Microsurgical treatment of arteriovenous malformations in pediatric patients: the Boston Children's Hospital experience

Free access

OBJECT

Outcomes of microsurgical treatment of arteriovenous malformations (AVMs) in children are infrequently reported across large cohorts.

METHODS

The authors undertook a retrospective review of departmental and hospital databases to obtain the medical data of all patients up to 18 years of age who were diagnosed with cerebral AVMs. Demographic and AVM angioarchitectural characteristics were analyzed, and for the patients who underwent surgery, the authors also analyzed the estimated intraoperative blood loss, postoperative angiographically confirmed obliteration rates, and neurological complications and outcomes classified according to the modified Rankin Scale (mRS).

RESULTS

Of 117 children with cerebral AVMs, 94 underwent microsurgical resection (80%). Twenty (21%) of these 94 patients underwent adjunctive preoperative embolization. The overall postoperative angiographically confirmed obliteration rate was 94%. As part of a new protocol, the last 50 patients in this series underwent immediate perioperative angiography, improving the subsequent obliteration rate from 86% to 100% (p = 0.01). No other factors, such as a hemorrhagic AVM, size of the AVM, location, drainage, or Spetzler-Martin grade, had a statistically significant impact on the obliteration rate. Perioperative neurological deficits occurred in 17% of the patients, but the vast majority of these (77%) were predictable visual field cuts. Arteriovenous malformations that were hemorrhagic or located in noneloquent regions were each associated with lower rates of postoperative neurological complications (p = 0.05 and 0.002, respectively). In total, 94% of the children had good functional outcomes (mRS Scores 0–2), and these outcomes were significantly influenced by the mRS score on presentation before surgery (p = 0.01). A review of 1- and 5-year follow-up data indicated an overall annual hemorrhage rate of 0.3% and a recurrence rate of 0.9%.

CONCLUSIONS

Microsurgical resection of AVMs in children is associated with high rates of angiographically confirmed obliteration and low rates of significant neurological complications. Implementation of a protocol using perioperative angiography in this series led to complete radiographically confirmed obliteration of all AVMs, with low annual repeat hemorrhage and recurrence rates.

ABBREVIATIONSAVM = arteriovenous malformation; mRS = modified Rankin Scale.

OBJECT

Outcomes of microsurgical treatment of arteriovenous malformations (AVMs) in children are infrequently reported across large cohorts.

METHODS

The authors undertook a retrospective review of departmental and hospital databases to obtain the medical data of all patients up to 18 years of age who were diagnosed with cerebral AVMs. Demographic and AVM angioarchitectural characteristics were analyzed, and for the patients who underwent surgery, the authors also analyzed the estimated intraoperative blood loss, postoperative angiographically confirmed obliteration rates, and neurological complications and outcomes classified according to the modified Rankin Scale (mRS).

RESULTS

Of 117 children with cerebral AVMs, 94 underwent microsurgical resection (80%). Twenty (21%) of these 94 patients underwent adjunctive preoperative embolization. The overall postoperative angiographically confirmed obliteration rate was 94%. As part of a new protocol, the last 50 patients in this series underwent immediate perioperative angiography, improving the subsequent obliteration rate from 86% to 100% (p = 0.01). No other factors, such as a hemorrhagic AVM, size of the AVM, location, drainage, or Spetzler-Martin grade, had a statistically significant impact on the obliteration rate. Perioperative neurological deficits occurred in 17% of the patients, but the vast majority of these (77%) were predictable visual field cuts. Arteriovenous malformations that were hemorrhagic or located in noneloquent regions were each associated with lower rates of postoperative neurological complications (p = 0.05 and 0.002, respectively). In total, 94% of the children had good functional outcomes (mRS Scores 0–2), and these outcomes were significantly influenced by the mRS score on presentation before surgery (p = 0.01). A review of 1- and 5-year follow-up data indicated an overall annual hemorrhage rate of 0.3% and a recurrence rate of 0.9%.

CONCLUSIONS

Microsurgical resection of AVMs in children is associated with high rates of angiographically confirmed obliteration and low rates of significant neurological complications. Implementation of a protocol using perioperative angiography in this series led to complete radiographically confirmed obliteration of all AVMs, with low annual repeat hemorrhage and recurrence rates.

Children with arteriovenous malformations (AVMs) face a lifetime risk of life-threatening intracranial hemorrhage along with symptoms such as chronic headaches, seizures, and progressive neurological deficits due to focal cerebral steal phenomena.5,7,18,33 Effective counseling about treatment options and anticipated results depends on a clear understanding of the procedural morbidity and obliteration rates. Reports of treatment outcome in moderate- to large-size pediatric cohorts are rare.2,5,16,18,22,27,31 We reviewed our own surgical experience with AVMs to elucidate obliteration rates and treatment results in the context of an evolved institutional protocol.

Methods

Patients and Study Design

With the approval of the Boston Children's Hospital Institutional Review Board, we undertook a retrospective review of departmental and hospital databases to identify all patients up to 18 years of age in whom cerebral AVMs were diagnosed from 1995 through October 2013. For this period, complete radiographic and clinical records were available from all databases. Patients with arteriovenous fistulas, vein of Galen malformations, and spinal vascular malformations of any type were excluded from the analysis. We reviewed the records of the patients to identify the following variables: patient age at presentation, sex, mode of presentation, AVM angioarchitectural characteristics (that is, location, size, venous drainage, and associated aneurysms), embolization procedures, early and late angiographically confirmed obliteration rates, the estimated surgical blood loss, and neurological complications. The neurological conditions at presentation and at last follow-up were classified by the modified Rankin Scale (mRS).39 Statistical analysis for categorical variables was performed with the Fisher exact test, and comparisons between mean values of continuous variables were performed with a 2-sided t-test.

