Original Article
Cerebral Arteriovenous Malformations and Epilepsy, Part 1: Predictors of Seizure Presentation Dale Ding1, Robert M. Starke1, Mark Quigg2, Chun-Po Yen1, Colin J. Przybylowski3, Blair K. Dodson1, Jason P. Sheehan1
OBJECTIVE: Seizures are relatively common in patients harboring cerebral arteriovenous malformations (AVMs). Because the pathogenesis of AVM-associated epilepsy is not well-defined, we aim to determine the factors associated with seizure presentation in AVM patients.
-
METHODS: We evaluated our institutional AVM radiosurgery database, from 1989e2013, to select patients in whom pertinent clinical information at presentation and adequate clinical and radiologic follow-up was available. Baseline patient demographics and AVM angioarchitectural features were compared between patients with and without seizure presentation. In addition to standard descriptive statistics, logistic regression analyses were performed to identify predictors of seizure presentation.
-
RESULTS: Of the 1007 AVM patients included for analysis, 229 patients presented with seizures (22.7%). The incidence of seizure presentation was significantly higher in cortical than noncortical AVMs (33.1% vs. 6.6%, P < 0.0001). Among the cortical locations, occipital AVMs had the lowest rate of seizure presentation (21.5%, P [ 0.0012), whereas the rates of seizure presentation in frontal (37.3%), temporal (37.7%), and parietal (34.0%) AVMs were similar. The lack of prior AVM hemorrhage (P < 0.0001), larger nidus diameter (P < 0.0001), and cortical location (P < 0.0001) were independent predictors of seizure presentation in the multivariate analysis. The
-
Key words Epilepsy - Intracranial arteriovenous malformations - Neurologic manifestations - Seizures - Vascular malformations -
Abbreviations and Acronyms AED: Antiepileptic drug ARUBA: A Randomized Trial of Unruptured Brain AVMs AVM: Arteriovenous malformation CCM: Cerebral cavernous malformation CI: Confidence interval CT: Computed tomography DSA: Digital subtraction angiography IRB: Institutional review board MRI: Magnetic resonance imaging
WORLD NEUROSURGERY - [-]: ---, - 2015
strongest independent predictors of seizure presentation were lack of prior AVM hemorrhage (OR 16.8) and cortical location (OR 4.2). CONCLUSIONS: Large, unruptured, cortical nidi are most prone to seizure presentation in patients referred for radiosurgery. Further investigations of the molecular biology, neuronal and glial physiology, and natural history of AVMassociated epilepsy appear warranted.
-
INTRODUCTION
E
pilepsy is a common symptom of cerebral arteriovenous malformations (AVMs), but its importance is often overlooked in favor of intracranial hemorrhage, which is the most frequent and feared component of an AVM’s natural history (4, 15, 20, 35). However, AVM-associated epilepsy may also be debilitating and adversely impact a patient’s quality of life (24). Seizures are the most common presentation of unruptured AVMs and are the second most frequent presentation for all AVM patients following hemorrhage (19, 34, 45). Prior studies have linked various factors to AVM-associated epilepsy, including male gender, younger age, larger AVM size, frontal or temporal nidus location, superficial AVM topography, AVM location at an arterial borderzone, absence of intranidal aneurysms, and presence of a venous varix (18, 19, 22, 47).
OR: Odds ratio RBAS: Radiosurgery-based AVM score SD: Standard deviation VRAS: Virginia Radiosurgery AVM Scale Departments of 1Neurological Surgery and 2Neurology and 3School of Medicine, University of Virginia, Charlottesville, VA, USA To whom correspondence should be addressed: Jason P. Sheehan, M.D., Ph.D. [E-mail: [email protected]] Citation: World Neurosurg. (2015). http://dx.doi.org/10.1016/j.wneu.2015.02.039 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
1
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
However, these previous studies were performed in relatively small or medium-sized cohorts. In this first of a two-part study, we relate baseline patient and AVM characteristics to seizures at presentation using a large patient sample. The aim of the present case-control study is to define the predictors of seizure presentation in AVMs before stereotactic radiosurgery. METHODS Patient Selection Criteria and Cohort Composition We reviewed a prospectively collected database approved by our institutional review board (IRB), of approximately 1400 patients harboring AVMs who were treated with Gamma Knife radiosurgery at the University of Virginia from 1989e2013. In order to maintain the consistency of the patient cohort throughout both parts of this study, the inclusion criteria were the same for Parts 1 and 2 of the study, specifically as follows: patients with (1) sufficient history at presentation to determine the presence or absence of seizures at the time of presentation, (2) similar information available at the time of follow-up, and (3) at least 2 years of postradiosurgery follow-up or radiologic evidence of AVM obliteration. Patients treated with dose- or volume-staged radiosurgical approaches were excluded. The case cohort comprised patients with seizures at or around the time of radiosurgery. The control cohort was composed of patients without seizures at or around the time of radiosurgery. Patient Demographics and AVM Angioarchitectural Features The baseline patient variables were gender; age; prior surgical resection or embolization; prior AVM hemorrhage; clinical presentation (hemorrhage, seizure, focal neurological deficit, headache, incidental diagnosis, or other symptoms); and seizure status at presentation (presence or absence of seizures at presentation). Patients with seizures at presentation for radiosurgery were classified as having seizures of any type, regardless of preexisting antiepileptic drug (AED) therapy, prior AVM treatments, or prior rupture. The angioarchitecture of the AVM nidus was defined by a combination of magnetic resonance imaging (MRI) and catheter digital subtraction angiography. Before 1991, angiography alone was used for diagnosis in some patients, whereas afterward, both angiography and MRI were used for nidus definition. The baseline AVM variables were location (cortical vs. noncortical, eloquent vs. noneloquent); size (maximal diameter, volume); venous drainage pattern (single vs. multiple draining veins, superficial vs. deep venous drainage); and presence of associated aneurysms (intranidal or perinidal). Cortical AVM location was defined as location of the center of the nidus in the frontal, temporal, parietal, or occipital lobes. Eloquent location was defined as sensorimotor, language, and visual cortex; hypothalamus and thalamus; internal capsule; brainstem; cerebellar peduncles; and deep cerebellar nuclei (40). On the basis of the patient and AVM characteristics, the Spetzler-Martin grade, modified radiosurgery-based AVM score (RBAS), and Virginia Radiosurgery AVM Scale (VRAS) were determined (40, 44, 49). Statistical Analysis Statistical analysis was performed with the Stata 10.0 software program (College Station, Texas, USA). Data are presented as
2
www.SCIENCEDIRECT.com
mean and standard deviation (SD) for continuous variables and as frequency for categorical variables. The incidence of seizure presentation was calculated separately for each individual location. Categorical variables were compared using the chi-square test, and continuous variables were compared with the independent (two sample), unpaired Student’s t-tests. A P-value of less than 0.05 was considered statistically significant. Univariate logistic regression analysis was performed to identify factors associated with seizure presentation. The covariates included in the logistic regression analysis were the patient and AVM factors listed earlier. The odds ratio (OR), 95% confidence interval (CI), and P-value were reported for each covariate. Interaction and confounding were assessed through stratification and relevant expansion covariates. Factors with a P-value of less than 0.20 in the univariate analysis were entered into a multivariate logistic regression analysis to determine independent predictors of seizure presentation. RESULTS Patient Demographics and Clinical Characteristics Table 1 shows the comparisons of patients who presented with seizures against the control cohort. Of the 1007 patients with sufficient clinical data at or around the time of radiosurgery, seizure was the presenting diagnostic incident in 229 patients, for a seizure presentation rate of 22.7%. The mean interval between the onset of seizures and radiosurgery was 58.6 82.7 months. The proportion of patients who presented with seizures was significantly lower for prior surgical resection (5.2% vs. 13.8%, P < 0.0001) and prior AVM hemorrhages (9.2%, vs. 69.5%, P < 0.0001) and higher for prior embolization (31% vs. 22%, P ¼ 0.005). AVM Angioarchitecture Table 2 details the AVM angioarchitectural comparisons of the patient cohorts with and without seizure presentation. Patients with seizure presentation harbored significantly larger AVMs on the basis of maximum diameter (mean 2.7 vs. 2.2 cm, P < 0.0001)
Table 1. Comparison of Demographics and Clinical Characteristics of Patients with and without Seizure Presentation AVMs in Patients with Seizure Presentation (n [ 229)
AVMs in Patients without Seizure Presentation (n [ 778)
P value
Female gender
106 (46.3%)
390 (50.1%)
0.307
Age (mean SD years)
35.2 14.4
33.6 16.2
0.186
Prior surgical resection
12 (5.2%)
107 (13.8%)
< 0.0001*
Factor
Prior embolization
71 (31.0%)
171 (22.0%)
Prior AVM hemorrhage
21 (9.2%)
541 (69.5%)
0.005* < 0.0001*
SD, standard deviation. *Statistically significant (P < 0.05).
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
Table 2. Comparison of AVM Angioarchitectural Features in Patients with and without Seizure Presentation AVMs in Patients with Seizure Presentation (n [ 229)
Factor
AVMs in Patients without Seizure Presentation (n [ 778)
P value
Diameter (mean SD cm)
2.7 0.9
2.2 0.9
< 0.0001*
Volume (mean SD mL)
4.8 3.8
3.1 3.1
< 0.0001*
Cortical location
203 (88.7%)
411 (52.8%)
< 0.0001*
Eloquent location
146 (63.8%)
529 (68.0%)
0.230
10 (4.4%)
42 (5.4%)
0.541
Associated aneurysms
y
102 (44.5%)
472 (60.7%)
< 0.0001*
Deep venous drainage
84 (36.7%)
471 (60.5%)
< 0.0001*
Spetzler-Martin grade
I: 36 (15.7%), II: 90 (39.3%), III: 80 (34.9%), IV: 22 (9.6%), V: 1 (0.4%)
Single draining vein
RBAS (mean SD) VRAS
I: 111 (14.3%), II: 264 (33.9%), III: 318 (40.9%), IV: 85 (10.9%), V: 0
1.22 0.46
1.15 0.46
0-1: 64 (27.9%), 2: 81 (35.4%), 3: 78 (34.1%), 4: 6 (2.6%)
0.156 0.034*
0-1: 175 (22.5%), 2: 309 (39.7%), 3: 198 (25.5%), 4: 96 (12.3%)
< 0.0001*
SD, standard deviation; RBAS, radiosurgery-based AVM score; VRAS, Virginia Radiosurgery AVM Scale. *Statistically significant (P < 0.05). yAssociated aneurysms ¼ intranidal or perinidal aneurysms.
and volume (mean 4.8 vs. 3.1 mL, P < 0.0001). A significantly smaller proportion of AVMs in patients with seizure presentation had a single draining vein (44.5% vs. 60.7% vs. P < 0.0001) and deep venous drainage (36.7% vs. 60.5%, P < 0.0001). Patients with seizure presentation had significantly higher RBAS (mean 1.22 vs. 1.15, P ¼ 0.034) and lower VRAS (P < 0.0001).
