Radiosurgery for low-grade intracranial arteriovenous malformations

Clinical article

Full access

Object

Low-grade, or Spetzler-Martin (SM) Grades I and II, arteriovenous malformations (AVMs) are associated with lower surgical morbidity rates than higher-grade lesions. While radiosurgery is now widely accepted as an effective treatment approach for AVMs, the risks and benefits of the procedure for low-grade AVMs, as compared with microsurgery, remain poorly understood. The authors of this study present the outcomes for a large cohort of low-grade AVMs treated with radiosurgery.

Methods

From an institutional radiosurgery database comprising approximately 1450 AVM cases, all patients with SM Grade I and II lesions were identified. Patients with less than 2 years of radiological follow-up, except those with complete AVM obliteration, were excluded from analysis. Univariate and multivariate Cox proportional-hazards and logistic regression analyses were used to determine factors associated with obliteration, radiation-induced changes (RICs), and hemorrhage following radiosurgery.

Results

Five hundred two patients harboring low-grade AVMs were eligible for analysis. The median age was 35 years, 50% of patients were male, and the most common presentation was hemorrhage (47%). The median AVM volume and prescription dose were 2.4 cm3 and 23 Gy, respectively. The median radiological and clinical follow-up intervals were 48 and 62 months, respectively. The cumulative obliteration rate was 76%. The median time to obliteration was 40 months, and the actuarial obliteration rates were 66% and 80% at 5 and 10 years, respectively. Independent predictors of obliteration were no preradiosurgery embolization (p < 0.001), decreased AVM volume (p = 0.005), single draining vein (p = 0.013), lower radiosurgery-based AVM scale score (p = 0.016), and lower Virginia Radiosurgery AVM Scale (Virginia RAS) score (p = 0.001). The annual postradiosurgery hemorrhage rate was 1.4% with increased AVM volume (p = 0.034) and lower prescription dose (p = 0.006) as independent predictors. Symptomatic and permanent RICs were observed in 8.2% and 1.4% of patients, respectively. No preradiosurgery hemorrhage (p = 0.011), a decreased prescription dose (p = 0.038), and a higher Virginia RAS score (p = 0.001) were independently associated with postradiosurgery RICs.

Conclusions

Spetzler-Martin Grade I and II AVMs are very amenable to successful treatment with stereotactic radiosurgery. While patient, physician, and institutional preferences frequently dictate the final course of treatment, radiosurgery offers a favorable risk-to-benefit profile for the management of low-grade AVMs.

Abbreviations used in this paper:AVM = arteriovenous malformation; DSA = digital subtraction angiography; RBAS = radiosurgery-based AVM scale; RIC = radiation-induced change; SM = Spetzler-Martin; Virginia RAS = Virginia Radiosurgery AVM Scale.

Object

Low-grade, or Spetzler-Martin (SM) Grades I and II, arteriovenous malformations (AVMs) are associated with lower surgical morbidity rates than higher-grade lesions. While radiosurgery is now widely accepted as an effective treatment approach for AVMs, the risks and benefits of the procedure for low-grade AVMs, as compared with microsurgery, remain poorly understood. The authors of this study present the outcomes for a large cohort of low-grade AVMs treated with radiosurgery.

Methods

From an institutional radiosurgery database comprising approximately 1450 AVM cases, all patients with SM Grade I and II lesions were identified. Patients with less than 2 years of radiological follow-up, except those with complete AVM obliteration, were excluded from analysis. Univariate and multivariate Cox proportional-hazards and logistic regression analyses were used to determine factors associated with obliteration, radiation-induced changes (RICs), and hemorrhage following radiosurgery.

Results

Five hundred two patients harboring low-grade AVMs were eligible for analysis. The median age was 35 years, 50% of patients were male, and the most common presentation was hemorrhage (47%). The median AVM volume and prescription dose were 2.4 cm3 and 23 Gy, respectively. The median radiological and clinical follow-up intervals were 48 and 62 months, respectively. The cumulative obliteration rate was 76%. The median time to obliteration was 40 months, and the actuarial obliteration rates were 66% and 80% at 5 and 10 years, respectively. Independent predictors of obliteration were no preradiosurgery embolization (p < 0.001), decreased AVM volume (p = 0.005), single draining vein (p = 0.013), lower radiosurgery-based AVM scale score (p = 0.016), and lower Virginia Radiosurgery AVM Scale (Virginia RAS) score (p = 0.001). The annual postradiosurgery hemorrhage rate was 1.4% with increased AVM volume (p = 0.034) and lower prescription dose (p = 0.006) as independent predictors. Symptomatic and permanent RICs were observed in 8.2% and 1.4% of patients, respectively. No preradiosurgery hemorrhage (p = 0.011), a decreased prescription dose (p = 0.038), and a higher Virginia RAS score (p = 0.001) were independently associated with postradiosurgery RICs.

Conclusions

Spetzler-Martin Grade I and II AVMs are very amenable to successful treatment with stereotactic radiosurgery. While patient, physician, and institutional preferences frequently dictate the final course of treatment, radiosurgery offers a favorable risk-to-benefit profile for the management of low-grade AVMs.

Abbreviations used in this paper:AVM = arteriovenous malformation; DSA = digital subtraction angiography; RBAS = radiosurgery-based AVM scale; RIC = radiation-induced change; SM = Spetzler-Martin; Virginia RAS = Virginia Radiosurgery AVM Scale.

Arteriovenous malformations (AVMs) are among the most complex intracranial vascular pathologies and are associated with wide variations in clinical presentation, location, size, and angioarchitecture.5 There is a paucity of prospective, multicenter data regarding the natural history of and long-term treatment outcomes for AVMs, owing to their rarity. Therefore, AVMs continue to represent a significant challenge to the cerebrovascular community, and opinions differ regarding the optimal management of these lesions.6,9,41,42 A number of AVM grading systems have been developed over the years with the ultimate goal of predicting treatment outcomes based on AVM subgroup.19,30,40,43,49 For surgical planning, however, no classification is more widely used than the Spetzler-Martin (SM) grading scale because of its simplicity and reliability in predicting microsurgical AVM outcomes.40

Based on the composition of the Spetzler-Martin grading scale, Grade I and II, or low-grade, AVMs are typically small, drain into the superficial venous system, and are located in noneloquent cortex. Despite having decreased rates of microsurgical morbidity compared with their higher-grade counterparts, low-grade AVMs have not been definitively shown to have more benign natural histories.12,23 In fact, small AVMs may be more prone to hemorrhage than larger ones as a result of higher pressures in the feeding arteries.39 While microsurgery is the traditional approach for AVM obliteration, in the last 2 decades radiosurgery has emerged as an acceptable treatment alternative.10,32,45 However, the risk-to-benefit profile of radiosurgery for low-grade AVMs has yet to be fully delineated.14,23 We hypothesize that radiosurgery offers an effective, low-risk treatment option for patients harboring Spetzler-Martin Grade I or II AVM. To test this hypothesis, we critically analyzed our low-grade AVM radiosurgery experience.

Methods

Patient Population

Over the period from 1989 to 2012, approximately 1450 AVM cases were treated with Gamma Knife radiosurgery at the University of Virginia and registered in a prospective, institutional review board–approved database. We performed a retrospective review of this prospectively collected database. We identified all patients with SM Grade I or II AVM who had at least 2 years of radiological follow-up. All patients with less than 2 years of follow-up were excluded except those with AVMs that were obliterated prior to the 2-year follow-up.

Radiological Evaluation and Radiosurgical Targeting

An attending neurosurgeon and neuroradiologist at the University of Virginia evaluated each patient's neuroimaging studies, including digital subtraction angiography (DSA), CT, and MRI. Each AVM's size, location, morphology, angioarchitecture, and SM grade were assigned at the time of radiosurgical treatment. Our Gamma Knife radiosurgery technique has been described elsewhere.45 Prior to 1991, DSA and CT were routinely used for radiosurgical planning. After 1991, MRI was routinely used to enhance the precision of nidus definition. Dose planning was performed with the Kula software from 1989 to June 1994, after which it was replaced by the Gamma Plan software. Repeat radiosurgery was considered for patients who demonstrated a patent nidus 3 years after a prior radiosurgery.