Surgical Technique

Prior to incision, we ensured that matched blood products were in the operating room, that the microscope was draped, and that temporary aneurysm clips were selected and loaded. In all cases, a craniotomy was performed, ensuring that the margins of the lesion lay well within the margins of the operative exposure, and the dissection was performed under the operating microscope once the dura was opened. When feasible, an early identified superficial draining vein could be followed proximally to localize the nidus. Easily identifiable and major arterial feeders were then coagulated.24 An anatomical, circumferential cortical and subcortical dissection was performed with serial coagulation of supplying arterial pedicles before disconnecting the venous drainage. Indocyanine green angiography was variably employed to help distinguish arterial feeders from arterialized veins, identify en passage vessels, and to confirm resection. Since 2008, we have adopted a formal protocol in which all patients remain under general anesthesia and are transferred to the dedicated angiography suite (50 patients in this study were treated with this protocol). If a residual AVM is detected, the patient is returned to the operating room for AVM resection; otherwise, anesthesia is reversed, and extubation is performed in the angiography suite.

Embolization Technique

As part of the new protocol, since 2008, all children were evaluated for potential preoperative embolization before microsurgical resection. In most cases, embolization was performed on the day prior to microsurgical resection, with the patient being monitored in the intensive care unit overnight. This timing allowed for a full day and early start for the subsequent resection and was appropriate for most cases; some small AVMs were embolized, followed immediately by resection in the same anesthesia period. All of these procedures were performed under general anesthesia.

The goal of the embolization was to deeply penetrate the nidus with the embolic agent (most commonly Onyx [ev3]) while carefully avoiding occlusion of the venous drainage. Arterial pedicles were chosen on the basis of the access they provide to the AVM nidus and their location. Deep pedicles that were not expected to be encountered along the surgical track from the periphery to the AVM were systematically embolized if feasible. If resection was scheduled for the following day, a dexmedetomidine drip was initiated during the embolization session and then continued after a “deep” extubation. This procedure ensured strict blood pressure control while allowing for serial neurological examinations until surgery the following morning.

Results

Overall Population and Rationale for Nonoperative Management

Our overall cohort consisted of 117 pediatric patients with cerebral AVMs (Table 1). The cohort displayed no significant bias in the sex ratio (47% male), and presentation included hemorrhage in 52% of the cases, headache without hemorrhage in 20%, seizures in the absence of hemorrhage in 12%, and an asymptomatic/incidental finding in 16%. Four of the children had hereditary hemorrhagic telangiectasia (HHT), and 2 had a family history of cerebral AVMs without a diagnosis of HHT. By mRS score, 68% of patients were functionally independent (that is, their score was 0–2). Two children died as a result of their hemorrhage, 7 were referred for stereotactic radiosurgery because the AVM was located in deep or eloquent cortex, and the remaining 14 patients underwent observation only (because of a high Spetzler-Martin grade or a preference expressed by the family) and/or underwent treatment at other institutions. The remaining 94 patients (80%) were treated with microsurgical resection, 20 of whom (21%) had adjunctive preoperative embolization.

TABLE 1

Characteristics of all patients and of those undergoing surgery*

VariableAll Patients (%)Surgical Cohort (%)
No. of patients117 (100)94 (100)
Male sex55 (47)43 (46)
Presentation
 Incidental19 (16)12 (13)
 Headache (no hemorrhage)23 (20)17 (18)
 Seizure (no hemorrhage)14 (12)12 (13)
 Hemorrhage61 (52)53 (56)
Presenting mRS score
 019 (16)12 (13)
 149 (42)44 (47)
 212 (10)9 (10)
 0–2 combined80 (68)65 (69)
 312 (10)9 (10)
 49 (8)9 (10)
 514 (12)11 (12)
 62 (2)0 (0)
Location
 Lobar92 (79)77 (82)
 Deep supratentorial11 (9)4 (4)
 Cerebellar14 (12)13 (14)
AVM diameter
 Up to 3 cm60 (70)52 (76)
 3–6 cm21 (24)14 (21)
 >6 cm5 (6)2 (3)
Deep venous drainage45 (42)31 (35)
Spetzler-Martin grade
 I13 (15)13 (19)
 II36 (42)31 (46)
 III27 (31)22 (32)
 IV7 (8)2 (3)
 V3 (3)0 (0)
Associated aneurysm§12 (13)9 (13)

All values indicate number (%) of patients; in cases in which the number of patients in subgroups differs from the total numbers in the main groups (that is, all patients and the surgical cohort), the total number in each subgroup is indicated.

The total numbers of patients assessed for this variable were 86 for all patients and 68 for the surgical cohort; percentages were calculated with these totals.

The total numbers of patients assessed for this variable were 108 for all patients and 88 for the surgical cohort; percentages were calculated with these totals.

The total numbers of patients assessed for this variable were 91 for all patients and 72 for the surgical cohort; percentages were calculated with these totals.

Surgical Population: Clinical and Radiographic Presentation

The surgical cohort also showed no significant bias in the sex ratio (46% male). Mean age at the time of surgery was 10.9 years (SD 4.5 years, range 3 months to 18 years). Most of the patients presented with hemorrhage (56%), and the remaining patients had headaches (18%), seizures (13%), or were asymptomatic (13%). According to the mRS, 69% of the patients were independent (that is, had mRS scores of 0–2). The AVM location was lobar in 82% of the cases, cerebellar in 14%, and deep supratentorial in 4%. Most AVMs (76%) were small (up to 3 cm in diameter), 21% were 3–6 cm, and 3% were larger than 6 cm in diameter. Thirty-five percent of the AVMs had deep venous drainage, and 13% had an aneurysm either on an associated feeding artery or at an intranidal location. Classified by Spetzler-Martin grade,36 19% of the AVMs were Grade I, 46% were Grade II, 32% were Grade III, and 3% were Grade IV. Demographic and AVM angioarchitectural characteristics did not significantly differ between the surgical cohort and the cohort at large (Table 1).