Predictors of Seizure Presentation Of the 229 patients who presented with seizures, 203 AVMs were in cortical locations (88.6%) and 26 AVMs were in noncortical locations (11.4%). Of the 778 patients without seizures at presentation, 411 AVMs were in cortical locations (52.8%) and 367 AVMs were in noncortical locations (47.2%). The incidence of seizure presentation was 33.1% for cortical AVMs (203/614 patients) compared with 6.6% for noncortical AVMs (26/393 patients). Seizure presentation was significantly more common in patients with cortically based AVMs (P < 0.0001). Table 3 details the incidences of seizure presentation for cortical AVMs, stratified by location. The incidence of seizure presentations was significantly lower for occipital nidi (21.5%, P ¼ 0.0012) and were similar among the remaining cortical loci. Table 4 details the incidences of seizure presentation for noncortical AVMs, stratified by location. The incidences of seizure presentation were significantly higher for insular nidi (32%, P < 0.0001) and lower for cerebellar nidi (0, P ¼ 0.017), and they were similar among the remaining noncortical loci. Figure 1 shows the incidences of seizure presentation based on AVM location. Table 5 details the results of the univariate and multivariate logistic regression analyses for predictors of seizure presentation. Univariate logistic regression analysis determined
WORLD NEUROSURGERY - [-]: ---, - 2015
that lack of prior surgical resection (P ¼ 0.001), presence of prior embolization (P ¼ 0.005), no prior AVM hemorrhage (P < 0.0001), larger nidus maximal diameter (P < 0.0001), larger nidus volume (P < 0.0001), cortical nidus location (P < 0.0001), superficial venous drainage (P < 0.0001), multiple draining veins (P < 0.0001), higher RBAS (P ¼ 0.035), and lower VRAS (P ¼ 0.006) were significantly associated. Multivariate logistic regression analysis determined lack of prior AVM hemorrhage (P < 0.0001), larger nidus maximal diameter (P < 0.0001), and cortical location (P < 0.0001) were independent predictors of seizure presentation. The strongest independent predictor of seizure presentation was lack of prior hemorrhage (OR 16.8, 95% CI 10.3e27.2), followed by cortical location (OR 4.2, 95% CI 2.6e6.8).
Table 3. Incidences of Seizure Presentation by Location for Cortical AVMs, Listed in Descending Order
Cortical Location
Incidence of Seizure Presentation at Specified Cortical Location (%y)
Incidence of Seizure Presentation not at Specified Cortical Location (%y)
Temporal
37.7 (66/175)
31.2 (137/439)
0.122
Frontal
37.3 (53/142)
31.8 (150/472)
0.218
Parietal
34.0 (55/162)
32.7 (148/452)
0.779
Occipital
21.5 (29/135)
36.3 (174/479)
0.0012*
P value
*Statistically significant (P < 0.05). yReported as % (A/B), where A ¼ number of patients with seizure presentation and B ¼ total number of patients.
www.WORLDNEUROSURGERY.org
3
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
Table 4. Incidences of Seizure Presentation by Location for Noncortical AVM, Listed in Descending Order Incidence of Seizure Presentation at Specified Noncortical Location (%*)
Incidence of Seizure Presentation not at Specified Noncortical Location (%*)
P value
32.0 (8/25)
4.9 (18/368)
< 0.0001y
Basal ganglia
9.6 (8/83)
5.8 (18/310)
0.212
Corpus callosum
7.9 (3/38)
6.5 (23/355)
0.739
Brainstem
4.8 (4/84)
7.1 (22/309)
0.441
Thalamus
3.1 (3/96)
7.7 (23/297)
0.113
0 (0/67)
8.0 (26/326)
0.017*
Noncortical Location Insula
Cerebellum
*Reported as % (A/B), where A ¼ number of patients with seizure presentation and B ¼ total number of patients. yStatistically significant (P < 0.05).
DISCUSSION In light of the recent prospective studies, A Randomized Trial of Unruptured Brain AVMs (ARUBA) and the Scottish Audit of Intracranial Vascular Malformations, which showed superior outcomes with medical therapy compared with intervention, the management of unruptured AVMs has come under intense scrutiny (2, 30, 43). Given that seizures are the most common presentation of unruptured AVMs, it behooves us to improve our understanding of AVM-associated epilepsy, including preferential locations and predictors of its occurrence. Pathogenesis of AVM-Associated Epilepsy The pathogenesis of epilepsy secondary to AVMs has been previously explored, although our current understanding of this phenomenon remains relatively limited. Kraemer and Awad proposed a number of potential mechanisms for epileptogenesis in patients harboring vascular malformations, including neuronal
Figure 1. Bar plot showing the frequency of AVM patients with seizure presentation, stratified by nidus location.
4
www.SCIENCEDIRECT.com
degeneration and aberrant signaling, alterations in glial structure and physiology, and generation of reactive oxygen species (26). Williamson et al. showed that neurons surrounding cerebral cavernous malformations (CCMs) were more predisposed to hyperexcitability than neurons surrounding gliomas (50). Specifically, CCM-adjacent neurons were more likely to have high-amplitude, complex spontaneous synaptic events, whereas tumor-adjacent neurons were more likely to have low-amplitude, spontaneous excitatory activity. The authors postulated that hemosiderin deposition altered the physiologic environment surrounding CCMs by abrogating glutamate uptake and causing injury-induced synaptic reorganization, thereby promoting neuronal hypersynchronization and propagating perilesional excitatory stimuli to other cortical regions (50). In an immunohistochemical analysis of 30 vascular malformations and intra-axial neoplasms, Wolf et al. showed that expression of both excitatory and inhibitory neurotransmitter receptors and enzymes was increased in some specimens and decreased in others (51). The authors proposed that their findings suggested an imbalance of excitatory and inhibitory signaling in the perilesional cortex contributed to the pathogenesis of epilepsy in these patients. One should note that many of these putative epileptogenic mechanisms are extrapolated from observations or studies of CCMs, without equivalent electrophysiologic experiments in AVMs. However, because the molecular biology and clinical course of CCMs and AVMs are fundamentally distinct, the findings may not be generalizable and should therefore be interpreted with caution. This further highlights the limits of our knowledge of AVM-associated epileptogenesis. Predictors of Seizure Presentation in AVM Patients Most AVM studies are focused on determining and abrogating an AVM’s hemorrhage risk (4, 5, 8-12, 14, 21, 33, 42, 53, 54). However, seizures are relatively common in AVM patients and may significantly alter the medical and interventional management of AVMs. Al-Shahi and Warlow reported a 1% annual risk of de novo seizures in patients with untreated AVMs (1). Because the natural history of AVMs is known to differ according to patient- and AVM-related features, however, an AVM’s associated risk of seizures may vary as well. Due to the negative effect of seizures on patient neurological outcomes and quality of life, accurately predicting the occurrence of AVM-associated epilepsy can potentially impact the management of AVM patients. The strongest independent predictor of seizure presentation in our series was lack of prior hemorrhage (OR 16.8, P < 0.0001). Prior studies regarding predictors of AVM-associated epilepsy either excluded hemorrhage as a covariate in the statistical analysis or only included patients with unruptured AVMs (18, 19). Although intracranial hemorrhages can result in seizures, they are more likely to manifest as focal neurological deficits or global clinical deterioration (32). And because the impetus for intervention is much higher with ruptured AVMs than unruptured ones, many of these lesions may be embolized, extirpated, or irradiated before they can cause seizures. Additionally, it is possible that the gliotic tissue surrounding a ruptured nidus is less prone to epileptogenesis than the brain parenchyma adjacent to unruptured AVMs. In the current series of 1007 patients, the incidence of seizure presentation near the time of radiosurgery was 23%. Galletti et al.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
Table 5. Univariate and Multivariate Logistic Regression Analyses for Predictors of Seizure Presentation; Only Covariates with P < 0.20 in the Univariate Analysis Were Listed Univariate Factor
Multivariate
Odds Ratio
95% CI
P value
Odds Ratio
95% CI
P value
Increasing age
1.006
0.997e1.016
0.186
NS
NS
NS
No prior surgical resection
2.884
1.557e5.339
0.001*
NS
NS
NS
Prior embolization
1.592
1.148e2.209
0.005*
NS
NS
NS
22.610
14.074e36.321
< 0.0001*
16.775
10.349e27.193
< 0.0001*
Larger diameter
1.056
1.039e1.074
< 0.0001*
1.038
1.017e1.059
< 0.0001*
Larger volume
1.144
1.093e1.198
< 0.0001*
NS
NS
NS
No prior AVM hemorrhage
Cortical location
6.972
< 0.0001*
4.223
2.631e6.778
< 0.0001*
Superficial venous drainage
2.648
1.953e3.592
< 0.0001*
NS
NS
NS
Multiple draining veins
1.423
1.224e1.654
< 0.0001*
NS
NS
NS
Lower Spetzler-Martin grade
1.122
Higher RBAS
1.392
Lower VRAS
1.233
4.527-10.737
0.947e1.329 1.024-1.892 1.062e1.431
0.184
NS
NS
NS
0.035*
NS
NS
NS
0.006*
NS
NS
NS
NS, not significant (P 0.05); RBAS, radiosurgery-based AVM score; VRAS, Virginia Radiosurgery AVM Scale. *Statistically significant (P < 0.05).
examined a cohort of 101 AVM patients with a 31% incidence of seizure presentation (18). Multivariate logistic regression analysis found frontal and temporal lobe locations and superficial topography to be independently associated with seizure presentation. Turjman et al. performed a detailed radiographic analysis of 100 AVM patients with a 47% incidence of seizure presentation (47). Nidus supply by the middle cerebral artery, the external carotid artery, or a cortical feeding artery, as well as cortical, temporal superficial, and parietal superficial nidus locations, were identified as predictors of AVM-associated epilepsy. Our finding of cortical location as an independent predictor of seizure presentation (P < 0.0001) is in general agreement with the literature, although we did not find a specific location with a particularly high incidence of seizure presentation (18, 19, 22, 47). The rates of seizure presentation were over 30% for frontal, temporal, and parietal locations but were relatively lower in occipital nidi at 22% (P ¼ 0.0012). Also, in agreement with previous studies (22), greater AVM diameter was an independent predictor of seizure presentation (P < 0.0001). Large AVMs may have greater association with epilepsy than small AVMs due to their increased vascular steal, venous congestion, induction of hypoxia-mediated signaling pathways, and propensity to cause pathogenic alterations in neuronal and glial components of the perinidal cortex (17, 25). Chronic vascular steal by AVMs may cause hypoperfusion-induced remodeling of the perinidal cortex, thus facilitating an epileptogenic milieu (25). Conversely, Fiestra et al. showed, using a combination blood oxygen level-dependent MRI and angiography, that impaired cerebrovascular reactivity and venous congestion, rather than arterial steal, correlated significantly with AVM-associated epilepsy (17).