Radiological and Clinical Follow-Up

For the first 2 years following radiosurgery, standard radiological follow-up consisted of MRI every 6 months. Subsequent to 2 years' follow-up, annual MRI was performed. Additional imaging, either CT or MRI, was performed for neurological deterioration. An attending neurosurgeon and neuroradiologist at our institution evaluated all postradiosurgery neuroimaging. Hemorrhage was defined by imaging only, regardless of clinical status. Obliteration was defined on MRI as the absence of flow voids or on DSA as the absence of abnormal arteriovenous shunting. Digital subtraction angiography was performed only to confirm the complete AVM obliteration demonstrated on MRI. Perinidal T2-weighted hyperintensities identified on postradiosurgery MRI were defined as radiation-induced changes (RICs). These changes were deemed symptomatic when correlated with clinical manifestation, typically headache, seizure, and/or focal neurological deficit. Since the University of Virginia is a tertiary care referral center for radiosurgery, clinical follow-up consisted of a combination of direct follow-up via return appointments or admissions to the University of Virginia clinics or hospital, respectively, and indirect follow-up via correspondence with outside hospitals and local primary care physicians.

Statistical Analysis

All statistical analyses were performed with the IBM SPSS 20 statistical software program. The actuarial obliteration rates following radiosurgical treatment were calculated with Kaplan-Meier analysis. Cox proportional-hazards regression analysis was used to identify predictors of obliteration, and logistic regression analysis was used to identify predictors of postradiosurgery hemorrhage and RICs. Factors analyzed in these analyses included sex, age, preradiosurgery hemorrhage, preradiosurgery embolization, AVM volume, AVM location (superficial vs deep and noneloquent vs eloquent), location of draining veins (superficial vs deep), number of draining veins (single vs multiple), prescription dose, number of isocenters, radiological presence of RIC, SM grade, radiosurgery-based AVM scale (RBAS) score, and Virginia Radiosurgery AVM Scale (Virginia RAS) score.

The patient, AVM, and treatment characteristics listed above were initially subjected to univariate analysis, either Cox proportional-hazards or logistic regression analysis. The covariates were subjected to multivariate analysis if the p value in the aforementioned univariate analysis was statistically significant (that is, < 0.05) using a Cox proportional-hazards regression model for covariates associated with obliteration and a logistic regression model for covariates associated with postradiosurgery hemorrhage or RIC with a backward likelihood ratio condition. The validity of the proportional-hazards assumption was detected using a log-minus-log survival plot. A hazard or odds ratio, 95% confidence interval, and p value were determined for each analyzed factor. Statistical significance was defined as a hazard or odds ratio with a 95% confidence interval not including 1.0 and a p value < 0.05. All statistical studies were 2-sided. The total number of risk years was calculated by the sum of the time intervals between radiosurgery and obliteration or the last radiological follow-up in nonobliterated AVMs. The annual postradiosurgery hemorrhage risk was defined by dividing the total number of individual postradiosurgery hemorrhages by the total number of risk years. Additionally, we identified 3 pairs of selected low-grade AVM subgroups and compared the outcomes between each pair: 1) nonembolized versus embolized lesions, 2) lesions with superficial versus deep venous drainage, and 3) lesions located in noneloquent versus eloquent cortex. The actuarial obliteration rates were compared with the log-rank test, the rates of cumulative RIC were compared with the chi-square test with Yates' correction, and the rates of symptomatic and permanent RICs were compared with Fisher's exact test.

Results

Patient Population

Five hundred two patients harboring SM Grade I or II AVM were eligible for analysis and consisted of 253 males (50.4%) and 249 females (49.6%) with a median age of 35.2 years. Seventy-nine patients (15.7%) were under the age of 18 years. Preradiosurgery therapies included microsurgical removal in 55 patients (11.0%) and embolization in 101 patients (20.1%). The most common presenting symptoms were hemorrhage in 235 patients (46.8%), seizure in 126 patients (25.1%), and headache in 83 patients (16.5%). The characteristics of the patient population are summarized in Table 1.

TABLE 1:

Summary of characteristics in 502 patients with SM Grade I or II AVM

ParameterNo.
sex
 M253 (50.4%)
 F249 (49.6%)
age in yrs
 median35.2
 mean (range)36.3 (4.1–81.8)
no. of pediatric patients (age <18 yrs)79 (15.7%)
preradiosurgery hemorrhage250 (49.8%)
preradiosurgery therapy
 embolization101 (20.1%)
 resection55 (11.0%)
presenting symptom
 hemorrhage235 (46.8%)
 seizure126 (25.1%)
 headache83 (16.5%)
 focal neurological deficit21 (4.2%)
 none15 (3.0%)

Radiosurgery Treatment Parameters and AVM Characteristics

The median radiosurgical parameters for the initial radiosurgery procedures were as follows: nidus volume 2.4 cm3, AVM margin dose 23 Gy, maximum dose 40 Gy, isodose line 50%, and number of isocenters 2. The AVM location was deep (basal ganglia, thalamus, or brainstem) in 32 cases (6.4%) and superficial in 470 (93.6%), as well as noneloquent in 298 (59.4%) and eloquent in 204 (40.6%). Venous drainage pattern was superficial in 391 cases (77.9%) and deep in 111 (22.1%). A single draining vein appeared in 292 cases (58.2%) and multiple draining veins in 210 (41.8%). The SM grade was I in 147 patients (29.3%) and II in 355 patients (70.7%).40 The RBAS score—a weighted sum of patient age, AVM volume, and AVM location—was a mean of 1.04, median of 1.03, and range of 0.21–2.95.49 The Virginia RAS score, which is based on history of preradiosurgery hemorrhage, AVM volume, and eloquence of AVM location, was 0–1 in 210 patients (41.8%), 2 points in 202 patients (40.2%), 3 in 77 patients (15.3%), and 4 in 13 patients (2.6%).43 Table 2 details the AVM characteristics and radiosurgical parameters.

TABLE 2:

Summary of AVM characteristics and treatment parameters

ParameterNo.
location
 superficial470 (93.6%)
 deep*32 (6.4%)
 noneloquent298 (59.4%)
 eloquent204 (40.6%)
venous drainage pattern
 superficial391 (77.9%)
 deep111 (22.1%)
no. of draining veins
 single292 (58.2%)
 multiple210 (41.8%)
max diameter in cm
 median2.0
 mean (range)2.0 (0.2–4.5)
AVM vol in cm3
 median2.4
 mean (range)2.8 (0.1–22.5)
prescription dose in Gy
 median23
 mean (range)22.1 (7–36)
max dose in Gy
 median40
 mean (range)40.7 (14–60)
isodose
 median50%
 mean (range)55.5% (44%–95%)
no. of isocenters
 median2
 mean (range)2.6 (1–22)
SM grade
 I147 (29.3%)
 II355 (70.7%)
RBAS score
 median1.03
 mean (range)1.04 (0.21–2.95)
 <1.00232 (46.2%)
 1.00–1.50209 (41.6%)
 1.51–2.0057 (11.4%)
 >2.004 (0.8%)
Virginia RAS score
 0–1210 (41.8%)
 2202 (40.2%)
 377 (15.3%)
 413 (2.6%)

Includes basal ganglia, thalamus, and brainstem.

Radiological Outcomes Following Radiosurgery

The overall radiological follow-up was a mean of 67.5 months (5.6 years), median of 47.9 months (4.0 years), and range of 5.7–239.4 months (0.5–20.0 years). Complete obliteration of the AVM nidus was confirmed by DSA in 304 patients (60.6%) and demonstrated by MRI alone in 78 patients (15.5%) for a cumulative AVM obliteration rate of 76.1%. Excluding all patients with less than 2 years of follow-up regardless of AVM obliteration yielded an obliteration rate of 71.0% (294 of 414 patients), consisting of 15.0% (62 of 414 patients) by MRI alone and 56% (232 of 414 patients) by DSA. The time to obliteration was a median of 39.5 months (3.3 years) with a range of 5.7–192.8 months (0.5–16.1 years). The actuarial obliteration rates at 3, 5, 7, and 10 years were 41%, 66%, 75%, and 80%, respectively. The obliteration rate of low-grade AVMs over time is shown in Fig. 1.

Fig. 1.
Fig. 1.

Kaplan-Meier plot demonstrating the obliteration rate over time for SM Grade I and II AVMs following treatment with radiosurgery. The table row beneath the x-axis shows the number of patients at each time point on the axis.

