You are here

Article Text

Article menu

Download PDFPDF

Hemorrhagic stroke
Original research
Acute embolization of ruptured brain arteriovenous malformations
  1. Andrew B Stemer1,
  2. William O Bank1,
  3. Rocco A Armonda2,
  4. Ai-Hsi Liu1,
  5. David W Herzig3,
  6. Randy S Bell2
  1. 1Department of Interventional Radiology, Washington Hospital Center, Washington, DC, USA
  2. 2National Capital Neurosurgery Consortium, Walter Reed Military Medical Center, Bethesda, Maryland, USA
  3. 3Department of Neurosurgery, Georgetown University Hospital, Washington, DC, USA
  1. Correspondence to Dr A B Stemer, Department of Interventional Radiology, Washington Hospital Center, 110 Irving Road NW, Washington, DC 20010, USA; astemer{at}gmail.com

Abstract

Background Ruptured brain arteriovenous malformations (bAVMs) are at increased risk of re-hemorrhage but management has historically been conservative. This is because: (1) ruptured bAVMs have not been considered as catastrophic as ruptured cerebral aneurysms, (2) surgical resection is aided by waiting for brain edema to resolve and clot to liquefy and (3) fear exists that partially treated bAVMs may be more dangerous than those untreated. The purpose of this study was to determine the feasibility and safety of acute embolization in patients with ruptured bAVMs.

Methods 21 consecutive patients who underwent acute embolization of ruptured bAVMs from 2007 to 2011 were retrospectively reviewed. All treatments consisted of embolization exclusively using Onyx liquid embolic agent. Neurological outcomes were assessed using the Glasgow Outcome Scale (GOS) at hospital discharge following initial treatment.

Results Nine patients (43%) were male, mean age was 38 years (range 8–75) and initial embolization was performed at a median of 4 days after ictus (mean 5.8, range 0–19). Spetzler–Martin grades ranged from I to VI (mean 3, median 3). In seven patients (33%) the AVM was completely occluded with a single treatment. With subsequent embolizations, complete occlusion occurred in four additional patients (19%). Six patients (29%) had post-embolization AVM resection and one patient (5%) had post-embolization gamma knife irradiation. None has suffered subsequent hemorrhages (mean follow-up 7.5 months). The mean discharge GOS of all patients was 4.4 (range 1–5).

Conclusion Treatment of ruptured bAVMs is often delayed but our experience with Onyx suggests that acute embolization is safe and feasible.

https://doi.org/10.1136/neurintsurg-2011-010214

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Brain arteriovenous malformations (bAVMs) initially present with intracranial hemorrhage in 30–50% of patients.1 ,2 Patient outcome after rupture of bAVMs varies between studies, but hemorrhage leads to death in 10–30% of patients and produces functional dependence after 2 years in 25% of survivors. 3 ,4 Historically, the neurosurgical approach has been conservative and treatment is often delayed for several weeks.5 This cautious approach has been widely accepted for the reasons outlined below.

    Re-rupture of bAVMs is thought to be extremely rare

    The natural history of bAVMs is poorly understood; their heterogeneity and relative rarity make large studies scarce. The specific factors causing initial bAVM rupture are debated, but unruptured bAVMs are crudely estimated to have a 2–4% chance of rupture per year.6 ,7 Morphological bAVM subgroups with associated aneurysms, deep venous drainage, deep location and venous outflow stenosis have been associated with higher risks of hemorrhage8 ,9 but a history of rupture is the most consistent predictor of re-hemorrhage.10 Once considered to be a rare event, re-hemorrhage can occur up to 33% in the first year.11 Similar to other neurovascular diseases such as subarachnoid hemorrhage and ischemic stroke, much of this risk may occur soon after ictus12 ,13 but acute re-hemorrhage rates have not been carefully estimated.

    Early surgical resection entails increased risks

    Acute neurosurgical resection of ruptured bAVMs is usually avoided because brain edema limits the ability to see the full size of the AVM and requires more extensive dissection, increased brain retraction and risk of perforation.14 Postponing surgery facilitates eventual resection by allowing time for the patient to recover, edema to resolve and clot to liquefy. No studies have justified this delay or compared neurologic outcomes following resections performed after different time intervals.