Surgical Outcomes

The surgical outcomes are summarized in Table 2. Of 87 patients undergoing formal postoperative angiography, 82 showed obliteration of the AVM (94%). In the most recent 50 cases, angiography was performed according to the protocol described above, that is, immediately after wound closure. This immediate postoperative angiography clearly improved obliteration rates, which increased from 86% to 100% after initiation of this protocol (p = 0.01) (Table 3). Preoperative embolization, hemorrhagic presentation, larger AVM size, AVM location in eloquent regions, deep venous drainage, and Spetzler-Martin grades of III or IV did not significantly affect obliteration rates in this study, but deep venous drainage tended to be associated with a higher rate of nonobliteration (p = 0.11).

TABLE 2

Overall surgical outcomes

ParameterRelative Rate (%)
Obliteration rate82/87 (94)
Preop embolization20/94 (21)
Periop angiography50/94 (53)
Neurological complications
 Total16/94 (17)
 Visual field deficit12/94 (13)
 Hemiparesis/hemiplegia3/94 (3)
 New seizures3/94 (3)
Neurological complication by Spetzler-Martin grade
 I0/13 (0)
 II5/31 (16)
 III–IV7/24 (29)
Neurological outcome by mRS score
 045/94 (48)
 130/94 (32)
 0–288/94 (94)
TABLE 3

Factors for complete AVM obliteration rates*

Characteristic/Risk FactorObliteration Rate (%)p Value
YesNo
Preop embolization19/19 (100)63/68 (93)0.58
Hemorrhagic presentation49/53 (92)33/34 (97)0.64
Larger AVM size (≥3 cm)16/16 (100)47/48 (98)1.0
AVM in eloquent region48/51 (94)34/36 (94)1.0
Deep venous drainage25/28 (89)53/54 (98)0.11
Spetzler-Martin Grade III or IV22/22 (100)41/42 (98)1.0
Use of early postop angiography/new protocol50/50 (100)32/37 (86)0.01

The table compares the obliteration rates in the presence or absence of the 7 different factors and AVM characteristics.

Overall, the mean estimated blood loss during AVM surgery was 212 ml (median 150 ml). Surgery on hemorrhagic (p = 0.02) and smaller (p = 0.0001) AVMs was associated with lower operative blood loss, while preoperative embolization, AVM location, deep venous drainage, and Spetzler-Martin grade did not appear to significantly influence operative blood loss (Table 4).

TABLE 4

Mean intraoperative blood loss*

Characteristic/Risk FactorMean Estimated Blood Loss (ml)p Value
YesNo
Preop embolization2801910.16
Hemorrhagic presentation1473150.002
Larger size (≥3 cm)4531600.0001
AVM in eloquent region2022280.65
Deep venous drainage1602420.18
Spetzler-Martin Grade III or IV2632100.46

The mean estimated blood loss in all procedures was 212 ml (median 150 ml). The table compares the blood loss in the presence or absence of the 6 different factors and AVM characteristics.

Sixteen patients (17%) had new and lasting neurological deficits after surgery, but in 12 of these patients (75%), these deficits were anticipated visual field cuts, given that these 12 AVMs were located close to optic radiations, the calcarine cortex, or both. Three patients (3%) had newonset seizures requiring antiepileptic medication for long-term seizure control, and 3 had new postoperative hemiparesis due to proximity of the AVMs to the motor/ premotor cortex in 2 cases and to the internal capsule in 1 case. Evaluating the impact of Spetzler-Martin grade on neurological complications, we noted that none of the patients with Grade I, 16% with Grade II, and 29% with Grade III or IV AVMs had neurological complications. Arteriovenous malformations in noneloquent regions (p = 0.002) and hemorrhagic AVMs (p = 0.05) were associated with lower postoperative rates of neurological complications (Table 5).

TABLE 5

Rates of postoperative neurological complications*

Characteristic/FactorNeurological Complications (%)p Value
YesNo
Preop embolization3/20 (15)13/74 (18)1.0
Hemorrhagic presentation6/57 (11)10/37 (27)0.05
Larger AVM size (≥3 cm)5/16 (31)7/52 (13)0.14
AVM in eloquent region15/55 (27)1/39 (3)0.002
Deep venous drainage6/31 (19)10/57 (18)1.0
Spetzler-Martin Grade III or IV7/24 (29)5/44 (11)0.10
Use of early postop angiography/new protocol7/50 (14)9/44 (20)0.43

The table compares neurological complications in the presence or absence of the 7 different factors and AVM characteristics.

Long-Term Follow-Up

The mean clinical follow-up duration in this series was 3.6 years. In total, 54% of the patients were asymptomatic (that is, had an mRS score of 0) at the follow-up, compared with 13% before the surgery. Ninety-four percent of patients were clinically independent at the follow-up (mRS Score 0–2), an improvement from a rate of 69% prior to surgery; a higher mRS score before surgery significantly predisposed patients to a higher mRS score after surgery (p = 0.01) (Table 6). Seventy-six patients had at least 1 year of clinical follow-up; 24 patients had at least 5 years of follow-up; and both of these groups had similar rates of clinical independence (that is, an mRS score < 3 in 93% and 88% of patients, respectively).