WORLD NEUROSURGERY - [-]: ---, - 2015
In an analysis of 542 AVM patients, Stapf et al. reported a 29% incidence of seizure presentation (41). Patients in the age ranges of 20 to 29 and 30 to 39 years had the highest rates of seizures presentation, compared with no patients with seizure presentation older than 60 years old. Hoh et al. studied 424 AVM patients with a 33% incidence of seizure presentation (22). Male gender, age younger than 65 years, AVM diameter greater than 3 cm, and temporal lobe location were significantly associated with seizure presentation. Garcin et al. analyzed a series of 155 AVM patients with a 29% incidence of seizure presentation (19). Multivariate logistic regression analysis identified male gender, frontal lobe location, and arterial borderzone location to be independent predictors of seizure presentation. In contrast to prior studies that reported an inverse correlation between age and frequency of seizure presentation (22, 41), the mean age of the patients in the present study with seizure presentation (35.2 years) was slightly higher than that of the patients without seizure presentation (33.6 years), although the difference was not significant (P ¼ 0.186). Male gender was more common in patients with seizure presentation (53.7% vs. 49.9%), unlike previous studies (19, 22), but the difference was not significant (P ¼ 0.307). Effect of AVM Intervention on Seizure Outcomes Although the primary goal of AVM intervention is obliteration of the nidus, in order to eliminate the future risk of hemorrhage, the impact of treatment on seizure outcomes is also an important consideration in the overall management of patients with AVMassociated epilepsy. Baranoski et al. performed a meta-analysis of AVM treatment-specific seizure outcomes and reported seizure control rates for 78%, 63%, and 49% for surgery, radiosurgery, and embolization (3). However, patients with obliterated AVMs after
www.WORLDNEUROSURGERY.org
5
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
radiosurgery had the highest rate of seizure control (85%). In a comparison of surgery to radiosurgery for AVM-associated epilepsy, Wang et al. found divergent seizure outcomes, depending on the baseline seizure status (48). Specifically, patients with seizures at presentation had significantly better seizure control after undergoing surgical AVM resection, whereas patients without seizures at presentation had significantly lower rates of de novo seizure after treatment with radiosurgery. Thus in addition to the angioarchitecture of an AVM nidus, the patient’s seizure status at presentation may influence the preferred treatment modality, if intervention is undertaken. A systematic review by Chen et al. reported seizure control and remission rates of 69% and 44%, respectively, after radiosurgery for patients with AVM-associated epilepsy (6). Seizure remission rates were found to be significantly higher in patients with complete AVM obliteration (82% vs. 41%, P ¼ 0.0007). Hoh et al. evaluated the seizure outcomes of 110 AVM patients who underwent multimodality AVM intervention (22). After a mean follow-up of 2.9 years, Engel class I outcome was achieved in 66% of patients. A short history of seizures, intracranial hemorrhage-related seizures, generalized seizures, deep and posterior fossa nidus locations, treatment with surgical resection, and complete AVM obliteration were significantly associated with Engel class I outcome. The seizure outcomes were similar among surgery, radiosurgery, and embolization when only obliterated nidi were considered (22). Thus AVM intervention appears to afford favorable seizure outcomes in the majority of patients with AVM-associated epilepsy. Although in some studies seizure remission appears to correlate with complete nidus obliteration, the association between the two outcomes is inconsistently reported in the literature (27, 39, 52). Study Limitations A limitation of our case-control study design is that all of the patients were referred for and underwent treatment with radiosurgery, so referral patterns and institutional treatment preferences may have biased the patient population of this study (6, 13, 31). Therefore patients selected for surgical resection, embolization without radiosurgery, and conservative management were not included in our analysis. As such, the clinical presentation and nidus characteristics of the patients in this study may not reflect those with significant hemorrhage from AVM rupture, surgically accessible AVMs, or large, unruptured AVMs in which the benefits of intervention were judged to outweigh the risks. Although the insula is a developmentally cortical structure, its location deep to the cortical surface and in close proximity to critical neural structures, such as the basal ganglia and internal capsule, distinguishes it from the lobar locations, as supported by prior studies of AVM-associated epilepsy, which have not classified the insula as a cortical structure (19, 47). Additionally, the insula is cytoarchitecturally and functionally distinct from the
REFERENCES 1. Al-Shahi R, Warlow C: A systematic review of the frequency and prognosis of arteriovenous malformations of the brain in adults. Brain 124: 1900-1926, 2001.
6
www.SCIENCEDIRECT.com
frontal, temporal, parietal, and occipital lobes. Finally, the insula is not a frequently identified origin of seizures, thus further distinguishing it from seizure-prone cortical sites, such as the temporal lobe (23, 38). If the insula has been classified as a cortical structure in the current study, the rate of seizure presentation for the cortical locations would be 33.0% (211/639 patients) and the rate of rate of seizure presentation in the non-cortical locations would be 4.9% (18/368 patients). These values were not significantly different from those reported previously in this study when the insula was categorized as a noncortical structure with respect to the rates of seizure presentation for either cortical (P ¼ 0.988) or noncortical (P ¼ 0.308) sites. Thus we do not believe that reclassifying the insular AVMs, which comprised only 2.5% (25/1007 patients) of the nidi in this study, as cortical AVMs would likely alter the major findings of our analysis. Because of the relatively long duration of the database (over 20 years), changes in neuroimaging modalities, interventional technology, and treatment paradigms over time may have affected the referral patterns to our center, thereby resulting in heterogeneous patient and AVM characteristics over time. Additionally, the disparities between the AVMs typically referred for microsurgery compared with radiosurgery may bias the angioarchitectural features of the nidi in our series (7, 16, 28, 29, 36, 46). Due to the nature of being a tertiary referral center for AVM radiosurgery, detailed clinical data, such as the type and severity of seizures, electroencephalography (EEG) findings, and prior AED therapy, were unavailable for some patients. Although the AVM patients with seizures were typically evaluated by an epileptologist, EEG confirmation of seizure activity was not required in the absence of generalized seizures. In a detailed analysis of seizure and anticonvulsant outcomes after AVM radiosurgery, we were able to obtain the data for seizure subtype and AED therapy for 73 of 229 AVM patients (32%) with seizures at presentation (37). We believe that the strength of the current study lies in the large number of patients, which provides higher statistical power to our analysis compared with prior studies. However, we acknowledge that the lack of complete details regarding seizure type and severity in a majority of patients represents a significant but common limitation of retrospective studies such as this one. CONCLUSIONS A combination of patient and nidus characteristics predicts seizure presentation. Unruptured, large, and cortically based AVMs are particularly prone to seizure presentation, and patients harboring such nidi likely warrant more rigorous neurological monitoring for signs of epilepsy. Future prospective studies are necessary to determine the natural history of seizures in unruptured AVM patients and the merits of long-term AED management in patients who are particularly prone to AVM-associated epilepsy.
2. Al-Shahi Salman R, White PM, Counsell CE, du Plessis J, van Beijnum J, Josephson CB, Wilkinson T, Wedderburn CJ, Chandy Z, St George EJ, Sellar RJ, Warlow CP: Outcome after conservative management or intervention for unruptured brain arteriovenous malformations. JAMA 311:1661-1669, 2014.
3. Baranoski JF, Grant RA, Hirsch LJ, Visintainer P, Gerrard JL, Gunel M, Matouk CC, Spencer DD, Bulsara KR: Seizure control for intracranial arteriovenous malformations is directly related to treatment modality: a meta-analysis. J Neurointerv Surg 6:684-690, 2013.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
ORIGINAL ARTICLE DALE DING ET AL.
4. Brown RD Jr, Wiebers DO, Forbes G, O’Fallon WM, Piepgras DG, Marsh WR, Maciunas RJ: The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 68:352-357, 1988. 5. Buell TJ, Ding D, Starke RM, Webster Crowley R, Liu KC: Embolization-induced angiogenesis in cerebral arteriovenous malformations. J Clin Neurosci 21:1866-1871, 2014. 6. Chen CJ, Chivukula S, Ding D, Starke RM, Lee CC, Yen CP, Xu Z, Sheehan JP: Seizure outcomes following radiosurgery for cerebral arteriovenous malformations. Neurosurg Focus 37:E17, 2014. 7. Ding D, Liu KC: Predictive capability of the Spetzler-Martin versus supplementary grading scale for microsurgical outcomes of cerebellar arteriovenous malformations. J Cerebrovasc Endovasc Neurosurg 15:307-310, 2013.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
18. Galletti F, Costa C, Cupini LM, Eusebi P, Hamam M, Caputo N, Siliquini S, Conti C, Moschini E, Lunardi P, Carletti S, Calabresi P: Brain arteriovenous malformations and seizures: an Italian study. J Neurol Neurosurg Psychiatry 85: 284-288, 2013. 19. Garcin B, Houdart E, Porcher R, Manchon E, Saint-Maurice JP, Bresson D, Stapf C: Epileptic seizures at initial presentation in patients with brain arteriovenous malformation. Neurology 78: 626-631, 2012. 20. Graf CJ, Perret GE, Torner JC: Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg 58:331-337, 1983. 21. Gross BA, Du R: Natural history of cerebral arteriovenous malformations: a meta-analysis. J Neurosurg 118:437-443, 2013.
unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. Lancet 383:614-621, 2014. 31. Moosa S, Chen CJ, Ding D, Lee CC, Chivukula S, Starke RM, Yen CP, Xu Z, Sheehan JP: Volumestaged versus dose-staged radiosurgery outcomes for large intracranial arteriovenous malformations. Neurosurg Focus 37:E18, 2014. 32. Morgenstern LB, Hemphill JC 3rd, Anderson C, Becker K, Broderick JP, Connolly ES Jr, Greenberg SM, Huang JN, MacDonald RL, Messe SR, Mitchell PH, Selim M, Tamargo RJ: Guidelines for the management of spontaneous intracerebral hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 41:2108-2129, 2010.