Obliteration rates were not significantly different between unruptured and ruptured AVMs (p = 0.132), nor between SM Grade I and II AVMs (p = 0.351). The actuarial obliteration rates of low-grade AVMs treated with embolization before radiosurgery were 24%, 34%, 49%, and 55% at 3, 5, 7, and 10 years, respectively. For low-grade AVMs without prior embolization, the actuarial obliteration rates were 45%, 74%, 81%, and 87%, respectively. Arteriovenous malformations without preradiosurgery embolization had significantly higher obliteration rates than embolized AVMs (p < 0.001). Figure 2 shows the obliteration rate over time for nonembolized and embolized low-grade AVMs. When classified by the Virginia RAS, AVMs with a score of 0–1 had actuarial obliteration rates of 54%, 78%, 85%, and 92% at 3, 5, 7, and 10 years, respectively; those with a score of 2–4 had actuarial obliteration rates of 32%, 58%, 67%, and 73%, respectively. Malformations with low Virginia RAS scores of 0–1 had significantly higher obliteration rates than those with intermediate or high scores of 2–4 (p = 0.001). Figure 3 shows the obliteration rate over time for AVMs with Virginia RAS scores of 0–1 versus 2–4.

Fig. 2.
Fig. 2.

Kaplan-Meier plot demonstrating the radiosurgical obliteration rate over time for SM Grade I and II AVMs with and without embolization prior to radiosurgery. Low-grade AVMs without preradiosurgery embolization had significantly higher rates of obliteration than AVMs receiving preradiosurgery embolization (p < 0.001). The table beneath the x-axis shows the number of patients at each time point on the axis for nonembolized and embolized low-grade AVMs.

Fig. 3.
Fig. 3.

Kaplan-Meier plot demonstrating the radiosurgical obliteration rate over time for SM Grade I and II AVMs categorized by Virginia RAS (Scores 0–1 vs 2–4). Low-grade AVMs with Virginia RAS scores of 0–1 had significantly higher rates of obliteration than those with scores of 2–4 (p = 0.001). The table beneath the x-axis shows the number of patients at each time point on the axis for low-grade AVMs with Virginia RAS scores of 0–1 and 2–4.

Predictors of AVM Obliteration

Univariate Cox proportional-hazards regression analysis identified no embolization preradiosurgery, decreased AVM volume, increased prescription dose, single draining vein, lower RBAS score, and lower Virginia RAS score as statistically significant predictors of obliteration. Multivariate analysis revealed factors independently associated with obliteration to be as follows: no preradiosurgery embolization (p < 0.001), decreased AVM volume (p = 0.005), single draining vein (p = 0.013), lower RBAS score (p = 0.016), and lower Virginia RAS score (p = 0.001). Table 3 details the results of the univariate and multivariate Cox proportional-hazards regression analyses for predictors of obliteration.

TABLE 3:

Factors predicting obliteration after radiosurgery

FactorUnivariate AnalysisMultivariate Analysis
HR95% CIp ValueHR95% CIp Value
male sex1.180.966–1.450.104
increased age1.000.997–1.010.273
preradiosurgery hemorrhage1.170.954–1.430.132
no preradiosurgery embolization2.331.75–3.10<0.001*1.971.47–2.64<0.001*
deep AVM location1.030.687–1.530.905
noneloquent AVM location1.030.835–1.260.802
decreased AVM vol1.221.15–1.30<0.001*1.111.03–1.190.005*
increased prescription dose1.081.05–1.11<0.001*1.030.993–1.070.108
fewer isocenters1.020.966–1.080.477
deep venous drainage1.200.947–1.520.130
single draining vein1.511.23–1.86<0.001*1.311.06–1.630.013*
no RIC1.010.816–1.240.944
lower SM grade1.110.893–1.380.352
lower RBAS score2.091.60–2.75<0.001*1.4731.08–2.010.016*
lower Virginia RAS score1.401.24–1.57<0.001*1.2551.11–1.420.001*

Statistically significant value.

Repeat Radiosurgery

Repeat radiosurgery was performed in 50 patients (10.0%) with incompletely obliterated AVMs, including 48 patients (9.6%) with one repeat treatment and 2 patients (0.4%) with two repeat treatments. The first repeat radiosurgical parameters were a median AVM volume of 0.9 cm3 (range 0.1–6.9 cm3), median prescription dose of 20 Gy (range 5–27 Gy), median maximum dose of 36 Gy (range 10–50 Gy), median isodose line of 50% (range 30%–96%), and median number of isocenters of 2 (range 1–20). The second repeat radiosurgery had parameters consisting of mean and median values of 0.6 and 0.1 cm3 for volume, 25 and 23 Gy for prescription dose, 50 and 46 Gy for maximum dose, 50% isodose line for both, and 1 and 2 isocenters, respectively.

Postradiosurgery Hemorrhage

Over 2158 risk years, 28 patients had 30 hemorrhages following radiosurgery, consisting of one hemorrhage in each of 26 patients and two hemorrhages in each of 2 patients, for an annual postradiosurgery hemorrhage rate of 1.4%. Seven patients experienced clinical deterioration due to postradiosurgery hemorrhage (1.4%). There were no cases of hemorrhage after complete AVM obliteration. Logistic regression analysis demonstrated increased AVM volume, decreased prescription dose, multiple draining veins, higher RBAS score, and higher Virginia RAS score as statistically significant predictors of postradiosurgery hemorrhage. Increased AVM volume (p = 0.034) and lower prescription dose (p = 0.006) were determined to be independently associated with postradiosurgery hemorrhage based on multivariate analysis. Table 4 details the results of the univariate and multivariate logistic regression analyses for predictors of postradiosurgery hemorrhage.

TABLE 4:

Factors predicting postradiosurgery hemorrhage

FactorUnivariate AnalysisMultivariate Analysis
OR95% CIp ValueOR95% CIp Value
female sex1.020.475–2.180.965
decreased age1.000.980–1.030.743
preradiosurgery hemorrhage1.010.471–2.160.983
preradiosurgery embolization1.640.700–3.840.255
deep location1.840.525–6.460.340
noneloquent location1.250.563–2.760.586
increased AVM vol1.241.09–1.400.001*1.141.01–1.300.034*
decreased prescription dose1.231.10–1.37<0.001*1.181.05–1.330.006*
fewer isocenters1.060.858–1.310.587
superficial venous drainage1.330.492–3.570.578
multiple draining veins2.641.19–5.850.016*2.110.920–4.840.078
lower SM grade1.150.509–2.610.732
higher RBAS score3.221.29–8.030.012*0.9750.263–3.620.970
higher Virginia RAS score1.651.09–2.500.018*1.180.732–1.900.497

Statistically significant value.

Radiation-Induced Changes and Cyst Formation Following Radiosurgery

Radiation-induced change was identified on postradiosurgery MRI in 184 patients (36.7%) at a mean of 12.2 months, median of 10.9 months, and range of 0.2–61.5 months following radiosurgery. The amount of time that the RICs endured was a mean of 22.1 months, median of 16.5 months, and range of 2.4–128.3 months. Radiation-induced changes were symptomatic in 41 patients (8.2%) consisting of headache in 14 patients (2.8%), neurological deficit in 18 (3.6%), and seizures in 9 (1.8%). Symptomatic RICs were transient in 34 patients (6.8%) and permanent in 7 (1.4%). Permanent RICs consisted of focal neurological deficits in 4 patients (0.8%) and seizures in 3 (0.6%). Excluding all patients with less than 2 years of radiological follow-up, even those with AVM obliteration, the rates of cumulative, symptomatic, and permanent RICs were 38.6% (160 of 414 patients), 8.2% (34 of 414 patients), and 1.4% (6 of 414 patients), respectively. Logistic regression analysis demonstrated no preradiosurgery hemorrhage, increased AVM volume, decreased prescription dose, higher RBAS score, and higher Virginia RAS score as statistically significant predictors of postradiosurgery RIC. Multivariate analysis demonstrated no preradiosurgery hemorrhage (p = 0.011), decreased prescription dose (p = 0.038), and higher Virginia RAS score (p = 0.001) to be independent predictors of RIC following radiosurgery. Table 5 details the results of the univariate and multivariate logistic regression analyses for predictors of RIC following radiosurgery. Cyst formation postradiosurgery was evident in 6 patients (1.2%) but was only symptomatic, resulting in new-onset seizures, in 1 patient (0.2%). No surgical intervention was undertaken for any of the patients with postradiosurgery cysts.