TABLE 6

Rates of poor neurological outcome for various AVM characteristics and surgical procedures*

Characteristic/factorNeurological Outcome (%)p Value
YesNo
mRS score >2 before surgery5/29 (17)1/60 (2)0.01
Preop embolization1/20 (5)5/74 (7)1.0
Hemorrhagic presentation5/57 (9)1/37 (3)0.40
Larger AVM size (≥3 cm)1/16 (6)2/52 (4)0.56
AVM in eloquent region3/55 (5)3/39 (8)0.69
Deep venous drainage4/31 (13)2/57 (4)0.18
Spetzler-Martin Grade III or IV2/24 (8)1/44 (2)0.28
Use of early postop angiography/new protocol2/50 (4)4/44 (9)0.41

Poor neurological outcome was defined as an mRS score > 2.

Over a combined 337.3 patient-years of clinical follow-up, 3 postoperative hemorrhages were observed in 3 patients (equivalent to a 0.9% annualized rate). Two hemorrhages occurred from known residual postoperative AVMs and 1 from a recurrent AVM. Of note, none of these 3 patients underwent perioperative angiography. In all 3 cases, the patient was returned to the operating room and the residual or recurrent AVM was resected. Considering only the 82 initially angiographically confirmed and obliterated AVMs, 1 postoperative hemorrhage originated from a recurrent AVM, corresponding to an annualized hemorrhage rate of 0.3% over 288.3 patient-years of clinical follow-up.

Including both postoperative MRI and MR angiography in these 82 cases, a total of 212.1 patient-years of imaging follow-up (mean 2.6 years) was recorded, during which 2 radiographic recurrences of AVMs were observed (corresponding to a 0.9% annualized recurrence rate). In both cases, the recurrent AVM was resected. Of the 82 patients, 55 had repeated formal digital subtraction angiography 1–2 years after the surgery, with a total of 66.6 patient-years of formal angiographic follow-up over which 2 recurrences were found. Fourteen patients had MRI follow-up at least 5 years after surgery and no recurrences of AVMs were observed in these cases.

Discussion

Cerebral AVMs have been proposed to form as a result of the failure of primitive vascular channels to differentiate into arteries, interposed capillaries, or veins. Although their natural history has not been prospectively studied in pediatric cohorts, a recent retrospective analysis of 120 pediatric patients with AVMs reported an annual hemorrhage rate of 4%.5 Assuming that in our surgical cohort the AVMs were present since birth, the retrospective annual hemorrhage rate in this series was 5.5% (that is, 57 events during 1026 patient-years). However, because this rate was derived from a surgical cohort, this estimate was potentially biased toward an increased representation of more aggressive malformations. Nonetheless, these rates are similar to or slightly higher than those reported for adults.4,11,37

The overall mortality rate due to AVM rupture in children is as high as 21% in some series.18 Furthermore, AVMs may undergo dynamic changes in morphology and develop associated arterial and venous aneurysms and venous stenoses as the patient ages. Young women must also anticipate an elevated risk of hemorrhage during pregnancy,10 and all children with an AVM must bear the psychological burden of possessing a lesion that may cause neurological devastation or even death at any time in their life. Thus, regardless of rupture status, in our opinion all AVMs in children should be evaluated for potential treatment.

Patient Selection

Of the AVMs in our series, 80% were managed surgically. Our “surgery-first” approach is based on the surgical advantages of immediate therapeutic cure and high obliteration rates. This approach is in contradistinction to the unknown long-term impact of cranial radiosurgery in children, both in regard to obliteration rate and other longterm complications of this treatment. Immediate therapeutic management is particularly advantageous in cases of ruptured AVMs, as such patients face a greater risk of rerupture,4,11,37 particularly in the first year.12 In children presenting with clinical and/or radiographic signs of herniation from an acute hemorrhage, we prefer to evacuate the clot without attempting a resection of the AVM. If feasible, after the administration of mannitol, we perform a preoperative CT angiogram en route to the operating room to help delineate a safe route to the clot without disrupting the AVM. Formal digital subtraction angiography and AVM treatment are performed after evacuation of the clot along an individualized treatment course based on the child's recovery.

It is interesting to note that in our series, microsurgical resection of hemorrhagic AVMs was associated with significantly lower postoperative neurological complication rates as well as lower intraoperative blood loss than resection of nonhemorrhagic AVMs. The lower neurological complication rate after surgery for ruptured AVMs is consistent with previous studies13,23 and has in fact been incorporated in a recently published AVM grading scale that includes an older patient age, nonhemorrhagic presentation, and diffuse nidus morphology as significant risk factors for surgical morbidity.23 Importantly, although new neurological deficits were noted in 17% of the patients at follow-up in this series, the vast majority—over three-quarters (77%)—of these deficits were relatively minor and were anticipated visual field deficits after resection of AVMs. Unanticipated neurological complications were observed in 4 patients (4%)—2 having new weakness and seizures requiring antiepileptic medication, 1 having a new seizure disorder, and another having a new weakness. As expected, in our study, Spetzler-Martin grade36 did correlate with surgical morbidity; however, location in an eloquent brain region emerged independently as a statistically significant factor that had the greatest impact on postoperative complications, consistent with our experience with other vascular malformations.15

We do recommend radiosurgery for AVMs of appropriate size that are located in eloquent areas, such as the motor cortex, functional speech cortex, the basal ganglia, thalamus, and brainstem. It is intriguing that an extensive number of studies have reported using radiosurgery to manage AVMs in children.1,3,5,6,20,25,28,30,34,40 By comparison, reports of surgical AVM series, summarized in Table 7,2,5,8,9,16–18,21,22,27,31 are sparse and contain fewer patients, but some describe better results after resection of AVMs in children than in adults.31 Furthermore, as most pediatric AVMs in nearly all series are hemorrhagic,2,8,9,16–18,22 a more aggressive, expedient therapy via microsurgery, as opposed to radiosurgery, should be sought in such cases. It is interesting to note that in the most recent iteration of radiosurgical grading schemes, hemorrhage was considered an adverse factor for radiosurgical success,38 an assessment that differs from the observed impact of hemorrhage on the outcomes of microsurgery.23 We did not observe significant operative blood loss, a potential limitation of microsurgery relative to radiosurgery, and our obliteration rates were high. As underscored by our data and by those of previous series, immediate perioperative angiography appears crucial in achieving total AVM obliteration.2 Radiosurgical series may rely on noninvasive imaging modalities to reach conclusions about obliteration status of AVMs,3,25,34 and potentially delayed radiosurgical complications are of particular concern in pediatric patients.14,19,29,32