8. Ding D, Starke RM, Yen CP, Sheehan JP: Radiosurgery for cerebellar arteriovenous malformations: does infratentorial location affect outcome? World Neurosurg 82:e209-217, 2014.
22. 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: 303-309; discussion 309-311, 2002.
33. Mouchtouris N, Jabbour PM, Starke RM, Hasan DM, Zanaty M, Theofanis T, Ding D, Tjoumakaris SI, Dumont AS, Ghobrial GM, Kung D, Rosenwasser RH, Chalouhi N: Biology of cerebral arteriovenous malformations with a focus on inflammation. J Cereb Blood Flow Metab 35: 167-175, 2014.
9. Ding D, Xu Z, Yen CP, Starke RM, Sheehan JP: Radiosurgery for unruptured cerebral arteriovenous malformations in pediatric patients. Acta Neurochir (Wien) 157:281-291, 2014.
23. Isnard J, Guenot M, Ostrowsky K, Sindou M, Mauguiere F: The role of the insular cortex in temporal lobe epilepsy. Ann Neurol 48:614-623, 2000.
34. Murphy MJ: Long-term follow-up of seizures associated with cerebral arteriovenous malformations. Results of therapy. Arch Neurol 42:477-479, 1985.
10. Ding D, Yen C, Starke RM, Xu Z, Sheehan JP: Effect of prior hemorrhage on intracranial arteriovenous malformation radiosurgery outcomes. Cerebrovasc Dis 39:53-62, 2014.
24. Jones JE, Berven NL, Ramirez L, Woodard A, Hermann BP: Long-term psychosocial outcomes of anterior temporal lobectomy. Epilepsia 43: 896-903, 2002.
35. 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 73:387-391, 1990.
11. Ding D, Yen CP, Starke RM, Xu Z, Sheehan JP: Radiosurgery for ruptured intracranial arteriovenous malformations. J Neurosurg 121:470-481, 2014.
25. Kim H, Abla AA, Nelson J, McCulloch CE, Bervini D, Morgan MK, Stapleton C, Walcott BP, Ogilvy CS, Spetzler RF, Lawton MT: Validation of the supplemented Spetzler-Martin grading system for brain arteriovenous malformations in a multicenter cohort of 1009 surgical patients. Neurosurgery 76:25-33, 2015.
36. Potts MB, Sheth SA, Louie J, Smyth MD, Sneed PK, McDermott MW, Lawton MT, Young WL, Hetts SW, Fullerton HJ, Gupta N: Stereotactic radiosurgery at a low marginal dose for the treatment of pediatric arteriovenous malformations: obliteration, complications, and functional outcomes. J Neurosurg Pediatr 14:1-11, 2014.
12. Ding D, Yen CP, Starke RM, Xu Z, Sun X, Sheehan JP: Outcomes following single-session radiosurgery for high-grade intracranial arteriovenous malformations. Br J Neurosurg 28:666-674, 2014. 13. Ding D, Yen CP, Starke RM, Xu Z, Sun X, Sheehan JP: Radiosurgery for Spetzler-Martin Grade III arteriovenous malformations. J Neurosurg 120:959-969, 2014. 14. Ding D, Yen CP, Xu Z, Starke RM, Sheehan JP: Radiosurgery for low-grade intracranial arteriovenous malformations. J Neurosurg 121:457-467, 2014. 15. Ding D, Yen CP, Xu Z, Starke RM, Sheehan JP: Radiosurgery for patients with unruptured intracranial arteriovenous malformations. J Neurosurg 118:958-966, 2013. 16. Ding D, Yen CP, Xu Z, Starke RM, Sheehan JP: Radiosurgery for primary motor and sensory cortex arteriovenous malformations: outcomes and the effect of eloquent location. Neurosurgery 73:816-824, 2013. 17. Fierstra J, Conklin J, Krings T, Slessarev M, Han JS, Fisher JA, Terbrugge K, Wallace MC, Tymianski M, Mikulis DJ: Impaired peri-nidal cerebrovascular reserve in seizure patients with brain arteriovenous malformations. Brain 134: 100-109, 2011.
26. Kraemer DL, Awad IA: Vascular malformations and epilepsy: clinical considerations and basic mechanisms. Epilepsia 35 (Suppl 6):S30-43, 1994. 27. Kurita H, Kawamoto S, Suzuki I, Sasaki T, Tago M, Terahara A, Kirino T: Control of epilepsy associated with cerebral arteriovenous malformations after radiosurgery. J Neurol Neurosurg Psychiatry 65:648-655, 1998. 28. Lawton MT, Du R, Tran MN, Achrol AS, McCulloch CE, Johnston SC, Quinnine NJ, Young WL: Effect of presenting hemorrhage on outcome after microsurgical resection of brain arteriovenous malformations. Neurosurgery 56: 485-493; discussion 485-493, 2005. 29. Lawton MT, Kim H, McCulloch CE, Mikhak B, Young WL: A supplementary grading scale for selecting patients with brain arteriovenous malformations for surgery. Neurosurgery 66: 702-713; discussion 713, 2010. 30. Mohr JP, Parides MK, Stapf C, Moquete E, Moy CS, Overbey JR, Al-Shahi Salman R, Vicaut E, Young WL, Houdart E, Cordonnier C, Stefani MA, Hartmann A, von Kummer R, Biondi A, Berkefeld J, Klijn CJ, Harkness K, Libman R, Barreau X, Moskowitz AJ: Medical management with or without interventional therapy for
WORLD NEUROSURGERY - [-]: ---, - 2015
37. Przybylowski CJ, Ding D, Starke RM, Yen CP, Quigg M, Dodson B, Ball BZ, Sheehan JP: Seizure and anticonvulsant outcomes following stereotactic radiosurgery for intracranial arteriovenous malformations. J Neurosurg, http://dx.doi.org/ 10.3171/2014.11.JNS141388 [Epub ahead of print], 2015, Jan 23. 38. Rossetti AO, Mortati KA, Black PM, Bromfield EB: Simple partial seizures with hemisensory phenomena and dysgeusia: an insular pattern. Epilepsia 46:590-591, 2005. 39. Schauble B, Cascino GD, Pollock BE, Gorman DA, Weigand S, Cohen-Gadol AA, McClelland RL: Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology 63:683-687, 2004. 40. Spetzler RF, Martin NA: A proposed grading system for arteriovenous malformations. J Neurosurg 65:476-483, 1986. 41. Stapf C, Khaw AV, Sciacca RR, Hofmeister C, Schumacher HC, Pile-Spellman J, Mast H, Mohr JP, Hartmann A: Effect of age on clinical and morphological characteristics in patients with brain arteriovenous malformation. Stroke 34: 2664-2669, 2003.
www.WORLDNEUROSURGERY.org
7
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
42. Stapf C, Mast H, Sciacca RR, Choi JH, Khaw AV, Connolly ES, Pile-Spellman J, Mohr JP: Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology 66: 1350-1355, 2006.
47. Turjman F, Massoud TF, Sayre JW, Vinuela F, Guglielmi G, Duckwiler G: Epilepsy associated with cerebral arteriovenous malformations: a multivariate analysis of angioarchitectural characteristics. AJNR Am J Neuroradiol 16:345-350, 1995.
43. Starke RM, Sheehan JP, Ding D, Liu KC, Kondziolka D, Crowley RW, Lunsford LD, Kassell NF: Conservative management or intervention for unruptured brain arteriovenous malformations. World Neurosurg 82:e668-669, 2014.
48. Wang JY, Yang W, Ye X, Rigamonti D, Coon AL, Tamargo RJ, Huang J: Impact on seizure control of surgical resection or radiosurgery for cerebral arteriovenous malformations. Neurosurgery 73: 648-655; discussion 655-646, 2013.
44. Starke RM, Yen CP, Ding D, Sheehan JP: A practical grading scale for predicting outcome after radiosurgery for arteriovenous malformations: analysis of 1012 treated patients. J Neurosurg 119:981-987, 2013. 45. Sun DQ, Carson KA, Raza SM, Batra S, Kleinberg LR, Lim M, Huang J, Rigamonti D: The radiosurgical treatment of arteriovenous malformations: obliteration, morbidities, and performance status. Int J Radiat Oncol Biol Phys 80:354-361, 2011. 46. Taylor CL, Dutton K, Rappard G, Pride GL, Replogle R, Purdy PD, White J, Giller C, Kopitnik TA Jr, Samson DS: Complications of preoperative embolization of cerebral arteriovenous malformations. J Neurosurg 100:810-812, 2004.
8
www.SCIENCEDIRECT.com
49. Wegner RE, Oysul K, Pollock BE, Sirin S, Kondziolka D, Niranjan A, Lunsford LD, Flickinger JC: A modified radiosurgery-based arteriovenous malformation grading scale and its correlation with outcomes. Int J Radiat Oncol Biol Phys 79:1147-1150, 2011. 50. Williamson A, Patrylo PR, Lee S, Spencer DD: Physiology of human cortical neurons adjacent to cavernous malformations and tumors. Epilepsia 44:1413-1419, 2003.
52. Yang SY, Kim DG, Chung HT, Paek SH: Radiosurgery for unruptured cerebral arteriovenous malformations: long-term seizure outcome. Neurology 78:1292-1298, 2012. 53. Yen CP, Ding D, Cheng CH, Starke RM, Shaffrey M, Sheehan J: Gamma Knife surgery for incidental cerebral arteriovenous malformations. J Neurosurg 121:1015-1021, 2014. 54. Yen CP, Sheehan JP, Schwyzer L, Schlesinger D: Hemorrhage risk of cerebral arteriovenous malformations before and during the latency period after Gamma Knife radiosurgery. Stroke 42:1691-1696, 2011.
Received 15 January 2015; accepted 15 February 2015 Citation: World Neurosurg. (2015). http://dx.doi.org/10.1016/j.wneu.2015.02.039 Journal homepage: www.WORLDNEUROSURGERY.org
51. Wolf HK, Roos D, Blumcke I, Pietsch T, Wiestler OD: Perilesional neurochemical changes in focal epilepsies. Acta Neuropathol 91:376-384, 1996.
Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
Cerebral Arteriovenous Malformations and Epilepsy, Part 1: Predictors of Seizure Presentation Dale Ding1, Robert M. Starke1, Mark Quigg2, Chun-Po Yen1, Colin J. Przybylowski3, Blair K. Dodson1, Jason P. Sheehan1
OBJECTIVE: Seizures are relatively common in patients harboring cerebral arteriovenous malformations (AVMs). Because the pathogenesis of AVM-associated epilepsy is not well-defined, we aim to determine the factors associated with seizure presentation in AVM patients.
-
METHODS: We evaluated our institutional AVM radiosurgery database, from 1989e2013, to select patients in whom pertinent clinical information at presentation and adequate clinical and radiologic follow-up was available. Baseline patient demographics and AVM angioarchitectural features were compared between patients with and without seizure presentation. In addition to standard descriptive statistics, logistic regression analyses were performed to identify predictors of seizure presentation.
-
RESULTS: Of the 1007 AVM patients included for analysis, 229 patients presented with seizures (22.7%). The incidence of seizure presentation was significantly higher in cortical than noncortical AVMs (33.1% vs. 6.6%, P < 0.0001). Among the cortical locations, occipital AVMs had the lowest rate of seizure presentation (21.5%, P [ 0.0012), whereas the rates of seizure presentation in frontal (37.3%), temporal (37.7%), and parietal (34.0%) AVMs were similar. The lack of prior AVM hemorrhage (P < 0.0001), larger nidus diameter (P < 0.0001), and cortical location (P < 0.0001) were independent predictors of seizure presentation in the multivariate analysis. The
-
Key words Epilepsy - Intracranial arteriovenous malformations - Neurologic manifestations - Seizures - Vascular malformations -
Abbreviations and Acronyms AED: Antiepileptic drug ARUBA: A Randomized Trial of Unruptured Brain AVMs AVM: Arteriovenous malformation CCM: Cerebral cavernous malformation CI: Confidence interval CT: Computed tomography DSA: Digital subtraction angiography IRB: Institutional review board MRI: Magnetic resonance imaging
WORLD NEUROSURGERY - [-]: ---, - 2015
strongest independent predictors of seizure presentation were lack of prior AVM hemorrhage (OR 16.8) and cortical location (OR 4.2). CONCLUSIONS: Large, unruptured, cortical nidi are most prone to seizure presentation in patients referred for radiosurgery. Further investigations of the molecular biology, neuronal and glial physiology, and natural history of AVMassociated epilepsy appear warranted.
-
INTRODUCTION
E
pilepsy is a common symptom of cerebral arteriovenous malformations (AVMs), but its importance is often overlooked in favor of intracranial hemorrhage, which is the most frequent and feared component of an AVM’s natural history (4, 15, 20, 35). However, AVM-associated epilepsy may also be debilitating and adversely impact a patient’s quality of life (24). Seizures are the most common presentation of unruptured AVMs and are the second most frequent presentation for all AVM patients following hemorrhage (19, 34, 45). Prior studies have linked various factors to AVM-associated epilepsy, including male gender, younger age, larger AVM size, frontal or temporal nidus location, superficial AVM topography, AVM location at an arterial borderzone, absence of intranidal aneurysms, and presence of a venous varix (18, 19, 22, 47).
OR: Odds ratio RBAS: Radiosurgery-based AVM score SD: Standard deviation VRAS: Virginia Radiosurgery AVM Scale Departments of 1Neurological Surgery and 2Neurology and 3School of Medicine, University of Virginia, Charlottesville, VA, USA To whom correspondence should be addressed: Jason P. Sheehan, M.D., Ph.D. [E-mail: [email protected]] Citation: World Neurosurg. (2015). http://dx.doi.org/10.1016/j.wneu.2015.02.039 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
1
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
However, these previous studies were performed in relatively small or medium-sized cohorts. In this first of a two-part study, we relate baseline patient and AVM characteristics to seizures at presentation using a large patient sample. The aim of the present case-control study is to define the predictors of seizure presentation in AVMs before stereotactic radiosurgery. METHODS Patient Selection Criteria and Cohort Composition We reviewed a prospectively collected database approved by our institutional review board (IRB), of approximately 1400 patients harboring AVMs who were treated with Gamma Knife radiosurgery at the University of Virginia from 1989e2013. In order to maintain the consistency of the patient cohort throughout both parts of this study, the inclusion criteria were the same for Parts 1 and 2 of the study, specifically as follows: patients with (1) sufficient history at presentation to determine the presence or absence of seizures at the time of presentation, (2) similar information available at the time of follow-up, and (3) at least 2 years of postradiosurgery follow-up or radiologic evidence of AVM obliteration. Patients treated with dose- or volume-staged radiosurgical approaches were excluded. The case cohort comprised patients with seizures at or around the time of radiosurgery. The control cohort was composed of patients without seizures at or around the time of radiosurgery. Patient Demographics and AVM Angioarchitectural Features The baseline patient variables were gender; age; prior surgical resection or embolization; prior AVM hemorrhage; clinical presentation (hemorrhage, seizure, focal neurological deficit, headache, incidental diagnosis, or other symptoms); and seizure status at presentation (presence or absence of seizures at presentation). Patients with seizures at presentation for radiosurgery were classified as having seizures of any type, regardless of preexisting antiepileptic drug (AED) therapy, prior AVM treatments, or prior rupture. The angioarchitecture of the AVM nidus was defined by a combination of magnetic resonance imaging (MRI) and catheter digital subtraction angiography. Before 1991, angiography alone was used for diagnosis in some patients, whereas afterward, both angiography and MRI were used for nidus definition. The baseline AVM variables were location (cortical vs. noncortical, eloquent vs. noneloquent); size (maximal diameter, volume); venous drainage pattern (single vs. multiple draining veins, superficial vs. deep venous drainage); and presence of associated aneurysms (intranidal or perinidal). Cortical AVM location was defined as location of the center of the nidus in the frontal, temporal, parietal, or occipital lobes. Eloquent location was defined as sensorimotor, language, and visual cortex; hypothalamus and thalamus; internal capsule; brainstem; cerebellar peduncles; and deep cerebellar nuclei (40). On the basis of the patient and AVM characteristics, the Spetzler-Martin grade, modified radiosurgery-based AVM score (RBAS), and Virginia Radiosurgery AVM Scale (VRAS) were determined (40, 44, 49). Statistical Analysis Statistical analysis was performed with the Stata 10.0 software program (College Station, Texas, USA). Data are presented as
2
www.SCIENCEDIRECT.com
mean and standard deviation (SD) for continuous variables and as frequency for categorical variables. The incidence of seizure presentation was calculated separately for each individual location. Categorical variables were compared using the chi-square test, and continuous variables were compared with the independent (two sample), unpaired Student’s t-tests. A P-value of less than 0.05 was considered statistically significant. Univariate logistic regression analysis was performed to identify factors associated with seizure presentation. The covariates included in the logistic regression analysis were the patient and AVM factors listed earlier. The odds ratio (OR), 95% confidence interval (CI), and P-value were reported for each covariate. Interaction and confounding were assessed through stratification and relevant expansion covariates. Factors with a P-value of less than 0.20 in the univariate analysis were entered into a multivariate logistic regression analysis to determine independent predictors of seizure presentation. RESULTS Patient Demographics and Clinical Characteristics Table 1 shows the comparisons of patients who presented with seizures against the control cohort. Of the 1007 patients with sufficient clinical data at or around the time of radiosurgery, seizure was the presenting diagnostic incident in 229 patients, for a seizure presentation rate of 22.7%. The mean interval between the onset of seizures and radiosurgery was 58.6 82.7 months. The proportion of patients who presented with seizures was significantly lower for prior surgical resection (5.2% vs. 13.8%, P < 0.0001) and prior AVM hemorrhages (9.2%, vs. 69.5%, P < 0.0001) and higher for prior embolization (31% vs. 22%, P ¼ 0.005). AVM Angioarchitecture Table 2 details the AVM angioarchitectural comparisons of the patient cohorts with and without seizure presentation. Patients with seizure presentation harbored significantly larger AVMs on the basis of maximum diameter (mean 2.7 vs. 2.2 cm, P < 0.0001)
Table 1. Comparison of Demographics and Clinical Characteristics of Patients with and without Seizure Presentation AVMs in Patients with Seizure Presentation (n [ 229)
AVMs in Patients without Seizure Presentation (n [ 778)
P value
Female gender
106 (46.3%)
390 (50.1%)
0.307
Age (mean SD years)
35.2 14.4
33.6 16.2
0.186
Prior surgical resection
12 (5.2%)
107 (13.8%)
< 0.0001*
Factor
Prior embolization
71 (31.0%)
171 (22.0%)
Prior AVM hemorrhage
21 (9.2%)
541 (69.5%)
0.005* < 0.0001*
SD, standard deviation. *Statistically significant (P < 0.05).
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
Table 2. Comparison of AVM Angioarchitectural Features in Patients with and without Seizure Presentation AVMs in Patients with Seizure Presentation (n [ 229)
Factor
AVMs in Patients without Seizure Presentation (n [ 778)
P value
Diameter (mean SD cm)
2.7 0.9
2.2 0.9
< 0.0001*
Volume (mean SD mL)
4.8 3.8
3.1 3.1
< 0.0001*
Cortical location
203 (88.7%)
411 (52.8%)
< 0.0001*
Eloquent location
146 (63.8%)
529 (68.0%)
0.230
10 (4.4%)
42 (5.4%)
0.541
Associated aneurysms
y
102 (44.5%)
472 (60.7%)
< 0.0001*
Deep venous drainage
84 (36.7%)
471 (60.5%)
< 0.0001*
Spetzler-Martin grade
I: 36 (15.7%), II: 90 (39.3%), III: 80 (34.9%), IV: 22 (9.6%), V: 1 (0.4%)
Single draining vein
RBAS (mean SD) VRAS
I: 111 (14.3%), II: 264 (33.9%), III: 318 (40.9%), IV: 85 (10.9%), V: 0
1.22 0.46
1.15 0.46
0-1: 64 (27.9%), 2: 81 (35.4%), 3: 78 (34.1%), 4: 6 (2.6%)
0.156 0.034*
0-1: 175 (22.5%), 2: 309 (39.7%), 3: 198 (25.5%), 4: 96 (12.3%)
< 0.0001*
SD, standard deviation; RBAS, radiosurgery-based AVM score; VRAS, Virginia Radiosurgery AVM Scale. *Statistically significant (P < 0.05). yAssociated aneurysms ¼ intranidal or perinidal aneurysms.
and volume (mean 4.8 vs. 3.1 mL, P < 0.0001). A significantly smaller proportion of AVMs in patients with seizure presentation had a single draining vein (44.5% vs. 60.7% vs. P < 0.0001) and deep venous drainage (36.7% vs. 60.5%, P < 0.0001). Patients with seizure presentation had significantly higher RBAS (mean 1.22 vs. 1.15, P ¼ 0.034) and lower VRAS (P < 0.0001).
