1 ± 32 months (range months). Two of the 5 patients who were known not to be on an AED prior to treatment with SRS were receiving AED therapy at a mean follow-up of 55.5 ± 29.9 months (range months). At last clinical follow-up (mean 37.1 ± 38.3 months, range months), 63 (80.8%) of the 78 patients were classified as Class I on the Engel Epilepsy Surgery Outcome Scale, 8 (10.3%) as Class II, 3 (3.8%) as Class III, and 4 (5.1%) as Class IV. Multivariate analysis did not demonstrate any significant correlation between the likelihood of an Engel Class I at last follow-up and patient demographics, comorbidities, AVM characteristics, history of operative intervention, pre- or posttreatment hemorrhage, or radiographic degree of resolution (Tables 1 and 2). A total of 21 (26.9%) of the 78 patients remained on AEDs at a mean follow-up of 31.3 ± 30.7 months (range months). Multivariate analysis did not demonstrate any significant correlation between the likelihood of continued AED therapy and factors such as patient age, average follow-up time, pre/posttreatment hemorrhage, or radiographic degree of resolution (Table 3). Of the 126 patients who did not present with seizures prior to treatment with SRS, only 5 patients (4.0%) had seizures in the post-srs period, 1 of whom had a single seizure, and their AED was discontinued after 1 year. Excluding this patient, 4 patients (3.2%) had clinically significant seizures at last follow-up. One of these patients was classified as Engel Class II, 1 as Engel Class III, and 2 as Engel Class IV. None of these patients had concomitant hemorrhage at the time of presentation. Figure 2 demonstrates the likelihood of presenting with seizures and longterm seizure outcomes as a function of the presence or absence of intracerebral hemorrhage (ICH) at presentation. Adverse Events All adverse events potentially related to the AVM or radiosurgical procedure were included. Fifty-four complications were encountered in 45 patients (22.1%): headache in 10 patients, hemorrhage in 13, neurological deficit (i.e., 846 J Neurosurg Volume 126 March 2017

3 Seizure outcome following radiosurgery for AVMs FIG. 1. Flow chart showing the radiographic evaluation of response to SRS for AVMs in 78 patients who presented with seizures. hemiparesis, decreased sensation, visual field cut) without hemorrhage in 10 patients, symptomatic cerebral edema in 6, new-onset seizure in 5, radiation necrosis in 3, dysphasia in 2, ventricular shunt in 2, syncope in 1, memory loss in 1, and alopecia in 1. Ten patients (4.9%) eventually underwent craniotomy for AVM resection with or without hematoma evacuation. Discussion This study found a clear tendency toward seizure improvement in AVM patients following treatment with TABLE 1. Demographic characteristics and comorbidities by post-srs Engel class in 78 AVM patients who presented with seizures* Characteristic Engel I Engel II IV p Value Total 63 (80.8) 15 (19.2) Mean age in yrs 39.2 ± ± Sex 0.77 M 32 (50.8) 7 (46.7) F 31 (49.2) 8 (53.3) Race 0.46 White 44 (69.8) 13 (86.7) Black 18 (28.6) 2 (13.3) Hispanic 1 (1.6) 0 (0) Smoking 23 (36.5) 6 (40) 1.00 Alcohol 7 (11.1) 3 (20) 0.39 Hypertension 16 (25.4) 4 (26.7) 1.00 CAD 1 (1.6) 1 (6.7) 0.35 Diabetes 1 (1.6) 1 (6.7) 0.35 CAD = coronary artery disease. * Values are number of patients (%) unless otherwise indicated. Mean value is presented with SD. SRS, with 81% of patients presenting with seizure, attaining Class I on the Engel Epilepsy Surgery Outcome Scale. There is significant variation in the literature with regard to the exact rates of improvement, but the majority of studies have indicated good long-term seizure outcome in AVM patients who undergo SRS. Our findings are comparable to those of a Mayo Clinic study published in 2000, which found that 78% of the cohort of AVM patients with seizures had achieved an Engel Class of 1 at their 3 year follow-up post-srs. 15 A retrospective study published in 2002 that looked at seizure outcome in AVM patients following multimodality treatment found a slightly lower proportion of patients (66%) achieving an Engel class of 1. 8 The same study included a review of the available literature and found rates of seizure improvement/cessation following SRS ranging from 19% to 85% across 7 studies (5 of which demonstrated rates greater than 50%). 