Although the management of unruptured AVMs has come under considerable scrutiny as a result of the ARUBA study (“A Randomized trial of Unruptured Brain Arteriovenous malformations”),26 the observations in that study have little relevance to pediatric patients, given that children were excluded from the ARUBA study and that the length of its clinical follow-up was shorter (the mean length of follow-up was < 3 years). Furthermore, although 76 of the 114 AVMs in the ARUBA interventional group (68%) were of Spetzler-Martin Grades I or II, only 5 were treated with microsurgery (4%) and 12 with microsurgery and embolization (11%). This makes it difficult to draw any conclusions from the ARUBA study about the impact of microsurgical resection on unruptured AVMs compared with that of observation only. Nonetheless, a review of the ARUBA study is relevant, as its data reveal an annual stroke rate of approximately 4% for untreated AVMs.

TABLE 7

Literature reports of surgical treatment for AVMs in children

Authors & YearNo. of PatientsObliteration Rate (%)Complication Rate (%)
Bristol et al. 200664*6519
Darsaut et al. 2011NS6710
Fong & Chan 198827NS15
Garza-Mercado et al. 1987128333
Hladky et al. 1994529019
Hoh et al. 200021955
Humphreys et al. 199612585NS
Kiri et al. 2005208910
Klimo et al. 2007357614
Nair et al. 20123610014
Sanchez-Mejia et al. 200632976
Current series949417

NS = not specified.

From a series of 82 children undergoing any treatment modality.

From a series of 120 children who underwent 104 surgical procedures; procedures included attempts at AVM obliteration and those aimed at only treating associated aneurysms.

The data reported here are important, as the natural history of untreated AVMs in children incorporates an annual hemorrhage rate of 4%–5.5%. These percentages contrast sharply with our postsurgery annual hemorrhage rate of 0.3%, representing a marked reduction in bleeding risk that is directly attributable to the surgical treatment. Consequently, the data of others, coupled with our experience reported here, support the practice of employing resection to improve the natural history and outcomes in children with intracranial AVMs.

Rationale for AVM Embolization

Although our results did not indicate greater obliteration rates and lower estimated blood loss after preoperative embolization, the absence of these improvements was likely an artifact of selection bias. We did not embolize all AVMs but rather only larger AVMs and those with deep arterial feeders. This selective embolization had the goal of reducing anticipated blood loss, of lowering the risk of perfusion pressure breakthrough, and of providing an Onyx-marked intraoperative guide to the lesion. Our surgical preference for Onyx stems not only from the roadmap it provides to feeding vessels at surgery, but also from the relative ease of coagulating and handling vessels embolized with Onyx as compared with N-butyl cyanoacrylate. In our cohort, 1 child had a postembolization hemorrhage (corresponding to a complication rate of 4.8%); however, the child recovered well and had no long-term neurological sequelae. A recent study of 25 children undergoing 38 procedures for Onyx embolization of AVMs reported an overall obliteration rate of 12%, with a procedural complication rate of 26.3% and no permanent neurological morbidity.35

Given the potential for long-term AVM recurrence in children, even if complete angiographically confirmed obliteration is serendipitously achieved, we would proceed to resection, wherever feasible, regardless of the postembolization angiographic appearance. Complete obliteration is not the primary goal of our embolization sessions. Our low embolization-related morbidity, coupled with our low total periprocedural morbidity and high AVM obliteration rates, highlights the potential advantages of adjunctive embolization used selectively in cases where an additional procedure provides a clear advantage.

Utility of Perioperative Angiography

In 2008, our institution developed a standardized protocol for the treatment of children with intracranial AVMs. An important component of this protocol is the inclusion of high-quality perioperative catheter angiography to confirm resection of the AVM prior to the conclusion of anesthesia. We observed no complications from the additional angiography. After the introduction of this protocol, the postoperative angiographically confirmed obliteration rates significantly improved from 86% to 100% (p = 0.01; Table 3). These results support the continued use of perioperative angiography. It is critical to emphasize that we use high-quality, biplane angiography, in contrast to the lower-quality portable units in the operating room. This markedly improves the resolution of the images and enhances the likelihood of identifying small residual AVMs.

Long-Term Results

Importantly, our series included both unruptured and ruptured AVMs with comparable results in both cohorts. Although AVMs that presented with hemorrhage did have lower volumes of operative blood loss and lower rates of perioperative neurological deficits (Tables 4 and 5), it is important to note that the presence or absence of a hemorrhage neither had a significant impact on the rate of surgical obliteration (Table 3) nor influenced the likelihood of long-term poor neurological outcomes (Table 6).

Considering a reported natural history of AVMs that includes a 4%–5.5% annual hemorrhage rate and an up to 20% mortality rate from each hemorrhage, we demonstrated that resection of AVMs confers a 0.3% annual hemorrhage rate and is associated with a 0.9% recurrence rate. Perioperative angiography significantly improved resection rates. Over a long period of follow-up (mean length 3.6 years), outcomes were excellent, with similar, durable 1- and 5-year rates of clinical independence (93% and 88%, respectively).