Predictors of Seizure Presentation Of the 229 patients who presented with seizures, 203 AVMs were in cortical locations (88.6%) and 26 AVMs were in noncortical locations (11.4%). Of the 778 patients without seizures at presentation, 411 AVMs were in cortical locations (52.8%) and 367 AVMs were in noncortical locations (47.2%). The incidence of seizure presentation was 33.1% for cortical AVMs (203/614 patients) compared with 6.6% for noncortical AVMs (26/393 patients). Seizure presentation was significantly more common in patients with cortically based AVMs (P < 0.0001). Table 3 details the incidences of seizure presentation for cortical AVMs, stratified by location. The incidence of seizure presentations was significantly lower for occipital nidi (21.5%, P ¼ 0.0012) and were similar among the remaining cortical loci. Table 4 details the incidences of seizure presentation for noncortical AVMs, stratified by location. The incidences of seizure presentation were significantly higher for insular nidi (32%, P < 0.0001) and lower for cerebellar nidi (0, P ¼ 0.017), and they were similar among the remaining noncortical loci. Figure 1 shows the incidences of seizure presentation based on AVM location. Table 5 details the results of the univariate and multivariate logistic regression analyses for predictors of seizure presentation. Univariate logistic regression analysis determined
WORLD NEUROSURGERY - [-]: ---, - 2015
that lack of prior surgical resection (P ¼ 0.001), presence of prior embolization (P ¼ 0.005), no prior AVM hemorrhage (P < 0.0001), larger nidus maximal diameter (P < 0.0001), larger nidus volume (P < 0.0001), cortical nidus location (P < 0.0001), superficial venous drainage (P < 0.0001), multiple draining veins (P < 0.0001), higher RBAS (P ¼ 0.035), and lower VRAS (P ¼ 0.006) were significantly associated. Multivariate logistic regression analysis determined lack of prior AVM hemorrhage (P < 0.0001), larger nidus maximal diameter (P < 0.0001), and cortical location (P < 0.0001) were independent predictors of seizure presentation. The strongest independent predictor of seizure presentation was lack of prior hemorrhage (OR 16.8, 95% CI 10.3e27.2), followed by cortical location (OR 4.2, 95% CI 2.6e6.8).
Table 3. Incidences of Seizure Presentation by Location for Cortical AVMs, Listed in Descending Order
Cortical Location
Incidence of Seizure Presentation at Specified Cortical Location (%y)
Incidence of Seizure Presentation not at Specified Cortical Location (%y)
Temporal
37.7 (66/175)
31.2 (137/439)
0.122
Frontal
37.3 (53/142)
31.8 (150/472)
0.218
Parietal
34.0 (55/162)
32.7 (148/452)
0.779
Occipital
21.5 (29/135)
36.3 (174/479)
0.0012*
P value
*Statistically significant (P < 0.05). yReported as % (A/B), where A ¼ number of patients with seizure presentation and B ¼ total number of patients.
www.WORLDNEUROSURGERY.org
3
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
Table 4. Incidences of Seizure Presentation by Location for Noncortical AVM, Listed in Descending Order Incidence of Seizure Presentation at Specified Noncortical Location (%*)
Incidence of Seizure Presentation not at Specified Noncortical Location (%*)
P value
32.0 (8/25)
4.9 (18/368)
< 0.0001y
Basal ganglia
9.6 (8/83)
5.8 (18/310)
0.212
Corpus callosum
7.9 (3/38)
6.5 (23/355)
0.739
Brainstem
4.8 (4/84)
7.1 (22/309)
0.441
Thalamus
3.1 (3/96)
7.7 (23/297)
0.113
0 (0/67)
8.0 (26/326)
0.017*
Noncortical Location Insula
Cerebellum
*Reported as % (A/B), where A ¼ number of patients with seizure presentation and B ¼ total number of patients. yStatistically significant (P < 0.05).
DISCUSSION In light of the recent prospective studies, A Randomized Trial of Unruptured Brain AVMs (ARUBA) and the Scottish Audit of Intracranial Vascular Malformations, which showed superior outcomes with medical therapy compared with intervention, the management of unruptured AVMs has come under intense scrutiny (2, 30, 43). Given that seizures are the most common presentation of unruptured AVMs, it behooves us to improve our understanding of AVM-associated epilepsy, including preferential locations and predictors of its occurrence. Pathogenesis of AVM-Associated Epilepsy The pathogenesis of epilepsy secondary to AVMs has been previously explored, although our current understanding of this phenomenon remains relatively limited. Kraemer and Awad proposed a number of potential mechanisms for epileptogenesis in patients harboring vascular malformations, including neuronal
Figure 1. Bar plot showing the frequency of AVM patients with seizure presentation, stratified by nidus location.
4
www.SCIENCEDIRECT.com
degeneration and aberrant signaling, alterations in glial structure and physiology, and generation of reactive oxygen species (26). Williamson et al. showed that neurons surrounding cerebral cavernous malformations (CCMs) were more predisposed to hyperexcitability than neurons surrounding gliomas (50). Specifically, CCM-adjacent neurons were more likely to have high-amplitude, complex spontaneous synaptic events, whereas tumor-adjacent neurons were more likely to have low-amplitude, spontaneous excitatory activity. The authors postulated that hemosiderin deposition altered the physiologic environment surrounding CCMs by abrogating glutamate uptake and causing injury-induced synaptic reorganization, thereby promoting neuronal hypersynchronization and propagating perilesional excitatory stimuli to other cortical regions (50). In an immunohistochemical analysis of 30 vascular malformations and intra-axial neoplasms, Wolf et al. showed that expression of both excitatory and inhibitory neurotransmitter receptors and enzymes was increased in some specimens and decreased in others (51). The authors proposed that their findings suggested an imbalance of excitatory and inhibitory signaling in the perilesional cortex contributed to the pathogenesis of epilepsy in these patients. One should note that many of these putative epileptogenic mechanisms are extrapolated from observations or studies of CCMs, without equivalent electrophysiologic experiments in AVMs. However, because the molecular biology and clinical course of CCMs and AVMs are fundamentally distinct, the findings may not be generalizable and should therefore be interpreted with caution. This further highlights the limits of our knowledge of AVM-associated epileptogenesis. Predictors of Seizure Presentation in AVM Patients Most AVM studies are focused on determining and abrogating an AVM’s hemorrhage risk (4, 5, 8-12, 14, 21, 33, 42, 53, 54). However, seizures are relatively common in AVM patients and may significantly alter the medical and interventional management of AVMs. Al-Shahi and Warlow reported a 1% annual risk of de novo seizures in patients with untreated AVMs (1). Because the natural history of AVMs is known to differ according to patient- and AVM-related features, however, an AVM’s associated risk of seizures may vary as well. Due to the negative effect of seizures on patient neurological outcomes and quality of life, accurately predicting the occurrence of AVM-associated epilepsy can potentially impact the management of AVM patients. The strongest independent predictor of seizure presentation in our series was lack of prior hemorrhage (OR 16.8, P < 0.0001). Prior studies regarding predictors of AVM-associated epilepsy either excluded hemorrhage as a covariate in the statistical analysis or only included patients with unruptured AVMs (18, 19). Although intracranial hemorrhages can result in seizures, they are more likely to manifest as focal neurological deficits or global clinical deterioration (32). And because the impetus for intervention is much higher with ruptured AVMs than unruptured ones, many of these lesions may be embolized, extirpated, or irradiated before they can cause seizures. Additionally, it is possible that the gliotic tissue surrounding a ruptured nidus is less prone to epileptogenesis than the brain parenchyma adjacent to unruptured AVMs. In the current series of 1007 patients, the incidence of seizure presentation near the time of radiosurgery was 23%. Galletti et al.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
Table 5. Univariate and Multivariate Logistic Regression Analyses for Predictors of Seizure Presentation; Only Covariates with P < 0.20 in the Univariate Analysis Were Listed Univariate Factor
Multivariate
Odds Ratio
95% CI
P value
Odds Ratio
95% CI
P value
Increasing age
1.006
0.997e1.016
0.186
NS
NS
NS
No prior surgical resection
2.884
1.557e5.339
0.001*
NS
NS
NS
Prior embolization
1.592
1.148e2.209
0.005*
NS
NS
NS
22.610
14.074e36.321
< 0.0001*
16.775
10.349e27.193
< 0.0001*
Larger diameter
1.056
1.039e1.074
< 0.0001*
1.038
1.017e1.059
< 0.0001*
Larger volume
1.144
1.093e1.198
< 0.0001*
NS
NS
NS
No prior AVM hemorrhage
Cortical location
6.972
< 0.0001*
4.223
2.631e6.778
< 0.0001*
Superficial venous drainage
2.648
1.953e3.592
< 0.0001*
NS
NS
NS
Multiple draining veins
1.423
1.224e1.654
< 0.0001*
NS
NS
NS
Lower Spetzler-Martin grade
1.122
Higher RBAS
1.392
Lower VRAS
1.233
4.527-10.737
0.947e1.329 1.024-1.892 1.062e1.431
0.184
NS
NS
NS
0.035*
NS
NS
NS
0.006*
NS
NS
NS
NS, not significant (P 0.05); RBAS, radiosurgery-based AVM score; VRAS, Virginia Radiosurgery AVM Scale. *Statistically significant (P < 0.05).
examined a cohort of 101 AVM patients with a 31% incidence of seizure presentation (18). Multivariate logistic regression analysis found frontal and temporal lobe locations and superficial topography to be independently associated with seizure presentation. Turjman et al. performed a detailed radiographic analysis of 100 AVM patients with a 47% incidence of seizure presentation (47). Nidus supply by the middle cerebral artery, the external carotid artery, or a cortical feeding artery, as well as cortical, temporal superficial, and parietal superficial nidus locations, were identified as predictors of AVM-associated epilepsy. Our finding of cortical location as an independent predictor of seizure presentation (P < 0.0001) is in general agreement with the literature, although we did not find a specific location with a particularly high incidence of seizure presentation (18, 19, 22, 47). The rates of seizure presentation were over 30% for frontal, temporal, and parietal locations but were relatively lower in occipital nidi at 22% (P ¼ 0.0012). Also, in agreement with previous studies (22), greater AVM diameter was an independent predictor of seizure presentation (P < 0.0001). Large AVMs may have greater association with epilepsy than small AVMs due to their increased vascular steal, venous congestion, induction of hypoxia-mediated signaling pathways, and propensity to cause pathogenic alterations in neuronal and glial components of the perinidal cortex (17, 25). Chronic vascular steal by AVMs may cause hypoperfusion-induced remodeling of the perinidal cortex, thus facilitating an epileptogenic milieu (25). Conversely, Fiestra et al. showed, using a combination blood oxygen level-dependent MRI and angiography, that impaired cerebrovascular reactivity and venous congestion, rather than arterial steal, correlated significantly with AVM-associated epilepsy (17).