8 De novo seizure rates following SRS ranged from 0% to 6.5% across 6 studies, 8 a range with which our data are compatible. A recent systematic review of 24 studies and 1157 patients found the overall rate of seizure control to be 62.8% among AVM patients treated with SRS, with an even greater proportion achieving seizure control (85.2%) when complete AVM obliteration was achieved. 1 In our study, follow-up radiographic data were available for only 75% of the cohort of AVM patients who presented with seizures, but there was no statistically significant correlation between the degree of obliteration and posttreatment Engel class or continued AED usage. Despite the known relative variation in epileptogenic susceptibility between different brain regions, our data did not demonstrate a significant correlation between the likelihood of achieving an Engel class of I and the location of the AVM (p = 0.18). A recent case-control study of 175 temporal lobe AVMs treated with radiosurgery, with AVM characteristics and seizure histories comparable to our cohort of 16 temporal AVM patients, also found no statistically significant correlation between temporal loca- J Neurosurg Volume 126 March

4 B. J. Ditty et al. TABLE 2. AVM characteristics by post-srs Engel Class in 78 AVM patients who presented with seizures* Characteristic Engel I Engel II IV p Value Total 63 (80.8) 15 (19.2) SM grade 0.84 I 4 (6.5) 1 (6.7) II 13 (21.0) 5 (33.3) III 30 (48.4) 6 (40) IV 14 (22.6) 3 (20) V 1 (1.6) 0 (0) Location 0.18 Frontal 12 (19) 2 (13.3) Parietal 17 (27) 4 (26.7) Temporal 14 (22.2) 2 (13.3) Occipital 8 (12.7) 1 (6.7) Cerebellar 1 (1.6) 0 (0) Thalamic 1 (1.6) 0 (0) Basal ganglia 3 (4.8) 1 (6.7) Frontoparietal 2 (3.2) 2 (13.3) Parietooccipital 0 (0) 3 (20) Frontotemporal 2 (3.2) 0 (0) Quadrigeminal plate 1 (1.6) 0 (0) Atrium 1 (1.6) 0 (0) Perisylvian 1 (1.6) 0 (0) Side 0.39 Lt 33 (52.4) 10 (66.7) Rt 30 (47.6) 5 (33.3) Prior craniotomy 3 (4.8) 2 (13.3) 0.24 Pre-SRS bleed 9 (14.3) 0 (0) 0.19 Post-SRS bleed 5 (7.9) 1 (6.7) 1.00 Response to SRS based on imaging 0.51 Complete 28 (44.4) 6 (40) Partial 12 (19) 5 (33.3) Stable 6 (9.5) 2 (13.3) No imaging follow-up 17 (27.0) 2 (13.3) SM = Spetzler-Martin. * Values are number of patients (%). Grades were available for 77 of 78 patients. TABLE 3. AED use following SRS in 78 AVM patients who presented with seizures* Characteristic Continued AEDs No AEDs p Value Total 21 (26.9) 57 (73.1) Mean age in yrs 36.7 ± ± Mean follow-up time in 31.3 ± ± mos Pre-SRS bleed 2 (9.5) 7 (12.3) 1.00 Post-SRS bleed 3 (14.3) 3 (5.3) 0.35 Response to SRS based 0.69 on imaging Complete 9 (42.9) 25 (43.9) Partial 6 (28.6) 11 (19.3) Stable 1 (4.8) 7 (12.3) No imaging follow-up 5 (23.8) 14 (24.6) * Values are number of patients (%) unless otherwise indicated. Mean values are presented with SDs. tion or degree of obliteration and the degree of seizure control. 4 Pretreatment and posttreatment AVM hemorrhage was associated with neurological morbidity, but it was not correlated with the likelihood of achieving an Engel class of 1 or discontinuing AEDs. One patient with pretreatment hemorrhage had residual right upper extremity weakness and headaches, while a second suffered devastating hemorrhages before and after treatment, resulting in a modified Rankin Scale score of 5. Including this patient, 4 of the 6 patients who experienced posttreatment hemorrhage suffered significant morbidity including ventriculoperitoneal shunt requirement, brain resection, and visual field deficits. The aforementioned systematic review by Baranoski et al. did demonstrate a significantly greater rate of seizure control after radiosurgery in patients with unruptured AVMs (p < 0.03); however, seizure outcomes were compared between a cohort of 187 unruptured AVMs versus only 13 ruptured AVMs. 1 Despite the hemorrhage-related neurological morbidity, a majority of these patients experienced good seizure outcomes. It has been suggested that while the likelihood of posttreatment hemorrhage is greatest during the latency interval between radiosurgery and AVM obliteration, the risk may not differ significantly from an AVM s natural history, and thus risk of hemorrhage should not be used to justify a delay in radiosurgical treatment of small AVMs. 14 The epileptogenic nature of AVMs is well recognized but not fully understood. Theorized mechanisms include ischemia of adjoining brain tissue from a steal phenomenon, gliosis from hemosiderin leakage, and subclinical hemorrhage. 