Conclusions

This series of pediatric patients with AVMs who underwent microsurgery demonstrates high obliteration rates (94% overall), acceptable rates of blood loss (mean blood loss 212 ml), and low rates of neurological complications, which were anticipated and manifested primarily as visual field deficits. Blood loss during surgery was lower for smaller and hemorrhagic AVMs. Surgeries on hemorrhagic AVMs and on AVMs in noneloquent locations were associated with lower rates of neurological complications. Long-term outcome, as measured by the mRS, was significantly influenced by the patient's mRS score at presentation. Rates of AVM obliteration improved as a result of a protocol modification mandating the use of perioperative angiography during the same anesthesia period. Surgical treatment markedly reduced the annual hemorrhage rate of AVMs compared with the rates reported from the natural history of untreated lesions. In aggregate, these data strongly support the use of resection in selected pediatric patients with intracranial AVMs.

Author Contributions

Conception and design: Gross, Scott, Smith. Acquisition of data: Gross, Storey. Analysis and interpretation of data: all authors. Drafting the article: Gross. Critically revising the article: Gross, Orbach, Scott, Smith. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Gross. Statistical analysis: Gross. Administrative/technical/material support: Scott, Smith. Study supervision: Scott, Smith.

References

  • 1

    Blamek SLarysz DMiszczyk L: Stereotactic linac radiosurgery and hypofractionated stereotactic radiotherapy for pediatric arteriovenous malformations of the brain: experiences of a single institution. Childs Nerv Syst 29:6516562013

  • 2

    Bristol REAlbuquerque FCSpetzler RFRekate HLMc-Dougall CGZambramski JM: Surgical management of arteriovenous malformations in children. J Neurosurg (2 Suppl) 105:88932006

  • 3

    Cohen-Gadol AAPollock BE: Radiosurgery for arteriovenous malformations in children. J Neurosurg 104 :6 Suppl3883912006

  • 4

    da Costa LWallace MCTer Brugge KGO'Kelly CWillinsky RATymianski M: The natural history and predictive features of hemorrhage from brain arteriovenous malformations. Stroke 40:1001052009

  • 5

    Darsaut TEGuzman RMarcellus MLEdwards MSTian LDo HM: Management of pediatric intracranial arteriovenous malformations: experience with multimodality therapy. Neurosurgery 69:5405562011

  • 6

    Dinca EBde Lacy PYianni JRowe JRadatz MWPreotiuc-Pietro D: Gamma knife surgery for pediatric arteriovenous malformations: a 25-year retrospective study. Clinical article. J Neurosurg Pediatr 10:4454502012

  • 7

    Ellis MJArmstrong DVachhrajani SKulkarni AVDirks PBDrake JM: Angioarchitectural features associated with hemorrhagic presentation in pediatric cerebral arteriovenous malformations. J Neurointerv Surg 5:1911952013

  • 8

    Fong DChan ST: Arteriovenous malformation in children. Childs Nerv Syst 4:1992031988

  • 9

    Garza-Mercado RCavazos ETamez-Montes D: Cerebral arteriovenous malformations in children and adolescents. Surg Neurol 27:1311401987

  • 10

    Gross BADu R: Hemorrhage from arteriovenous malformations during pregnancy. Neurosurgery 71:3493562012

  • 11

    Gross BADu R: Natural history of cerebral arteriovenous malformations: a meta-analysis. Clinical article. J Neurosurg 118:4374432013

  • 12

    Gross BADu R: Rate of re-bleeding of arteriovenous malformations in the first year after rupture. J Clin Neurosci 19:108710882012

  • 13

    Gross BADuckworth EAMGetch CCBendok BRBatjer HH: Challenging traditional beliefs: microsurgery for arteriovenous malformations of the basal ganglia and thalamus. Neurosurgery 63:3934112008

  • 14

    Gross BARopper AEDu R: Vascular complications of stereotactic radiosurgery for arteriovenous malformations. Clin Neurol Neurosurg 115:7137172013

  • 15

    Gross BASmith ERGoumnerova LProctor MRMadsen JRScott RM: Resection of supratentorial lobar cavernous malformations in children. Clinical article. J Neurosurg Pediatr 12:3673732013

  • 16

    Hladky JPLejeune JPBlond SPruvo JPDhellemmes P: Cerebral arteriovenous malformations in children: report on 62 cases. Childs Nerv Syst 10:3283331994

  • 17

    Hoh BLOgilvy CSButler WELoeffler JSPutman CMChapman PH: Multimodality treatment of nongalenic arteriovenous malformations in pediatric patients. Neurosurgery 47:3463582000

  • 18

    Humphreys RPHoffman HJDrake JMRutka JT: Choices in the 1990s for the management of pediatric cerebral arteriovenous malformations. Pediatr Neurosurg 25:2772851996

  • 19

    Izawa MChernov MHayashi MNakaya KKamikawa SKato K: Management and prognosis of cysts developed on long-term follow-up after Gamma Knife radiosurgery for intracranial arteriovenous malformations. Surg Neurol 68:4004062007

  • 20

    Kano HKondziolka DFlickinger JCYang HCFlannery TJAwan NR: Stereotactic radiosurgery for arteriovenous malformations, part 2: management of pediatric patients. Clinical article. J Neurosurg Pediatr 9:1102012

  • 21

    Kiriş TSencer ASahinbaş MSencer SImer MIzgi N: Surgical results in pediatric Spetzler-Martin grades I-III intracranial arteriovenous malformations. Childs Nerv Syst 21:69762005

  • 22

    Klimo P JrRao GBrockmeyer D: Pediatric arteriovenous malformations: a 15-year experience with an emphasis on residual and recurrent lesions. Childs Nerv Syst 23:31372007

  • 23

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

  • 24

    Maher COScott RM: Linear vein-based arteriovenous malformations in children. Clinical article. J Neurosurg Pediatr 4:12162009