TABLE 5:

Factors predicting RICs following radiosurgery

FactorUnivariate AnalysisMultivariate Analysis
OR95% CIp ValueOR95% CIp Value
female sex1.070.743–1.550.707
increased age1.010.995–1.020.266
no preradiosurgery hemorrhage1.761.22–2.560.003*1.631.12–2.390.011*
preradiosurgery embolization1.090.689–1.730.709
superficial location1.260.577–2.760.561
eloquent location1.280.879–1.850.201
increased AVM vol1.111.02–1.210.021*1.030.928–1.140.574
decreased prescription dose1.091.02–1.160.010*1.071.00–1.140.038*
more isocenters1.040.954–1.120.405
superficial venous drainage1.190.755–1.860.461
single draining vein1.000.690–1.460.992
higher SM grade1.420.933–2.150.102
higher RBAS score1.661.04–2.660.035*1.280.763–2.130.354
higher Virginia RAS score1.261.03–1.550.026*1.431.15–1.780.001*

Statistically significant value.

Clinical Outcomes Following Radiosurgery

Overall clinical follow-up was a mean of 77.4 months (6.5 years), median of 61.6 months (5.1 years), and range of 6.8–239.4 months (0.6–20.0 years). Patients presenting with seizures had decreased seizure frequency (56 patients) or were seizure-free (26 patients) following radiosurgery in 16.3% of cases. Increased seizure frequency occurred in 6 patients (1.2%), and new-onset seizures were observed in 9 patients (1.8%) after radiosurgery (38 patients were stable from a seizure standpoint). Clinical improvement following radiosurgery was reported in 108 patients (21.5%). Permanent clinical deterioration was observed in 28 patients (5.6%); combined with 30 patients with transient clinical decline (6.0%), the overall rate of temporary and permanent clinical morbidity following radiosurgery was 11.6%.

Selected Low-Grade AVM Subgroups

As listed above and shown in Fig. 2, low-grade AVMs without embolization before radiosurgery had significantly higher obliteration rates than those treated with prior embolization (45% vs 24% at 3 years, 74% vs 34% at 5 years, p < 0.001). The rates of RIC were not significantly different between the nonembolized and embolized cohorts including cumulative (36.4% vs 37.6%, p = 0.912), symptomatic (9.2% vs 4.0%, p = 0.103), and permanent (1.5% vs 1.0%, p = 1.000) RICs. The postradiosurgery hemorrhage risks for nonembolized and embolized lesions were 1.4% and 1.5%, respectively.

The respective 3- and 5-year obliteration rates were 39% and 65% for low-grade AVMs with superficial venous drainage and 50% and 69% for those with deep venous drainage (p = 0.130). The rates of RIC were not significantly different between the lesions with superficial and those with deep venous drainage including cumulative (37.6% vs 33.3%, p = 0.477), symptomatic (7.2% vs 8.4%, p = 0.824), and permanent RICs (0.9% vs 1.5%, p = 1.000). The postradiosurgery hemorrhage risks for the superficial and deep venous drainage cohorts were 1.5% and 1.1%, respectively.

The respective 3- and 5-year obliteration rates were 43% and 65% for low-grade AVMs with a noneloquent location and 38% and 67% for those with an eloquent location (p = 0.802). The rates of RIC were not significantly different between the lesions located in noneloquent and those located in eloquent cortex including cumulative (33.9% vs 40.7%, p = 0.145), symptomatic (6.7% vs 10.3%, p = 0.203), and permanent RIC (1.3% vs 1.5%, p = 1.000). The postradiosurgery hemorrhage risks for the noneloquent and eloquent cohorts were 1.4% and 1.3%, respectively.

Discussion

Low-grade AVMs, SM Grade I or II, are often potentially treatable via all three standard therapeutic modalities, including endovascular embolization, microsurgical extirpation, and radiosurgery. The natural history of low-grade compared with high-grade AVMs remains the subject of controversy. The traditional dogma that smaller AVMs are more likely to rupture because of higher intranidal pressure has been challenged by natural history studies demonstrating higher rates of rupture in larger lesions.39 Smaller AVMs may be more likely to present with hemorrhage, but recent studies have shown that large AVMs can be associated with higher prospective hemorrhage rates.18,44 Nevertheless, the combination of the young patient age at which most AVMs are diagnosed and the natural history of AVMs exposes patients harboring these lesions to a substantial lifetime risk of hemorrhagerelated morbidity and mortality.

Microsurgical Removal of Low-Grade AVMs

In general, microsurgical removal represents an excellent treatment approach for low-grade AVMs. Early microsurgical series reported extremely low surgical complication rates, although in relatively few patients.12 Sisti et al. resected 67 small AVMs, less than 3 cm in diameter, with a morbidity of 1.5% and no mortality.38 Heros et al. reported 0% early and late neurological morbidity following microsurgical removal of Grade I AVMs (12 patients) and 5.7% early and 2.9% late neurological morbidity after the resection of Grade II AVMs (35 patients).13 Morgan et al. described their surgical outcomes in 220 patients with low-grade AVMs, 8% of whom underwent preoperative embolization.23 Overall morbidity and mortality combined was 1.4%: 0.6% (1 of 180 patients) for noneloquent and 5% (2 of 40 patients) for eloquent AVMs. As the authors note, two eloquent AVMs were managed conservatively owing to a perceived high risk of neurological injury. Therefore, the adjusted rate of operative morbidity for eloquent AVMs may have been as high as 9.5%.

The impressive results of many microsurgical AVM series have resulted in an aggressive stance by some cerebrovascular surgeons heavily favoring open resection over radiosurgery for low-grade AVMs.7 Pikus et al. treated 72 AVM cases with microsurgery, including 13 patients with SM Grade I, 13 with Grade II, and 28 with Grade III lesions.29 For AVMs smaller than 3 cm in diameter, there was no operative morbidity, and among all Grade I–III AVMs in the series, there was 1 case of operative morbidity (1.9%). After comparing their results with those in the available radiosurgical AVM literature at the time, the authors concluded that microsurgery was superior to radiosurgery for the treatment of low- and intermediate-grade AVMs. In contrast to the very low complication rates reported in the aforementioned studies, neurological decline or mortality rates of 8.9% (5 of 56 patients) for SM Grade I AVMs and 24.4% (30 of 123 patients) for Grade II AVMs were reported by Lawton et al.19 Beyond surgical technique differences, variations in morbidity and mortality rates in microsurgical series for SM Grade I and II AVMs may be attributable to patient selection, patient attributes (for example, age and medical comorbidities), study size, duration of follow-up, and rigorousness of follow-up.

Endovascular Embolization of Low-Grade AVMs

Low-grade AVMs, by nature of their relatively smaller size and frequent superficial venous drainage, can be successfully treated with stand-alone embolization therapy in select cases. In highly selected case series, obliteration rates with permanent embolic agents such as ethylene vinyl copolymer, otherwise known as Onyx (ev3), are up to 50%.35 More accurate estimates of complete AVM occlusion with embolization alone are significantly lower, around 20%–30%, even for low-grade lesions.16,28 While curative embolization of low-grade AVMs is feasible, the associated morbidity rate, over 10% in some series, and the risk of recanalization after initially complete angiographic occlusion remain barriers to its widespread acceptance as a primary management approach.16,25,28,35

As our understanding of the risks and benefits of endovascular embolization has advanced over time, we have typically adopted a conservative endovascular approach in the management of low-grade AVMs. Additionally, data from our radiosurgery experience, corroborated by similar reports from other centers, have demonstrated a deleterious effect of embolization on subsequent radiosurgical obliteration rates.2,8,37 At our institution, preradiosurgery embolization is reserved for lesions supplied by high-flow feeding arteries harboring perinidal or intranidal aneurysms and those with intranidal arteriovenous shunts. For the majority of low-grade AVMs, the neoadjuvant role of embolization prior to microsurgery or radiosurgery seems similarly minor. However, as advances in endovascular technology continue to be made and our understanding of multimodality AVM treatment continues to evolve, the impact of embolization on AVMs, such as after rather than before radiosurgery, has yet to reach its maximum potential.