11,19 One study implicated abnormal electrophysiological properties of neurons surrounding an AVM and iron-induced free radical damage in seizure pathogenesis. 11 The study included a detailed discussion about how hemosiderin deposits in the brain tissue surrounding an AVM can inhibit the reuptake of glutamate as well as inhibit glutamine synthetase, thus allowing for glutamateinduced cytotoxiticty. 11 Another study described secondary epileptogenesis as a possibility in AVM patients with seizure foci distant from the site of a resected AVM. 23 These changes in surrounding brain tissue may account for failure of AVM therapy to achieve complete seizure control. Kraemer and Awad (1994) specifically discussed the role of perinidal ischemic gliosis and residual gliotic scars as sources for persistent seizures in AVM patients status post therapy. 11 Previous investigators have attempted to identify risk factors and predictors of seizures in patients with AVMs. A number of studies have shown statistically significant correlations based on angioarchitecture and location. 6 In 1995, Turjman et al. demonstrated that cortical location, superficial temporal location, superficial parietal location, cortical feeder location, external carotid feeder, and the 848 J Neurosurg Volume 126 March 2017

5 Seizure outcome following radiosurgery for AVMs FIG. 2. Schematic showing the likelihood of presenting with seizures and long-term seizure outcomes as a function of ICH at presentation for 204 AVM patients treated with SRS. presence of a middle cerebral artery feeder were independent predictors of seizure secondary to AVMs. 19 More recently, Shankar et al. found a statistically significant association between seizures and AVMs with characteristics of high flow (including pial recruitment and perinidal angiogenesis, fistulization, and intranidal aneurysm) and venous outflow obstruction (including venous ectasia, pial long draining vein, venous outflow restriction draining vein, and pseudophlebitic cortical vein pattern). 16 The mechanism by which SRS leads to a reduction in seizures has not been fully elucidated. Improved cerebral hemodynamics following radiosurgical obliteration may contribute to control of seizures and other adverse symptoms. 12 This is plausible in light of theories that attribute vascular steal phenomena to the epileptogenicity of AVMs. 22 Animal models studies have found that exposure to ionizing radiation produces alterations in synaptic transmission 17 and biochemical/structural changes that reduce susceptibility to seizures. 2 A Finnish study found that seizure reduction following radiosurgery for AVMs did not rely on the angiographic result; the authors suggested that seizure control may be attributable to the effects of ionizing radiation. 7 Persistent or de novo post-srs seizures may also be related to the radiotherapy itself. There appears to be a 2% 3% risk of permanent neurological deficits secondary to radiation injury in patients who undergo SRS. 10 These data are consistent with our finding of a 3% rate of de novo seizures that were persistent at last follow-up. In contrast, one study showed that as many as 36% of AVM patients treated with SRS experienced some type of immediate side effect (occurring during and lasting no more than the 2-week period following the procedure), which included but was not limited to new seizures. 21 Despite evidence of improvement in seizures following SRS for AVMs as demonstrated by our study and others, there remains the question of what the natural history of these seizures would be without therapy. The prospective Scottish Intracranial Vascular Malformation Study (SIVMS) showed no statistical difference in the incidence of de novo or recurrent seizures over a 5-year period between AVM patients who received therapy and those managed conservatively. 