  • 25

    Maity AShu HKTan JERuffer JSutton LNTochner Z: Treatment of pediatric intracranial arteriovenous malformations with linear-accelerator-based stereotactic radiosurgery: the University of Pennsylvania experience. Pediatr Neurosurg 40:2072142004

  • 26

    Mohr JPParides MKStapf CMoquete EMoy CSOverbey JR: Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicenter, non-blinded, randomised trial. Lancet 383:6146212014

  • 27

    Nair APKumar RMehrotra ASrivastava AKSahu RNNair P: Clinical, radiological profile and outcome in pediatric Spetzler-Martin grades I-III arteriovenous malformations. Childs Nerv Syst 28:5935982012

  • 28

    Pan DHKuo YHGuo WYChung WYWu HMLiu KD: Gamma Knife surgery for cerebral arteriovenous malformations in children: a 13-year experience. J Neurosurg Pediatr 1:2963042008

  • 29

    Pan HCSheehan JStroila MSteiner MSteiner L: Late cyst formation following gamma knife surgery of arteriovenous malformations. J Neurosurg 102 Suppl:1241272005

  • 30

    Reyns NBlond SGauvrit JYTouzet GCoche BPruvo JP: Role of radiosurgery in the management of cerebral arteriovenous malformations in the pediatric age group: data from a 100-patient series. Neurosurgery 60:2682762007

  • 31

    Sanchez-Mejia ROChennupati SKGupta NFullerton HYoung WLLawton MT: Superior outcomes in children compared with adult after microsurgical resection of brain arteriovenous malformations. J Neurosurg (2 Suppl) 105:82872006

  • 32

    Shin MKawahara NMaruyama KTago MUeki KKirino T: Risk of hemorrhage from an arteriovenous malformation confirmed to have been obliterated on angiography after stereotactic radiosurgery. J Neurosurg 102:8428462005

  • 33

    Smith ERButler WEOgilvy CS: Surgical approaches to vascular anomalies of the child's brain. Curr Opin Neurol 15:1651712002

  • 34

    Smyth MDSneed PKCiricillo SFEdwards MSWara WMLarson DA: Stereotactic radiosurgery for pediatric intracranial arteriovenous malformations: the University of California at San Francisco experience. J Neurosurg 97:48552002

  • 35

    Soltanolkotabi MSchoeneman SEAlden TDHurley MCAnsari SADiPatri AJ Jr: Onyx embolization of intracranial arteriovenous malformations in pediatric patients. Clinical article. J Neurosurg Pediatr 11:4314372013

  • 36

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

  • 37

    Stapf CMast HSciacca RRChoi JHKhaw AVConnolly ES: Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology 66:135013552006

  • 38

    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

  • 39

    van Swieten JCKoudstaal PJVisser MCSchouten HJvan Gijn J: Interobserver agreement for the assessment of handicap in stroke patients. Stroke 19:6046071988

  • 40

    Yeon JYShin HJKim JSHong SCLee JI: Clinico-radiological outcomes following gamma knife radiosurgery for pediatric arteriovenous malformations. Childs Nerv Syst 27:110911192011

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

Article Information

Correspondence Bradley A. Gross, Department of Neurological Surgery, Boston Children's Hospital and Harvard Medical School, 300 Longwood Ave., Boston, MA 02115. email: bgross1@partners.org.

INCLUDE WHEN CITING Published online October 31, 2014; DOI: 10.3171/2014.9.PEDS146.

DISCLOSURE The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. This study received support from the Stuart and Jane Weitzman Vascular Anomalies Center Fund.

© AANS, except where prohibited by US copyright law.

Headings

References

  • 1

    Blamek SLarysz DMiszczyk L: Stereotactic linac radiosurgery and hypofractionated stereotactic radiotherapy for pediatric arteriovenous malformations of the brain: experiences of a single institution. Childs Nerv Syst 29:6516562013

  • 2

    Bristol REAlbuquerque FCSpetzler RFRekate HLMc-Dougall CGZambramski JM: Surgical management of arteriovenous malformations in children. J Neurosurg (2 Suppl) 105:88932006

  • 3

    Cohen-Gadol AAPollock BE: Radiosurgery for arteriovenous malformations in children. J Neurosurg 104 :6 Suppl3883912006

  • 4

    da Costa LWallace MCTer Brugge KGO'Kelly CWillinsky RATymianski M: The natural history and predictive features of hemorrhage from brain arteriovenous malformations. Stroke 40:1001052009

  • 5

    Darsaut TEGuzman RMarcellus MLEdwards MSTian LDo HM: Management of pediatric intracranial arteriovenous malformations: experience with multimodality therapy. Neurosurgery 69:5405562011

  • 6

    Dinca EBde Lacy PYianni JRowe JRadatz MWPreotiuc-Pietro D: Gamma knife surgery for pediatric arteriovenous malformations: a 25-year retrospective study. Clinical article. J Neurosurg Pediatr 10:4454502012

  • 7

    Ellis MJArmstrong DVachhrajani SKulkarni AVDirks PBDrake JM: Angioarchitectural features associated with hemorrhagic presentation in pediatric cerebral arteriovenous malformations. J Neurointerv Surg 5:1911952013

  • 8

    Fong DChan ST: Arteriovenous malformation in children. Childs Nerv Syst 4:1992031988

  • 9

    Garza-Mercado RCavazos ETamez-Montes D: Cerebral arteriovenous malformations in children and adolescents. Surg Neurol 27:1311401987

  • 10

    Gross BADu R: Hemorrhage from arteriovenous malformations during pregnancy. Neurosurgery 71:3493562012

  • 11

    Gross BADu R: Natural history of cerebral arteriovenous malformations: a meta-analysis. Clinical article. J Neurosurg 118:4374432013