WORLD NEUROSURGERY - [-]: ---, - 2015
In an analysis of 542 AVM patients, Stapf et al. reported a 29% incidence of seizure presentation (41). Patients in the age ranges of 20 to 29 and 30 to 39 years had the highest rates of seizures presentation, compared with no patients with seizure presentation older than 60 years old. Hoh et al. studied 424 AVM patients with a 33% incidence of seizure presentation (22). Male gender, age younger than 65 years, AVM diameter greater than 3 cm, and temporal lobe location were significantly associated with seizure presentation. Garcin et al. analyzed a series of 155 AVM patients with a 29% incidence of seizure presentation (19). Multivariate logistic regression analysis identified male gender, frontal lobe location, and arterial borderzone location to be independent predictors of seizure presentation. In contrast to prior studies that reported an inverse correlation between age and frequency of seizure presentation (22, 41), the mean age of the patients in the present study with seizure presentation (35.2 years) was slightly higher than that of the patients without seizure presentation (33.6 years), although the difference was not significant (P ¼ 0.186). Male gender was more common in patients with seizure presentation (53.7% vs. 49.9%), unlike previous studies (19, 22), but the difference was not significant (P ¼ 0.307). Effect of AVM Intervention on Seizure Outcomes Although the primary goal of AVM intervention is obliteration of the nidus, in order to eliminate the future risk of hemorrhage, the impact of treatment on seizure outcomes is also an important consideration in the overall management of patients with AVMassociated epilepsy. Baranoski et al. performed a meta-analysis of AVM treatment-specific seizure outcomes and reported seizure control rates for 78%, 63%, and 49% for surgery, radiosurgery, and embolization (3). However, patients with obliterated AVMs after
www.WORLDNEUROSURGERY.org
5
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
radiosurgery had the highest rate of seizure control (85%). In a comparison of surgery to radiosurgery for AVM-associated epilepsy, Wang et al. found divergent seizure outcomes, depending on the baseline seizure status (48). Specifically, patients with seizures at presentation had significantly better seizure control after undergoing surgical AVM resection, whereas patients without seizures at presentation had significantly lower rates of de novo seizure after treatment with radiosurgery. Thus in addition to the angioarchitecture of an AVM nidus, the patient’s seizure status at presentation may influence the preferred treatment modality, if intervention is undertaken. A systematic review by Chen et al. reported seizure control and remission rates of 69% and 44%, respectively, after radiosurgery for patients with AVM-associated epilepsy (6). Seizure remission rates were found to be significantly higher in patients with complete AVM obliteration (82% vs. 41%, P ¼ 0.0007). Hoh et al. evaluated the seizure outcomes of 110 AVM patients who underwent multimodality AVM intervention (22). After a mean follow-up of 2.9 years, Engel class I outcome was achieved in 66% of patients. A short history of seizures, intracranial hemorrhage-related seizures, generalized seizures, deep and posterior fossa nidus locations, treatment with surgical resection, and complete AVM obliteration were significantly associated with Engel class I outcome. The seizure outcomes were similar among surgery, radiosurgery, and embolization when only obliterated nidi were considered (22). Thus AVM intervention appears to afford favorable seizure outcomes in the majority of patients with AVM-associated epilepsy. Although in some studies seizure remission appears to correlate with complete nidus obliteration, the association between the two outcomes is inconsistently reported in the literature (27, 39, 52). Study Limitations A limitation of our case-control study design is that all of the patients were referred for and underwent treatment with radiosurgery, so referral patterns and institutional treatment preferences may have biased the patient population of this study (6, 13, 31). Therefore patients selected for surgical resection, embolization without radiosurgery, and conservative management were not included in our analysis. As such, the clinical presentation and nidus characteristics of the patients in this study may not reflect those with significant hemorrhage from AVM rupture, surgically accessible AVMs, or large, unruptured AVMs in which the benefits of intervention were judged to outweigh the risks. Although the insula is a developmentally cortical structure, its location deep to the cortical surface and in close proximity to critical neural structures, such as the basal ganglia and internal capsule, distinguishes it from the lobar locations, as supported by prior studies of AVM-associated epilepsy, which have not classified the insula as a cortical structure (19, 47). Additionally, the insula is cytoarchitecturally and functionally distinct from the
REFERENCES 1. Al-Shahi R, Warlow C: A systematic review of the frequency and prognosis of arteriovenous malformations of the brain in adults. Brain 124: 1900-1926, 2001.
6
www.SCIENCEDIRECT.com
frontal, temporal, parietal, and occipital lobes. Finally, the insula is not a frequently identified origin of seizures, thus further distinguishing it from seizure-prone cortical sites, such as the temporal lobe (23, 38). If the insula has been classified as a cortical structure in the current study, the rate of seizure presentation for the cortical locations would be 33.0% (211/639 patients) and the rate of rate of seizure presentation in the non-cortical locations would be 4.9% (18/368 patients). These values were not significantly different from those reported previously in this study when the insula was categorized as a noncortical structure with respect to the rates of seizure presentation for either cortical (P ¼ 0.988) or noncortical (P ¼ 0.308) sites. Thus we do not believe that reclassifying the insular AVMs, which comprised only 2.5% (25/1007 patients) of the nidi in this study, as cortical AVMs would likely alter the major findings of our analysis. Because of the relatively long duration of the database (over 20 years), changes in neuroimaging modalities, interventional technology, and treatment paradigms over time may have affected the referral patterns to our center, thereby resulting in heterogeneous patient and AVM characteristics over time. Additionally, the disparities between the AVMs typically referred for microsurgery compared with radiosurgery may bias the angioarchitectural features of the nidi in our series (7, 16, 28, 29, 36, 46). Due to the nature of being a tertiary referral center for AVM radiosurgery, detailed clinical data, such as the type and severity of seizures, electroencephalography (EEG) findings, and prior AED therapy, were unavailable for some patients. Although the AVM patients with seizures were typically evaluated by an epileptologist, EEG confirmation of seizure activity was not required in the absence of generalized seizures. In a detailed analysis of seizure and anticonvulsant outcomes after AVM radiosurgery, we were able to obtain the data for seizure subtype and AED therapy for 73 of 229 AVM patients (32%) with seizures at presentation (37). We believe that the strength of the current study lies in the large number of patients, which provides higher statistical power to our analysis compared with prior studies. However, we acknowledge that the lack of complete details regarding seizure type and severity in a majority of patients represents a significant but common limitation of retrospective studies such as this one. CONCLUSIONS A combination of patient and nidus characteristics predicts seizure presentation. Unruptured, large, and cortically based AVMs are particularly prone to seizure presentation, and patients harboring such nidi likely warrant more rigorous neurological monitoring for signs of epilepsy. Future prospective studies are necessary to determine the natural history of seizures in unruptured AVM patients and the merits of long-term AED management in patients who are particularly prone to AVM-associated epilepsy.
2. Al-Shahi Salman R, White PM, Counsell CE, du Plessis J, van Beijnum J, Josephson CB, Wilkinson T, Wedderburn CJ, Chandy Z, St George EJ, Sellar RJ, Warlow CP: Outcome after conservative management or intervention for unruptured brain arteriovenous malformations. JAMA 311:1661-1669, 2014.
3. Baranoski JF, Grant RA, Hirsch LJ, Visintainer P, Gerrard JL, Gunel M, Matouk CC, Spencer DD, Bulsara KR: Seizure control for intracranial arteriovenous malformations is directly related to treatment modality: a meta-analysis. J Neurointerv Surg 6:684-690, 2013.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
ORIGINAL ARTICLE DALE DING ET AL.
4. Brown RD Jr, Wiebers DO, Forbes G, O’Fallon WM, Piepgras DG, Marsh WR, Maciunas RJ: The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 68:352-357, 1988. 5. Buell TJ, Ding D, Starke RM, Webster Crowley R, Liu KC: Embolization-induced angiogenesis in cerebral arteriovenous malformations. J Clin Neurosci 21:1866-1871, 2014. 6. Chen CJ, Chivukula S, Ding D, Starke RM, Lee CC, Yen CP, Xu Z, Sheehan JP: Seizure outcomes following radiosurgery for cerebral arteriovenous malformations. Neurosurg Focus 37:E17, 2014. 7. Ding D, Liu KC: Predictive capability of the Spetzler-Martin versus supplementary grading scale for microsurgical outcomes of cerebellar arteriovenous malformations. J Cerebrovasc Endovasc Neurosurg 15:307-310, 2013.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
18. Galletti F, Costa C, Cupini LM, Eusebi P, Hamam M, Caputo N, Siliquini S, Conti C, Moschini E, Lunardi P, Carletti S, Calabresi P: Brain arteriovenous malformations and seizures: an Italian study. J Neurol Neurosurg Psychiatry 85: 284-288, 2013. 19. Garcin B, Houdart E, Porcher R, Manchon E, Saint-Maurice JP, Bresson D, Stapf C: Epileptic seizures at initial presentation in patients with brain arteriovenous malformation. Neurology 78: 626-631, 2012. 20. Graf CJ, Perret GE, Torner JC: Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg 58:331-337, 1983. 21. Gross BA, Du R: Natural history of cerebral arteriovenous malformations: a meta-analysis. J Neurosurg 118:437-443, 2013.
unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. Lancet 383:614-621, 2014. 31. Moosa S, Chen CJ, Ding D, Lee CC, Chivukula S, Starke RM, Yen CP, Xu Z, Sheehan JP: Volumestaged versus dose-staged radiosurgery outcomes for large intracranial arteriovenous malformations. Neurosurg Focus 37:E18, 2014. 32. Morgenstern LB, Hemphill JC 3rd, Anderson C, Becker K, Broderick JP, Connolly ES Jr, Greenberg SM, Huang JN, MacDonald RL, Messe SR, Mitchell PH, Selim M, Tamargo RJ: Guidelines for the management of spontaneous intracerebral hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 41:2108-2129, 2010.