9 For patients who had seizures on presentation, the study found no statistical difference in the likelihood of 2-year seizure freedom whether an intervention was instituted or not. 9 This lack of statistical significance was present regardless of the presenting symptom, modality of treatment in the intervention group, or completeness of AVM obliteration. 9 There have been a number of studies comparing seizure outcomes in AVM patients based on the treatment modality. Recently, a retrospective study performed at The Johns Hopkins University School of Medicine suggested that seizure control is more effective with surgical resection than radiosurgical ablation of AVMs, but that de novo seizures are more likely following surgical resection than radiosurgery. 20 The latter assertion is also apparent in the literature review by Hoh et al., which showed newonset seizure rates following surgical excision of AVMs ranging from 8% 57% over 7 studies. 8 In contrast, the J Neurosurg Volume 126 March

6 B. J. Ditty et al. retrospective study conducted by the authors themselves looked at data from completely obliterated AVMs and found no statistically significant difference in seizure outcomes between surgery, embolization, and radiosurgery. 8 The previously mentioned SIVMS also demonstrated the lack of a statistically significant difference in seizure outcomes between the different treatment modalities. 9 It is likely that additional studies will be needed to further explore whether seizure outcomes in AVM patients depend on the treatment modality or are specifically attributable to treatment at all. Limitations This case series has several limitations. The study is retrospective and takes place at a single center. Retrospective data collection is subject to bias and relies heavily on the quality of the medical record. The single center experience limits the generalizability and reproducibility of the findings. The overall number of patients, while reasonable for a rare disease process, is small. This results in a potentially underpowered study and limits subgroup analysis. Follow up of these patients ranged widely and included 10 patients (12.8%) with follow-up of less than 6 months the traditional cutoff for seizure outcome. While the inclusion of these patients can falsely elevate seizure freedom rates, the effect was felt to be small and justified the inclusion of all patients presenting with this rare disease. Of the 10 patients with follow-up of less than 6 months, 9 were categorized as Engel Class I. If all patients with less than 6 months of follow-up were removed (leaving 68 patients presenting with seizure), the change in the frequency of Engel Class I outcomes would be modest (80.8% to 79.4%). Radiographic follow-up imaging was only available in 75% of patients, limiting the ability to evaluate the role of radiographic treatment response and seizure outcome. Moreover, some of the follow-up imaging consisted of CTA or MRI, which are less reliable imaging techniques than cerebral angiography for assessing the results of SRS in the treatment of AVMs. These limitations are common in the investigation of rare neurosurgical pathology and are best overcome by prospective multicenter collaboration. Conclusions For patients presenting with AVM-related seizures, SRS appears to reduce seizure frequency and allows for the discontinuation of AEDs. The majority of SRS-treated patients achieved Engel Class I seizure outcome, while de novo seizures following SRS were rare. Acknowledgments Dr. Ditty completed this work as a Women s Leadership Council Clinical Scholar in the Department of Neurosurgery at the University of Alabama at Birmingham. References 1. Baranoski JF, Grant RA, Hirsch LJ, Visintainer P, Gerrard JL, Günel M, et al: Seizure control for intracranial arteriovenous malformations is directly related to treatment modality: a meta-analysis. J Neurointerv Surg 6: , Chen ZF, Kamiryo T, Henson SL, Yamamoto H, Bertram EH, Schottler F, et al: Anticonvulsant effects of gamma surgery in a model of chronic spontaneous limbic epilepsy in rats. J Neurosurg 94: , Crawford PM, West CR, Shaw MD, Chadwick DW: Cerebral arteriovenous malformations and epilepsy: factors in