Role of Radiosurgery for Treatment of SM Grade I and II AVMs

In a large cohort of low-grade AVM patients with long-term radiological follow-up (median 47.9 months), we reported a cumulative obliteration rate of 76%, which was 66% at 5 years and 80% at 10 years based on Kaplan-Meier survival analysis. No preradiosurgery obliteration (p < 0.001), decreased AVM volume (p = 0.005), single draining vein (p = 0.013), lower RBAS score (p = 0.016), and lower Virginia RAS score (p = 0.001) were identified as independent predictors of obliteration. As stated earlier, embolization prior to radiosurgery is a well-known deterrent to successful AVM obliteration.2,8,37 While beam scattering or absorption by embolic agents has been hypothesized as a possible explanation for this phenomenon, in vitro studies performed by Bing et al. did not demonstrate significant reductions in radiation dose to embolized AVMs.3 Alternate mechanisms for decreased radiosurgical obliteration of embolized AVMs include postembolization AVM recanalization or desensitization to radiation, embolization-induced angiogenesis, and increased difficulty of radiosurgical targeting due to an embolic agent artifact or alterations in the angioarchitecture.1,17,46,48 Smaller AVM volume is a widely accepted predictor of obliteration and is reconfirmed by the present study.10,14,32 Singular venous drainage probably indicates a more compact nidus, which enhances the efficacy of radiosurgical targeting.

The biggest disadvantage of radiosurgery compared with microsurgery is the latency period between radiosurgery and obliteration (typically 6 months–3 years) during which patients remain at risk for AVM rupture. In the majority of radiosurgery AVM series, the hemorrhage risk during the latency period appears to compare favorably with the natural history.15,20,21,50 The annual postradiosurgery hemorrhage risk of 1.4% in our series is lower than the 2%–4% per year rupture rate reported in AVM natural history studies.4,11,27,31 Increased AVM volume (p = 0.034) and lower prescription dose (p = 0.006) were independent predictors of postradiosurgery hemorrhage. The low annual postradiosurgery hemorrhage rate in the current study suggests that irradiation of SM Grade I and II AVMs results in vascular stabilization, thereby resulting in partial protection from latency period hemorrhage. However, some of the protection may be a result of complete obliteration occurring in AVMs for which there is a delay in recognizing obliteration in between fixed intervals of radiological assessment.

Postradiosurgery RIC was symptomatic in 8.2% and permanent in 1.4% of patients, with multivariate analysis determining that no preradiosurgery hemorrhage (p = 0.011), a decreased prescription dose (p = 0.038), and a higher Virginia RAS score (p = 0.001) were independently associated with RIC. While not definitively proven, AVM rupture prior to radiosurgery may create a perilesional gliotic margin, which protects adjacent normal parenchyma from RIC.47 The overall rate of permanent clinical decline was 5.6%, which included 3% of patients with new-onset or exacerbation of preexisting seizures.

In a similar study of 217 low-grade AVM patients treated with radiosurgery, Kano et al. reported actuarial obliteration rates of 58%, 87%, and 93% at 3, 5, and 10 years, respectively.14 Despite a similar median AVM volume and prescription dose, Kano et al. were able to achieve a higher obliteration rate than we did in the current study. Note that the aforementioned study excluded all patients with preradiosurgery embolization, which we determined was an independent predictor of decreased obliteration (p < 0.001) and was used in 20% of the AVMs in our series. The overall obliteration rate of nonembolized AVMs in our series was 81% (324 of 401 patients). The postradiosurgery hemorrhage rate reported by Kano et al. was 2.3% per year but was reduced significantly from 3.9% in the 1st year to 0.6% in the 2nd year after treatment.14 While no patients developed permanent neurological deficits from RIC, 6 patients died of latency period hemorrhage for a mortality rate of 2.8%.

Using retrospective studies, we had difficulty accurately comparing the complication rates between microsurgery and radiosurgery series.12,13,23,29 For instance, because of selection biases of the treating physicians and institutional referral patterns, the AVM characteristics of radiosurgery and microsurgery cohorts can differ significantly. In the microsurgical series by Morgan et al., 18% (40 of 220) of the AVMs were in eloquent cortex.23 By contrast, 41% of the patients in our series had AVMs in an eloquent location, which is known to negatively impact microsurgical outcomes but was not found to increase radiosurgery-induced complications in our cohort (p = 0.201).26,36 Nataf et al. compared the microsurgery and radiosurgery outcomes in two AVM cohorts, each composed of 39 patients, which were matched for age, initial symptoms, AVM size, AVM location, SM grade, and prior embolization.24 The obliteration rate was higher in the microsurgery cohort (91%) than in the radiosurgery cohort (81%), although the difference was not statistically significant (p = 0.10). Neurological deficit was more common in the microsurgery group (p < 0.001), whereas posttreatment AVM rupture was higher in the radiosurgery group (p = 0.04). Additionally, the functional outcomes at 12 and 24 months, as measured by the Glasgow Outcome and modified Rankin Scales, were better in the radiosurgery cohort.

Nevertheless, for low-grade AVMs, especially SM Grade I lesions, evaluation for microsurgical removal at an experienced cerebrovascular center should remain the initial step in patient management. Those patients deemed unsuitable for surgical treatment, because of AVM location or angioarchitecture or medical comorbidities, and those refusing craniotomy can be subsequently treated with radiosurgery.34 Moreover, patients with SM Grade I or II AVM with residual nidus after resection should be considered for radiosurgery. Finally, pending the publication of results of the ARUBA trial (A Randomized Trial of Unruptured Brain Arteriovenous Malformations), it remains to be determined whether conservative management is superior to any intervention for unruptured AVMs, which comprise 50% of our low-grade AVM cohort.22

Study Limitations

Our study is inherently limited by the biases associated with a retrospective single-center review of a single cohort without a comparable control cohort. The lack of prospective randomization results in a selection bias, which makes it difficult to obtain a precise comparison between our radiosurgical outcomes and those associated with AVM natural history and with other AVM treatments. Without randomization, we were unable to accurately ascertain the effect of prior embolization on radiosurgical efficacy. Additionally, 16% of patients with documented complete AVM obliteration did not have angiographic confirmation. While DSA remains the gold standard for evaluating the patency of an AVM nidus, Pollock et al. showed that for AVM neuroimaging follow-up, MRI has a 100% specificity and 91% negative predictive value, as compared with DSA.33 Moreover, the precedent of documenting AVM obliteration using MRI alone has been set by similar studies from other institutions.14 We acknowledge that despite the long-term clinical follow-up of patients in our study, the clinical data were not as complete as the radiological follow-up. Therefore, we were unfortunately unable to precisely determine the effect of either postradiosurgery hemorrhage or symptomatic RIC on a patient's functional status.

By including those patients with less than 2 years of radiological follow-up who had AVM obliteration and excluding others with short follow-up, we may have favorably biased our results toward higher obliteration rates. However, excluding patients with less than 2 years of follow-up would have biased results the other way, that is, toward lower rates of obliteration. In an attempt to maintain equipoise when reporting our results, we included the obliteration and RIC rates of all patients, save those with less than 2 years of follow-up, even those with obliteration. In our cohort of low-grade AVMs, the rates of symptomatic (8.2%) and permanent (1.4%) RICs were the same with both exclusion criteria. We also note that the exclusion of patients with less than 2 years' follow-up may have biased our results toward more favorable outcomes by potentially overestimating obliteration rates and underestimating RIC rates. Lastly, we emphasize that this study was not designed to compare radiosurgical with microsurgical outcomes for low-grade AVMs. While such comparisons will nonetheless be made, an accurate comparison of the two approaches will ultimately rely on future prospective studies.

Conclusions

Spetzler-Martin Grade I and II AVMs represent the substratification of AVMs most amenable to successful treatment, with the lowest risk of morbidity associated with microsurgery or radiosurgery. Because of the significant lifetime hemorrhage risk posed even by low-grade AVMs and their relatively favorable risk-to-benefit profile, as compared with Grade III–V lesions, aggressive intervention is typically pursued especially in cases of prior rupture. While microsurgery should be considered for low-grade AVMs given the procedure's generally low complication rate and immediate impact, we have found radiosurgery to be a safe and effective alternative. Embolization should be used selectively because of its detrimental effect on radiosurgical outcomes. For low-grade AVMs, embolization to reduce nidus size is generally not warranted prior to stereotactic radiosurgery. Although successful treatment with radiosurgery manifests in a delayed fashion, irradiation of AVMs can provide a degree of vascular stabilization during the latency period and eliminates the hemorrhage risk once obliteration is achieved.