    Fear that partial occlusion is ineffective or harmful

    Endovascular treatments have recently achieved cure rates of over 50%15 but were previously not acutely performed because cure rates rarely exceeded 30%16 ,17and partial treatment was thought to be ineffective if not frankly harmful.18 Some reported that partially occluded bAVMs tended to re-hemorrhage but more recent evidence suggests that acute partial occlusion is beneficial.19

    The purpose of this study was to demonstrate that acute endovascular embolization of ruptured bAVMs is safe and feasible. Our goals for acute embolization were: (1) to identify rupture points when possible (ie, perinidal aneurysm or venous outflow stenosis), (2) to decrease the risk of re-hemorrhage in patients where acute re-hemorrhage risks are unknown and the usual treatment may be delayed more than is optimal and (3) to initiate bAVM treatment in conjunction with the diagnostic cerebral angiogram. Early treatment (surgical resection or endovascular embolization) for ruptured bAVMs with large hematomas or associated aneurysms has been reported previously.20 This is the first study to address the safety and feasibility of acute embolization in a consecutive series of ruptured brain bAVMs independent of specific risk factors.

    Methods

    This was a retrospective review of prospectively collected data. Between April 2007 and February 2011, 21 consecutive patients (15 adult and six pediatric) presented with acutely ruptured bAVMs and underwent acute endovascular embolization. All patients had a CT brain showing intracranial hemorrhage and a cerebral angiogram showing bAVM.

    After the diagnostic portion of the angiogram, images were reviewed to identify possible rupture points (such as perinidal aneurysms or venous outflow stenosis) that could be targeted during stage I embolization. The decision to pursue endovascular embolization (and subsequent bAVM management) was made collaboratively with the neurovascular team, including the interventionist, neurosurgeon and intensivist. If a rupture point was not evident, stage I embolization targeted the area with the maximum amount of high flow shunting to achieve the greatest reduction in flow through the malformation.

    All patients were embolized exclusively with ethylene vinyl alcohol copolymer (Onyx; ev3, Covidien Vascular Therapies, Irvine, California, USA). Onyx embolizations were performed using a slowly staged reflux technique. Onyx-34 was used initially and continued until either the embolization was complete or it was no longer effective, at which point it was followed with Onyx-18. Stage I embolizations were terminated when the most likely rupture points were felt to be occluded, when flow through the malformation was adequately reduced or completely obliterated, or when continued treatment was judged to be unsafe to the patient.

    The degree of occlusion achieved after each embolization was determined by the operator; 100% occlusion was determined only on follow-up angiography after review by three independent operators and the consulting neurosurgeon. All patients had postoperative CT scans and were admitted to the intensive care unit. Neurologic status was measured using the Glasgow Outcome Scale (GOS) on hospital arrival and hospital discharge after stage I embolization.

    Follow-up studies were tailored to each patient's specific needs. After surgical resection, patients underwent cerebral angiography within 7 days and again at approximately 3 months. Patients who required staged embolizations were typically embolized at approximately 4–6 week intervals. Those with complete endovascular occlusion of their bAVMs were followed with cerebral angiogram at 6–12 months. This study was approved by the institutional review board.

    Results

    The mean age of our patients was 38 years (range 8–75 years), 12 were women and nine were men. Spetzler–Martin grades21 ranged from 1 to 6 (mean 3, median 3). Patients underwent stage I embolization at a median of 4 days post-ictus (mean 5.8; range 0–19 days) (table 1) using 2.0 ml of Onyx (mean; range 0.7–5.5 ml) injected over 38 min (mean; range 10–94). The shortest embolization (10 min) was an outlier as this procedure was prematurely terminated after the perinidal aneurysm was occluded but before the malformation was further occluded. Excluding this embolization the mean injection time was 40 min (range 22–94). Seven patients had perinidal aneurysms, two had venous outflow stenosis and one had cortical venous reflux. Five patients underwent pre-embolization hematoma evacuation, six underwent post-embolization surgical resection (mean 5.3 days post embolization, range 1–14), and one received post-embolization radiotherapy (table 2).