the development of epilepsy. Epilepsia 27: , Ding D, Quigg M, Starke RM, Xu Z, Yen CP, Przybylowski CJ, et al: Radiosurgery for temporal lobe arteriovenous malformations: effect of temporal location on seizure outcomes. J Neurosurg 123: , Engel J (ed): Surgical Treatment of the Epilepsies, ed 2. New York: Raven Press, Galletti F, Costa C, Cupini LM, Eusebi P, Hamam M, Caputo N, et al: Brain arteriovenous malformations and seizures: an Italian study. J Neurol Neurosurg Psychiatr 85: , Heikkinen ER, Konnov B, Melnikov L, Yalynych N, Zubkov YN, Garmashov YA, et al: Relief of epilepsy by radiosurgery of cerebral arteriovenous malformations. Stereotact Funct Neurosurg 53: , 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 51: , Josephson CB, Bhattacharya JJ, Counsell CE, Papanastassiou V, Ritchie V, Roberts R, et al: Seizure risk with AVM treatment or conservative management: prospective, population-based study. Neurology 79: , Kano H, Lunsford LD, Flickinger JC, Yang HC, Flannery TJ, Awan NR, et al: Stereotactic radiosurgery for arteriovenous malformations, Part 1: management of Spetzler-Martin Grade I and II arteriovenous malformations. J Neurosurg 116:11 20, Kraemer DL, Awad IA: Vascular malformations and epilepsy: clinical considerations and basic mechanisms. Epilepsia 35 (Suppl 6):S30 S43, Lim YJ, Lee CY, Koh JS, Kim TS, Kim GK, Rhee BA: Seizure control of Gamma Knife radiosurgery for nonhemorrhagic arteriovenous malformations. Acta Neurochir Suppl 99:97 101, Murphy MJ: Long-term follow-up of seizures associated with cerebral arteriovenous malformations. Results of therapy. Arch Neurol 42: , Pollock BE, Lunsford LD, Kondziolka D, Maitz A, Flickinger JC: Patient outcomes after stereotactic radiosurgery for operable arteriovenous malformations. Neurosurgery 35:1 8, Schäuble B, Cascino GD, Pollock BE, Gorman DA, Weigand S, Cohen-Gadol AA, et al: Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology 63: , Shankar JJS, Menezes RJ, Pohlmann-Eden B, Wallace C, terbrugge K, Krings T: Angioarchitecture of brain AVM determines the presentation with seizures: proposed scoring system. AJNR Am J Neuroradiol 34: , Tolliver JM, Pellmar TC: Ionizing radiation alters neuronal excitability in hippocampal slices of the guinea pig. Radiat Res 112: , Trinka E, Bauer G, Oberaigner W, Ndayisaba JP, Seppi K, Granbichler CA: Cause-specific mortality among patients with epilepsy: results from a 30-year cohort study. Epilepsia 54: , Turjman F, Massoud TF, Sayre JW, Viñuela F, Guglielmi G, Duckwiler G: Epilepsy associated with cerebral arteriovenous malformations: a multivariate analysis of angioarchitectural characteristics. AJNR Am J Neuroradiol 16: , J Neurosurg Volume 126 March 2017

7 Seizure outcome following radiosurgery for AVMs 20. Wang JY, Yang W, Ye X, Rigamonti D, Coon AL, Tamargo RJ, et al: Impact on seizure control of surgical resection or radiosurgery for cerebral arteriovenous malformations. Neurosurgery 73: , Werner-Wasik M, Rudoler S, Preston PE, Hauck WW, Downes BM, Leeper D, et al: Immediate side effects of stereotactic radiotherapy and radiosurgery. Int J Radiat Oncol Biol Phys 43: , Yeh HS, Kashiwagi S, Tew JM Jr, Berger TS: Surgical management of epilepsy associated with cerebral arteriovenous malformations. J Neurosurg 72: , Yeh HS, Privitera MD: Secondary epileptogenesis in cerebral arteriovenous malformations. Arch Neurol 48: , 1991 Disclosures The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author Contributions Conception and design: Ditty, Miller, Harrigan. Acquisition of data: Omar, Miller. Analysis and interpretation of data: Ditty, Miller, Kicielinski, Fisher, Harrigan. Drafting the article: Ditty, Omar, Miller. Critically revising the article: Foreman, Ditty, Omar, Kicielinski, Fisher, Harrigan. Reviewed submitted version of manuscript: Foreman, Omar, Kicielinski, Fisher, Harrigan. Approved the final version of the manuscript on behalf of all authors: Foreman. Statistical analysis: Omar, Miller, Kicielinski. Study supervision: Fisher, Harrigan. Correspondence Paul M. Foreman, Department of Neurosurgery, University of Alabama at Birmingham, Faculty Office Tower 1005, th St. S, Birmingham, AL pforeman@uabmc.edu. J Neurosurg Volume 126 March