Disclosure

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 to the study and manuscript preparation include the following. Conception and design: Sheehan, Ding. Acquisition of data: Sheehan, Ding, Yen. Analysis and interpretation of data: Sheehan, Ding, Yen, Xu. Drafting the article: Ding, Yen, Starke. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Sheehan. Statistical analysis: Ding, Xu.

References

  • 1

    Achrol ASGuzman RVarga MAdler JRSteinberg GKChang SD: Pathogenesis and radiobiology of brain arteriovenous malformations: implications for risk stratification in natural history and posttreatment course. Neurosurg Focus 26:5E92009

    • Search Google Scholar
    • Export Citation
  • 2

    Andrade-Souza YMRamani MScora DTsao MNter-Brugge KSchwartz ML: Embolization before radiosurgery reduces the obliteration rate of arteriovenous malformations. Neurosurgery 60:4434522007

    • Search Google Scholar
    • Export Citation
  • 3

    Bing FDoucet RLacroix FBahary JPDarsaut TRoy D: Liquid embolization material reduces the delivered radiation dose: clinical myth or reality?. AJNR Am J Neuroradiol 33:3203222012

    • Search Google Scholar
    • Export Citation
  • 4

    Brown RD JrWiebers DOForbes GO'Fallon WMPiepgras DGMarsh WR: The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 68:3523571988

    • Search Google Scholar
    • Export Citation
  • 5

    Choi JHMohr JP: Brain arteriovenous malformations in adults. Lancet Neurol 4:2993082005

  • 6

    Cockroft KM: Unruptured brain arteriovenous malformations should be treated conservatively: no. Stroke 38:331033112007

  • 7

    Davidson ASMorgan MK: How safe is arteriovenous malformation surgery? A prospective, observational study of surgery as first-line treatment for brain arteriovenous malformations. Neurosurgery 66:4985052010

    • Search Google Scholar
    • Export Citation
  • 8

    Ding DYen CPXu ZStarke RMSheehan JP: Radiosurgery for patients with unruptured intracranial arteriovenous malformations. Clinical article. J Neurosurg 118:9589662013

    • Search Google Scholar
    • Export Citation
  • 9

    Fiehler JStapf C: ARUBA—beating natural history in unruptured brain AVMs by intervention. Neuroradiology 50:4654672008

  • 10

    Friedman WABova FJMendenhall WM: Linear accelerator radiosurgery for arteriovenous malformations: the relationship of size to outcome. J Neurosurg 82:1801891995

    • Search Google Scholar
    • Export Citation
  • 11

    Graf CJPerret GETorner JC: Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg 58:3313371983

    • Search Google Scholar
    • Export Citation
  • 12

    Hamilton MGSpetzler RF: The prospective application of a grading system for arteriovenous malformations. Neurosurgery 34:271994

  • 13

    Heros RCKorosue KDiebold PM: Surgical excision of cerebral arteriovenous malformations: late results. Neurosurgery 26:5705781990

    • Search Google Scholar
    • Export Citation
  • 14

    Kano HLunsford LDFlickinger JCYang HCFlannery TJAwan NR: Stereotactic radiosurgery for arteriovenous malformations, Part 1: management of Spetzler-Martin Grade I and II arteriovenous malformations. Clinical article. J Neurosurg 116:11202012

    • Search Google Scholar
    • Export Citation
  • 15

    Karlsson BLax ISöderman M: Risk for hemorrhage during the 2-year latency period following gamma knife radiosurgery for arteriovenous malformations. Int J Radiat Oncol Biol Phys 49:104510512001

    • Search Google Scholar
    • Export Citation
  • 16

    Katsaridis VPapagiannaki CAimar E: Curative embolization of cerebral arteriovenous malformations (AVMs) with Onyx in 101 patients. Neuroradiology 50:5895972008

    • Search Google Scholar
    • Export Citation
  • 17

    Kiliç KKonya DKurtkaya OSav APamir MNKiliç T: Inhibition of angiogenesis induced by cerebral arteriovenous malformations using Gamma Knife irradiation. J Neurosurg 106:4634692007

    • Search Google Scholar
    • Export Citation
  • 18

    Langer DJLasner TMHurst RWFlamm ESZager ELKing JT Jr: Hypertension, small size, and deep venous drainage are associated with risk of hemorrhagic presentation of cerebral arteriovenous malformations. Neurosurgery 42:4814891998

    • Search Google Scholar
    • Export Citation
  • 19

    Lawton MTKim HMcCulloch CEMikhak BYoung WL: A supplementary grading scale for selecting patients with brain arteriovenous malformations for surgery. Neurosurgery 66:7027132010

    • Search Google Scholar
    • Export Citation
  • 20

    Maruyama KKawahara NShin MTago MKishimoto JKurita H: The risk of hemorrhage after radiosurgery for cerebral arteriovenous malformations. N Engl J Med 352:1461532005

    • Search Google Scholar
    • Export Citation
  • 21

    Maruyama KShin MTago MKishimoto JMorita AKawahara N: Radiosurgery to reduce the risk of first hemorrhage from brain arteriovenous malformations. Neurosurgery 60:4534592007

    • Search Google Scholar
    • Export Citation
  • 22

    Mohr JPMoskowitz AJStapf CHartmann ALord KMarshall SM: The ARUBA trial: current status, future hopes. Stroke 41:e537e5402010

    • Search Google Scholar
    • Export Citation
  • 23

    Morgan MKRochford AMTsahtsarlis ALittle NFaulder KC: Surgical risks associated with the management of Grade I and II brain arteriovenous malformations. Neurosurgery 54:8328392004

    • Search Google Scholar
    • Export Citation
  • 24

    Nataf FSchlienger MBayram MGhossoub MGeorge BRoux FX: Microsurgery or radiosurgery for cerebral arteriovenous malformations? A study of two paired series. Neurosurgery 61:39502007

    • Search Google Scholar
    • Export Citation
  • 25

    Natarajan SKGhodke BBritz GWBorn DESekhar LN: Multimodality treatment of brain arteriovenous malformations with microsurgery after embolization with onyx: single-center experience and technical nuances. Neurosurgery 62:121312262008

    • Search Google Scholar
    • Export Citation
  • 26

    O'Laoire SA: Microsurgical treatment of arteriovenous malformations in critical areas of the brain. Br J Neurosurg 9:3473601995

  • 27

    Ondra SLTroupp HGeorge EDSchwab K: The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg 73:3873911990

    • Search Google Scholar
    • Export Citation
  • 28

    Panagiotopoulos VGizewski EAsgari SRegel JForsting MWanke I: Embolization of intracranial arteriovenous malformations with ethylene-vinyl alcohol copolymer (Onyx). AJNR Am J Neuroradiol 30:991062009

    • Search Google Scholar
    • Export Citation
  • 29

    Pikus HJBeach MLHarbaugh RE: Microsurgical treatment of arteriovenous malformations: analysis and comparison with stereotactic radiosurgery. J Neurosurg 88:6416461998

    • Search Google Scholar
    • Export Citation
  • 30

    Pollock BEFlickinger JC: Modification of the radiosurgery-based arteriovenous malformation grading system. Neurosurgery 63:2392432008

    • Search Google Scholar
    • Export Citation
  • 31

    Pollock BEFlickinger JCLunsford LDBissonette DJKondziolka D: Factors that predict the bleeding risk of cerebral arteriovenous malformations. Stroke 27:161996

    • Search Google Scholar
    • Export Citation
  • 32

    Pollock BEFlickinger JCLunsford LDMaitz AKondziolka D: Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 42:123912471998

    • Search Google Scholar
    • Export Citation
  • 33

    Pollock BEKondziolka DFlickinger JCPatel AKBissonette DJLunsford LD: Magnetic resonance imaging: an accurate method to evaluate arteriovenous malformations after stereotactic radiosurgery. J Neurosurg 85:104410491996

    • Search Google Scholar
    • Export Citation
  • 34

    Pollock BELunsford LDKondziolka DMaitz AFlickinger JC: Patient outcomes after stereotactic radiosurgery for “operable” arteriovenous malformations. Neurosurgery 35:181994

    • Search Google Scholar
    • Export Citation
  • 35

    Saatci IGeyik SYavuz KCekirge HS: Endovascular treatment of brain arteriovenous malformations with prolonged intranidal Onyx injection technique: long-term results in 350 consecutive patients with completed endovascular treatment course. Clinical article. J Neurosurg 115:78882011