  • 12

    Gross BADu R: Rate of re-bleeding of arteriovenous malformations in the first year after rupture. J Clin Neurosci 19:108710882012

  • 13

    Gross BADuckworth EAMGetch CCBendok BRBatjer HH: Challenging traditional beliefs: microsurgery for arteriovenous malformations of the basal ganglia and thalamus. Neurosurgery 63:3934112008

  • 14

    Gross BARopper AEDu R: Vascular complications of stereotactic radiosurgery for arteriovenous malformations. Clin Neurol Neurosurg 115:7137172013

  • 15

    Gross BASmith ERGoumnerova LProctor MRMadsen JRScott RM: Resection of supratentorial lobar cavernous malformations in children. Clinical article. J Neurosurg Pediatr 12:3673732013

  • 16

    Hladky JPLejeune JPBlond SPruvo JPDhellemmes P: Cerebral arteriovenous malformations in children: report on 62 cases. Childs Nerv Syst 10:3283331994

  • 17

    Hoh BLOgilvy CSButler WELoeffler JSPutman CMChapman PH: Multimodality treatment of nongalenic arteriovenous malformations in pediatric patients. Neurosurgery 47:3463582000

  • 18

    Humphreys RPHoffman HJDrake JMRutka JT: Choices in the 1990s for the management of pediatric cerebral arteriovenous malformations. Pediatr Neurosurg 25:2772851996

  • 19

    Izawa MChernov MHayashi MNakaya KKamikawa SKato K: Management and prognosis of cysts developed on long-term follow-up after Gamma Knife radiosurgery for intracranial arteriovenous malformations. Surg Neurol 68:4004062007

  • 20

    Kano HKondziolka DFlickinger JCYang HCFlannery TJAwan NR: Stereotactic radiosurgery for arteriovenous malformations, part 2: management of pediatric patients. Clinical article. J Neurosurg Pediatr 9:1102012

  • 21

    Kiriş TSencer ASahinbaş MSencer SImer MIzgi N: Surgical results in pediatric Spetzler-Martin grades I-III intracranial arteriovenous malformations. Childs Nerv Syst 21:69762005

  • 22

    Klimo P JrRao GBrockmeyer D: Pediatric arteriovenous malformations: a 15-year experience with an emphasis on residual and recurrent lesions. Childs Nerv Syst 23:31372007

  • 23

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

  • 24

    Maher COScott RM: Linear vein-based arteriovenous malformations in children. Clinical article. J Neurosurg Pediatr 4:12162009

  • 25

    Maity AShu HKTan JERuffer JSutton LNTochner Z: Treatment of pediatric intracranial arteriovenous malformations with linear-accelerator-based stereotactic radiosurgery: the University of Pennsylvania experience. Pediatr Neurosurg 40:2072142004

  • 26

    Mohr JPParides MKStapf CMoquete EMoy CSOverbey JR: Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicenter, non-blinded, randomised trial. Lancet 383:6146212014

  • 27

    Nair APKumar RMehrotra ASrivastava AKSahu RNNair P: Clinical, radiological profile and outcome in pediatric Spetzler-Martin grades I-III arteriovenous malformations. Childs Nerv Syst 28:5935982012

  • 28

    Pan DHKuo YHGuo WYChung WYWu HMLiu KD: Gamma Knife surgery for cerebral arteriovenous malformations in children: a 13-year experience. J Neurosurg Pediatr 1:2963042008

  • 29

    Pan HCSheehan JStroila MSteiner MSteiner L: Late cyst formation following gamma knife surgery of arteriovenous malformations. J Neurosurg 102 Suppl:1241272005

  • 30

    Reyns NBlond SGauvrit JYTouzet GCoche BPruvo JP: Role of radiosurgery in the management of cerebral arteriovenous malformations in the pediatric age group: data from a 100-patient series. Neurosurgery 60:2682762007

  • 31

    Sanchez-Mejia ROChennupati SKGupta NFullerton HYoung WLLawton MT: Superior outcomes in children compared with adult after microsurgical resection of brain arteriovenous malformations. J Neurosurg (2 Suppl) 105:82872006

  • 32

    Shin MKawahara NMaruyama KTago MUeki KKirino T: Risk of hemorrhage from an arteriovenous malformation confirmed to have been obliterated on angiography after stereotactic radiosurgery. J Neurosurg 102:8428462005

  • 33

    Smith ERButler WEOgilvy CS: Surgical approaches to vascular anomalies of the child's brain. Curr Opin Neurol 15:1651712002

  • 34

    Smyth MDSneed PKCiricillo SFEdwards MSWara WMLarson DA: Stereotactic radiosurgery for pediatric intracranial arteriovenous malformations: the University of California at San Francisco experience. J Neurosurg 97:48552002

  • 35

    Soltanolkotabi MSchoeneman SEAlden TDHurley MCAnsari SADiPatri AJ Jr: Onyx embolization of intracranial arteriovenous malformations in pediatric patients. Clinical article. J Neurosurg Pediatr 11:4314372013

  • 36

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

  • 37

    Stapf CMast HSciacca RRChoi JHKhaw AVConnolly ES: Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology 66:135013552006

  • 38

    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

  • 39

    van Swieten JCKoudstaal PJVisser MCSchouten HJvan Gijn J: Interobserver agreement for the assessment of handicap in stroke patients. Stroke 19:6046071988

  • 40

    Yeon JYShin HJKim JSHong SCLee JI: Clinico-radiological outcomes following gamma knife radiosurgery for pediatric arteriovenous malformations. Childs Nerv Syst 27:110911192011

TrendMD

Cited By

Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 269 269 39
PDF Downloads 883 883 207
EPUB Downloads 0 0 0

PubMed

Google Scholar