8. Ding D, Starke RM, Yen CP, Sheehan JP: Radiosurgery for cerebellar arteriovenous malformations: does infratentorial location affect outcome? World Neurosurg 82:e209-217, 2014.
22. 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: 303-309; discussion 309-311, 2002.
33. Mouchtouris N, Jabbour PM, Starke RM, Hasan DM, Zanaty M, Theofanis T, Ding D, Tjoumakaris SI, Dumont AS, Ghobrial GM, Kung D, Rosenwasser RH, Chalouhi N: Biology of cerebral arteriovenous malformations with a focus on inflammation. J Cereb Blood Flow Metab 35: 167-175, 2014.
9. Ding D, Xu Z, Yen CP, Starke RM, Sheehan JP: Radiosurgery for unruptured cerebral arteriovenous malformations in pediatric patients. Acta Neurochir (Wien) 157:281-291, 2014.
23. Isnard J, Guenot M, Ostrowsky K, Sindou M, Mauguiere F: The role of the insular cortex in temporal lobe epilepsy. Ann Neurol 48:614-623, 2000.
34. Murphy MJ: Long-term follow-up of seizures associated with cerebral arteriovenous malformations. Results of therapy. Arch Neurol 42:477-479, 1985.
10. Ding D, Yen C, Starke RM, Xu Z, Sheehan JP: Effect of prior hemorrhage on intracranial arteriovenous malformation radiosurgery outcomes. Cerebrovasc Dis 39:53-62, 2014.
24. Jones JE, Berven NL, Ramirez L, Woodard A, Hermann BP: Long-term psychosocial outcomes of anterior temporal lobectomy. Epilepsia 43: 896-903, 2002.
35. 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 73:387-391, 1990.
11. Ding D, Yen CP, Starke RM, Xu Z, Sheehan JP: Radiosurgery for ruptured intracranial arteriovenous malformations. J Neurosurg 121:470-481, 2014.
25. Kim H, Abla AA, Nelson J, McCulloch CE, Bervini D, Morgan MK, Stapleton C, Walcott BP, Ogilvy CS, Spetzler RF, Lawton MT: Validation of the supplemented Spetzler-Martin grading system for brain arteriovenous malformations in a multicenter cohort of 1009 surgical patients. Neurosurgery 76:25-33, 2015.
36. Potts MB, Sheth SA, Louie J, Smyth MD, Sneed PK, McDermott MW, Lawton MT, Young WL, Hetts SW, Fullerton HJ, Gupta N: Stereotactic radiosurgery at a low marginal dose for the treatment of pediatric arteriovenous malformations: obliteration, complications, and functional outcomes. J Neurosurg Pediatr 14:1-11, 2014.
12. Ding D, Yen CP, Starke RM, Xu Z, Sun X, Sheehan JP: Outcomes following single-session radiosurgery for high-grade intracranial arteriovenous malformations. Br J Neurosurg 28:666-674, 2014. 13. Ding D, Yen CP, Starke RM, Xu Z, Sun X, Sheehan JP: Radiosurgery for Spetzler-Martin Grade III arteriovenous malformations. J Neurosurg 120:959-969, 2014. 14. Ding D, Yen CP, Xu Z, Starke RM, Sheehan JP: Radiosurgery for low-grade intracranial arteriovenous malformations. J Neurosurg 121:457-467, 2014. 15. Ding D, Yen CP, Xu Z, Starke RM, Sheehan JP: Radiosurgery for patients with unruptured intracranial arteriovenous malformations. J Neurosurg 118:958-966, 2013. 16. Ding D, Yen CP, Xu Z, Starke RM, Sheehan JP: Radiosurgery for primary motor and sensory cortex arteriovenous malformations: outcomes and the effect of eloquent location. Neurosurgery 73:816-824, 2013. 17. Fierstra J, Conklin J, Krings T, Slessarev M, Han JS, Fisher JA, Terbrugge K, Wallace MC, Tymianski M, Mikulis DJ: Impaired peri-nidal cerebrovascular reserve in seizure patients with brain arteriovenous malformations. Brain 134: 100-109, 2011.
26. Kraemer DL, Awad IA: Vascular malformations and epilepsy: clinical considerations and basic mechanisms. Epilepsia 35 (Suppl 6):S30-43, 1994. 27. Kurita H, Kawamoto S, Suzuki I, Sasaki T, Tago M, Terahara A, Kirino T: Control of epilepsy associated with cerebral arteriovenous malformations after radiosurgery. J Neurol Neurosurg Psychiatry 65:648-655, 1998. 28. Lawton MT, Du R, Tran MN, Achrol AS, McCulloch CE, Johnston SC, Quinnine NJ, Young WL: Effect of presenting hemorrhage on outcome after microsurgical resection of brain arteriovenous malformations. Neurosurgery 56: 485-493; discussion 485-493, 2005. 29. Lawton MT, Kim H, McCulloch CE, Mikhak B, Young WL: A supplementary grading scale for selecting patients with brain arteriovenous malformations for surgery. Neurosurgery 66: 702-713; discussion 713, 2010. 30. Mohr JP, Parides MK, Stapf C, Moquete E, Moy CS, Overbey JR, Al-Shahi Salman R, Vicaut E, Young WL, Houdart E, Cordonnier C, Stefani MA, Hartmann A, von Kummer R, Biondi A, Berkefeld J, Klijn CJ, Harkness K, Libman R, Barreau X, Moskowitz AJ: Medical management with or without interventional therapy for
WORLD NEUROSURGERY - [-]: ---, - 2015
37. Przybylowski CJ, Ding D, Starke RM, Yen CP, Quigg M, Dodson B, Ball BZ, Sheehan JP: Seizure and anticonvulsant outcomes following stereotactic radiosurgery for intracranial arteriovenous malformations. J Neurosurg, http://dx.doi.org/ 10.3171/2014.11.JNS141388 [Epub ahead of print], 2015, Jan 23. 38. Rossetti AO, Mortati KA, Black PM, Bromfield EB: Simple partial seizures with hemisensory phenomena and dysgeusia: an insular pattern. Epilepsia 46:590-591, 2005. 39. Schauble B, Cascino GD, Pollock BE, Gorman DA, Weigand S, Cohen-Gadol AA, McClelland RL: Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology 63:683-687, 2004. 40. Spetzler RF, Martin NA: A proposed grading system for arteriovenous malformations. J Neurosurg 65:476-483, 1986. 41. Stapf C, Khaw AV, Sciacca RR, Hofmeister C, Schumacher HC, Pile-Spellman J, Mast H, Mohr JP, Hartmann A: Effect of age on clinical and morphological characteristics in patients with brain arteriovenous malformation. Stroke 34: 2664-2669, 2003.
www.WORLDNEUROSURGERY.org
7
ORIGINAL ARTICLE DALE DING ET AL.
PREDICTORS OF AVM-ASSOCIATED SEIZURE PRESENTATION
42. Stapf C, Mast H, Sciacca RR, Choi JH, Khaw AV, Connolly ES, Pile-Spellman J, Mohr JP: Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology 66: 1350-1355, 2006.
47. Turjman F, Massoud TF, Sayre JW, Vinuela F, Guglielmi G, Duckwiler G: Epilepsy associated with cerebral arteriovenous malformations: a multivariate analysis of angioarchitectural characteristics. AJNR Am J Neuroradiol 16:345-350, 1995.
43. Starke RM, Sheehan JP, Ding D, Liu KC, Kondziolka D, Crowley RW, Lunsford LD, Kassell NF: Conservative management or intervention for unruptured brain arteriovenous malformations. World Neurosurg 82:e668-669, 2014.
48. Wang JY, Yang W, Ye X, Rigamonti D, Coon AL, Tamargo RJ, Huang J: Impact on seizure control of surgical resection or radiosurgery for cerebral arteriovenous malformations. Neurosurgery 73: 648-655; discussion 655-646, 2013.
44. Starke RM, Yen CP, Ding D, Sheehan JP: A practical grading scale for predicting outcome after radiosurgery for arteriovenous malformations: analysis of 1012 treated patients. J Neurosurg 119:981-987, 2013. 45. Sun DQ, Carson KA, Raza SM, Batra S, Kleinberg LR, Lim M, Huang J, Rigamonti D: The radiosurgical treatment of arteriovenous malformations: obliteration, morbidities, and performance status. Int J Radiat Oncol Biol Phys 80:354-361, 2011. 46. Taylor CL, Dutton K, Rappard G, Pride GL, Replogle R, Purdy PD, White J, Giller C, Kopitnik TA Jr, Samson DS: Complications of preoperative embolization of cerebral arteriovenous malformations. J Neurosurg 100:810-812, 2004.
8
www.SCIENCEDIRECT.com
49. Wegner RE, Oysul K, Pollock BE, Sirin S, Kondziolka D, Niranjan A, Lunsford LD, Flickinger JC: A modified radiosurgery-based arteriovenous malformation grading scale and its correlation with outcomes. Int J Radiat Oncol Biol Phys 79:1147-1150, 2011. 50. Williamson A, Patrylo PR, Lee S, Spencer DD: Physiology of human cortical neurons adjacent to cavernous malformations and tumors. Epilepsia 44:1413-1419, 2003.
52. Yang SY, Kim DG, Chung HT, Paek SH: Radiosurgery for unruptured cerebral arteriovenous malformations: long-term seizure outcome. Neurology 78:1292-1298, 2012. 53. Yen CP, Ding D, Cheng CH, Starke RM, Shaffrey M, Sheehan J: Gamma Knife surgery for incidental cerebral arteriovenous malformations. J Neurosurg 121:1015-1021, 2014. 54. Yen CP, Sheehan JP, Schwyzer L, Schlesinger D: Hemorrhage risk of cerebral arteriovenous malformations before and during the latency period after Gamma Knife radiosurgery. Stroke 42:1691-1696, 2011.
Received 15 January 2015; accepted 15 February 2015 Citation: World Neurosurg. (2015). http://dx.doi.org/10.1016/j.wneu.2015.02.039 Journal homepage: www.WORLDNEUROSURGERY.org
51. Wolf HK, Roos D, Blumcke I, Pietsch T, Wiestler OD: Perilesional neurochemical changes in focal epilepsies. Acta Neuropathol 91:376-384, 1996.
Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.02.039