    • Search Google Scholar
    • Export Citation
  • 36

    Schaller CSchramm J: Microsurgical results for small arteriovenous malformations accessible for radiosurgical or embolization treatment. Neurosurgery 40:6646741997

    • Search Google Scholar
    • Export Citation
  • 37

    Schwyzer LYen CPEvans AZavoian SSteiner L: Long-term results of gamma knife surgery for partially embolized arteriovenous malformations. Neurosurgery 71:113911482012

    • Search Google Scholar
    • Export Citation
  • 38

    Sisti MBKader AStein BM: Microsurgery for 67 intracranial arteriovenous malformations less than 3 cm in diameter. J Neurosurg 79:6536601993

    • Search Google Scholar
    • Export Citation
  • 39

    Spetzler RFHargraves RWMcCormick PWZabramski JMFlom RAZimmerman RS: Relationship of perfusion pressure and size to risk of hemorrhage from arteriovenous malformations. J Neurosurg 76:9189231992

    • Search Google Scholar
    • Export Citation
  • 40

    Spetzler RFMartin NA: A proposed grading system for arteriovenous malformations. J Neurosurg 65:4764831986

  • 41

    Stapf CMohr JP: Unruptured brain arteriovenous malformations should be treated conservatively: yes. Stroke 38:330833092007

  • 42

    Stapf CMohr JPChoi JHHartmann AMast H: Invasive treatment of unruptured brain arteriovenous malformations is experimental therapy. Curr Opin Neurol 19:63682006

    • Search Google Scholar
    • Export Citation
  • 43

    Starke RMYen CPDing DSheehan JP: A practical grading scale for predicting outcome after radiosurgery for arteriovenous malformations: analysis of 1012 treated patients. Clinical article. J Neurosurg 119:9819872013

    • Search Google Scholar
    • Export Citation
  • 44

    Stefani MAPorter PJterBrugge KGMontanera WWillinsky RAWallace MC: Large and deep brain arteriovenous malformations are associated with risk of future hemorrhage. Stroke 33:122012242002

    • Search Google Scholar
    • Export Citation
  • 45

    Steiner LLindquist CAdler JRTorner JCAlves WSteiner M: Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 77:181992

    • Search Google Scholar
    • Export Citation
  • 46

    Sure UBattenberg EDempfle ATirakotai WBien SBertalanffy H: Hypoxia-inducible factor and vascular endothelial growth factor are expressed more frequently in embolized than in nonembolized cerebral arteriovenous malformations. Neurosurgery 55:6636702004

    • Search Google Scholar
    • Export Citation
  • 47

    Takashima SBecker LE: Neuropathology of cerebral arteriovenous malformations in children. J Neurol Neurosurg Psychiatry 43:3803851980

    • Search Google Scholar
    • Export Citation
  • 48

    Valle RDZenteno MJaramillo JLee ADe Anda S: Definition of the key target volume in radiosurgical management of arteriovenous malformations: a new dynamic concept based on angiographic circulation time. J Neurosurg 109:Suppl41502008

    • Search Google Scholar
    • Export Citation
  • 49

    Wegner REOysul KPollock BESirin SKondziolka DNiranjan A: A modified radiosurgery-based arteriovenous malformation grading scale and its correlation with outcomes. Int J Radiat Oncol Biol Phys 79:114711502011

    • Search Google Scholar
    • Export Citation
  • 50

    Yen CPSheehan JPSchwyzer LSchlesinger D: Hemorrhage risk of cerebral arteriovenous malformations before and during the latency period after GAMMA knife radiosurgery. Stroke 42:169116962011

    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

Article Information

Contributor Notes

Address correspondence to: Jason P. Sheehan, M.D., Ph.D., University of Virginia, Department of Neurosurgery, P.O. Box 800212, Charlottesville, VA 22908. email: jps2f@virginia.edu.Please include this information when citing this paper: published online March 7, 2014; DOI: 10.3171/2014.1.JNS131713.
Headings
Figures
  • View in gallery

    Kaplan-Meier plot demonstrating the obliteration rate over time for SM Grade I and II AVMs following treatment with radiosurgery. The table row beneath the x-axis shows the number of patients at each time point on the axis.

  • View in gallery

    Kaplan-Meier plot demonstrating the radiosurgical obliteration rate over time for SM Grade I and II AVMs with and without embolization prior to radiosurgery. Low-grade AVMs without preradiosurgery embolization had significantly higher rates of obliteration than AVMs receiving preradiosurgery embolization (p < 0.001). The table beneath the x-axis shows the number of patients at each time point on the axis for nonembolized and embolized low-grade AVMs.

  • View in gallery

    Kaplan-Meier plot demonstrating the radiosurgical obliteration rate over time for SM Grade I and II AVMs categorized by Virginia RAS (Scores 0–1 vs 2–4). Low-grade AVMs with Virginia RAS scores of 0–1 had significantly higher rates of obliteration than those with scores of 2–4 (p = 0.001). The table beneath the x-axis shows the number of patients at each time point on the axis for low-grade AVMs with Virginia RAS scores of 0–1 and 2–4.

References
  • 1

    Achrol ASGuzman RVarga MAdler JRSteinberg GKChang SD: Pathogenesis and radiobiology of brain arteriovenous malformations: implications for risk stratification in natural history and posttreatment course. Neurosurg Focus 26:5E92009

    • Search Google Scholar
    • Export Citation
  • 2

    Andrade-Souza YMRamani MScora DTsao MNter-Brugge KSchwartz ML: Embolization before radiosurgery reduces the obliteration rate of arteriovenous malformations. Neurosurgery 60:4434522007

    • Search Google Scholar
    • Export Citation
  • 3

    Bing FDoucet RLacroix FBahary JPDarsaut TRoy D: Liquid embolization material reduces the delivered radiation dose: clinical myth or reality?. AJNR Am J Neuroradiol 33:3203222012

    • Search Google Scholar
    • Export Citation
  • 4

    Brown RD JrWiebers DOForbes GO'Fallon WMPiepgras DGMarsh WR: The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 68:3523571988

    • Search Google Scholar
    • Export Citation
  • 5

    Choi JHMohr JP: Brain arteriovenous malformations in adults. Lancet Neurol 4:2993082005

  • 6

    Cockroft KM: Unruptured brain arteriovenous malformations should be treated conservatively: no. Stroke 38:331033112007

  • 7

    Davidson ASMorgan MK: How safe is arteriovenous malformation surgery? A prospective, observational study of surgery as first-line treatment for brain arteriovenous malformations. Neurosurgery 66:4985052010

    • Search Google Scholar
    • Export Citation
  • 8

    Ding DYen CPXu ZStarke RMSheehan JP: Radiosurgery for patients with unruptured intracranial arteriovenous malformations. Clinical article. J Neurosurg 118:9589662013

    • Search Google Scholar
    • Export Citation
  • 9

    Fiehler JStapf C: ARUBA—beating natural history in unruptured brain AVMs by intervention. Neuroradiology 50:4654672008

  • 10

    Friedman WABova FJMendenhall WM: Linear accelerator radiosurgery for arteriovenous malformations: the relationship of size to outcome. J Neurosurg 82:1801891995

    • Search Google Scholar
    • Export Citation
  • 11

    Graf CJPerret GETorner JC: Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg 58:3313371983

    • Search Google Scholar
    • Export Citation
  • 12

    Hamilton MGSpetzler RF: The prospective application of a grading system for arteriovenous malformations. Neurosurgery 34:271994

  • 13

    Heros RCKorosue KDiebold PM: Surgical excision of cerebral arteriovenous malformations: late results. Neurosurgery 26:5705781990

    • Search Google Scholar
    • Export Citation
  • 14

    Kano HLunsford LDFlickinger JCYang HCFlannery TJAwan NR: Stereotactic radiosurgery for arteriovenous malformations, Part 1: management of Spetzler-Martin Grade I and II arteriovenous malformations. Clinical article. J Neurosurg 116:11202012

    • Search Google Scholar
    • Export Citation
  • 15

    Karlsson BLax ISöderman M: Risk for hemorrhage during the 2-year latency period following gamma knife radiosurgery for arteriovenous malformations. Int J Radiat Oncol Biol Phys 49:104510512001

    • Search Google Scholar
    • Export Citation
  • 16

    Katsaridis VPapagiannaki CAimar E: Curative embolization of cerebral arteriovenous malformations (AVMs) with Onyx in 101 patients. Neuroradiology 50:5895972008

    • Search Google Scholar
    • Export Citation
  • 17

    Kiliç KKonya DKurtkaya OSav APamir MNKiliç T: Inhibition of angiogenesis induced by cerebral arteriovenous malformations using Gamma Knife irradiation. J Neurosurg 106:4634692007

    • Search Google Scholar
    • Export Citation
  • 18

    Langer DJLasner TMHurst RWFlamm ESZager ELKing JT Jr: Hypertension, small size, and deep venous drainage are associated with risk of hemorrhagic presentation of cerebral arteriovenous malformations. Neurosurgery 42:4814891998

    • Search Google Scholar
    • Export Citation
  • 19

    Lawton MTKim HMcCulloch CEMikhak BYoung WL: A supplementary grading scale for selecting patients with brain arteriovenous malformations for surgery. Neurosurgery 66:7027132010

    • Search Google Scholar
    • Export Citation
  • 20

    Maruyama KKawahara NShin MTago MKishimoto JKurita H: The risk of hemorrhage after radiosurgery for cerebral arteriovenous malformations. N Engl J Med 352:1461532005

    • Search Google Scholar
    • Export Citation
  • 21

    Maruyama KShin MTago MKishimoto JMorita AKawahara N: Radiosurgery to reduce the risk of first hemorrhage from brain arteriovenous malformations. Neurosurgery 60:4534592007

    • Search Google Scholar
    • Export Citation
  • 22

    Mohr JPMoskowitz AJStapf CHartmann ALord KMarshall SM: The ARUBA trial: current status, future hopes. Stroke 41:e537e5402010

    • Search Google Scholar
    • Export Citation
  • 23

    Morgan MKRochford AMTsahtsarlis ALittle NFaulder KC: Surgical risks associated with the management of Grade I and II brain arteriovenous malformations. Neurosurgery 54:8328392004

    • Search Google Scholar
    • Export Citation
  • 24

    Nataf FSchlienger MBayram MGhossoub MGeorge BRoux FX: Microsurgery or radiosurgery for cerebral arteriovenous malformations? A study of two paired series. Neurosurgery 61:39502007

    • Search Google Scholar
    • Export Citation
  • 25

    Natarajan SKGhodke BBritz GWBorn DESekhar LN: Multimodality treatment of brain arteriovenous malformations with microsurgery after embolization with onyx: single-center experience and technical nuances. Neurosurgery 62:121312262008

    • Search Google Scholar
    • Export Citation
  • 26

    O'Laoire SA: Microsurgical treatment of arteriovenous malformations in critical areas of the brain. Br J Neurosurg 9:3473601995

  • 27

    Ondra SLTroupp HGeorge EDSchwab K: The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg 73:3873911990

    • Search Google Scholar
    • Export Citation
  • 28

    Panagiotopoulos VGizewski EAsgari SRegel JForsting MWanke I: Embolization of intracranial arteriovenous malformations with ethylene-vinyl alcohol copolymer (Onyx). AJNR Am J Neuroradiol 30:991062009

    • Search Google Scholar
    • Export Citation
  • 29

    Pikus HJBeach MLHarbaugh RE: Microsurgical treatment of arteriovenous malformations: analysis and comparison with stereotactic radiosurgery. J Neurosurg 88:6416461998

    • Search Google Scholar
    • Export Citation
  • 30

    Pollock BEFlickinger JC: Modification of the radiosurgery-based arteriovenous malformation grading system. Neurosurgery 63:2392432008

    • Search Google Scholar
    • Export Citation
  • 31

    Pollock BEFlickinger JCLunsford LDBissonette DJKondziolka D: Factors that predict the bleeding risk of cerebral arteriovenous malformations. Stroke 27:161996

    • Search Google Scholar
    • Export Citation
  • 32

    Pollock BEFlickinger JCLunsford LDMaitz AKondziolka D: Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 42:123912471998

    • Search Google Scholar
    • Export Citation
  • 33

    Pollock BEKondziolka DFlickinger JCPatel AKBissonette DJLunsford LD: Magnetic resonance imaging: an accurate method to evaluate arteriovenous malformations after stereotactic radiosurgery. J Neurosurg 85:104410491996

    • Search Google Scholar
    • Export Citation
  • 34

    Pollock BELunsford LDKondziolka DMaitz AFlickinger JC: Patient outcomes after stereotactic radiosurgery for “operable” arteriovenous malformations. Neurosurgery 35:181994

    • Search Google Scholar
    • Export Citation
  • 35

    Saatci IGeyik SYavuz KCekirge HS: Endovascular treatment of brain arteriovenous malformations with prolonged intranidal Onyx injection technique: long-term results in 350 consecutive patients with completed endovascular treatment course. Clinical article. J Neurosurg 115:78882011

    • Search Google Scholar
    • Export Citation
  • 36

    Schaller CSchramm J: Microsurgical results for small arteriovenous malformations accessible for radiosurgical or embolization treatment. Neurosurgery 40:6646741997

    • Search Google Scholar
    • Export Citation
  • 37

    Schwyzer LYen CPEvans AZavoian SSteiner L: Long-term results of gamma knife surgery for partially embolized arteriovenous malformations. Neurosurgery 71:113911482012

    • Search Google Scholar
    • Export Citation
  • 38

    Sisti MBKader AStein BM: Microsurgery for 67 intracranial arteriovenous malformations less than 3 cm in diameter. J Neurosurg 79:6536601993

    • Search Google Scholar
    • Export Citation
  • 39

    Spetzler RFHargraves RWMcCormick PWZabramski JMFlom RAZimmerman RS: Relationship of perfusion pressure and size to risk of hemorrhage from arteriovenous malformations. J Neurosurg 76:9189231992

    • Search Google Scholar
    • Export Citation
  • 40

    Spetzler RFMartin NA: A proposed grading system for arteriovenous malformations. J Neurosurg 65:4764831986

  • 41

    Stapf CMohr JP: Unruptured brain arteriovenous malformations should be treated conservatively: yes. Stroke 38:330833092007

  • 42

    Stapf CMohr JPChoi JHHartmann AMast H: Invasive treatment of unruptured brain arteriovenous malformations is experimental therapy. Curr Opin Neurol 19:63682006

    • Search Google Scholar
    • Export Citation
  • 43

    Starke RMYen CPDing DSheehan JP: A practical grading scale for predicting outcome after radiosurgery for arteriovenous malformations: analysis of 1012 treated patients. Clinical article. J Neurosurg 119:9819872013

    • Search Google Scholar
    • Export Citation
  • 44

    Stefani MAPorter PJterBrugge KGMontanera WWillinsky RAWallace MC: Large and deep brain arteriovenous malformations are associated with risk of future hemorrhage. Stroke 33:122012242002

    • Search Google Scholar
    • Export Citation
  • 45

    Steiner LLindquist CAdler JRTorner JCAlves WSteiner M: Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 77:181992

    • Search Google Scholar
    • Export Citation
  • 46

    Sure UBattenberg EDempfle ATirakotai WBien SBertalanffy H: Hypoxia-inducible factor and vascular endothelial growth factor are expressed more frequently in embolized than in nonembolized cerebral arteriovenous malformations. Neurosurgery 55:6636702004

    • Search Google Scholar
    • Export Citation
  • 47

    Takashima SBecker LE: Neuropathology of cerebral arteriovenous malformations in children. J Neurol Neurosurg Psychiatry 43:3803851980

    • Search Google Scholar
    • Export Citation
  • 48

    Valle RDZenteno MJaramillo JLee ADe Anda S: Definition of the key target volume in radiosurgical management of arteriovenous malformations: a new dynamic concept based on angiographic circulation time. J Neurosurg 109:Suppl41502008

    • Search Google Scholar
    • Export Citation
  • 49

    Wegner REOysul KPollock BESirin SKondziolka DNiranjan A: A modified radiosurgery-based arteriovenous malformation grading scale and its correlation with outcomes. Int J Radiat Oncol Biol Phys 79:114711502011

    • Search Google Scholar
    • Export Citation
  • 50

    Yen CPSheehan JPSchwyzer LSchlesinger D: Hemorrhage risk of cerebral arteriovenous malformations before and during the latency period after GAMMA knife radiosurgery. Stroke 42:169116962011

    • Search Google Scholar
    • Export Citation
TrendMD
Cited By
Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 383 294 19
PDF Downloads 216 143 13
EPUB Downloads 0 0 0
PubMed
Google Scholar