    View this table:
    Table 1

    Demographics, characteristics and time to treatment of 21 patients and their arteriovenous malformations

    View this table:
    Table 2

    Pre and post embolization interventions

    Between 1 and 5 embolization stages were performed (mean 1.7) although most (13 of 21, 62%) required only a single stage before the bAVM was completely occluded (figure 1) or the residual malformation was surgically resected or irradiated. In seven patients (33%) complete occlusion occurred after stage I embolization alone. These malformations tended to be smaller (two or three) and a mean of 1.0 ml of Onyx was used.

    Figure 1
    Figure 1

    Images of an 11-year-old patient who presented with severe headache. Preoperative CT brain (A) showed a right occipital hematoma. Cerebral angiogram (B) on post-bleed day 4 showed a Spetzler–Martin grade II arteriovenous malformation (AVM) supplied by branches of the right posterior cerebral artery. Embolization was performed on the same date and complete occlusion achieved with 1.3 ml of Onyx. (C) Unsubtracted view of Onyx cast. (D) Fourteen month follow-up angiogram confirmed complete AVM occlusion.

    Eight patients had more than one stage of embolization and these were typically performed 4–6 weeks after the previous stage. With subsequent embolizations, complete occlusion occurred in four additional patients (19%) (table 3). Follow-up embolizations used a mean of 3.4 ml of Onyx per embolization and patients who achieved complete endovascular occlusion with multiple stages had a mean total of 7.8 ml of Onyx.

    View this table:
    Table 3

    Embolization results

    Eighteen of 21 patients (86%) had follow-up angiograms (mean 7.5 months post-embolization) although patients who underwent post-embolization resection had their follow-up angiograms sooner (mean 2.5 months) than patients without resection (mean 10 months). One patient died of unrelated causes before follow-up angiography was scheduled, one patient was lost to follow-up and one patient refused follow-up angiography. No patients suffered re-hemorrhage after the initial embolization.

    Morbidity and mortality

    Mean GOS on initial presentation was 4.0 (nine patients had a GOS of 5) and mean GOS on hospital discharge after stage I embolization was 4.4 (14 patients had a GOS of 5). Fifteen patients (71%) were discharged directly home, three patients to acute rehabilitation facilities, two patients to long term care facilities and one patient had care withdrawn in the hospital.

    Of 12 patients who presented with a GOS <5, seven remained with a GOS <5 at discharge. Of these seven, three were discharged with an improved GOS, two had an unchanged GOS compared with admission and two were discharged with a lower GOS. Two patients had a history of bAVM rupture (including one with an initial hemorrhage 3 weeks earlier) and five had bAVMs located in eloquent locations (four cerebellar).

    Twelve month follow-up examinations revealed that three patients had improved to a GOS of 5, two patients had unchanged neurologic deficits, one patient died of unrelated causes at a long term care facility and one patient had care withdrawn in the hospital (table 4).

    View this table:
    Table 4

    Characteristics of patients with a discharge Glasgow Outcome Scale <5

    Complications

    Two asymptomatic technical complications occurred during stage I embolizations (a right posterior cerebral artery occlusion without infarction, a retained microcatheter in the left vertebral artery) and two other patients died. Of the patients who died, the first was a 61-year-old woman who presented with a re-hemorrhage of her grade III cerebellar AVM 3 weeks after initial hemorrhage. On arrival, her GOS was 3, pre-embolization hematoma evacuation was performed and followed by stage I embolization with complete occlusion of the AVM. Her neurologic status did not change and eventually tracheostomy and percutaneous endoscopic gastrostromy tubes were placed prior to her discharge to a long term care facility (GOS 3). She died 5 months later in the long term care facility of unrelated causes.

    The second patient who died was a 64-year-old woman whose bAVM was diagnosed during an epilepsy evaluation 40 years prior to rupture. She was informed by physicians at different institutions over several decades that her bAVM was inoperable (grade VI). Her epilepsy was well controlled and the AVM otherwise asymptomatic until she became acutely unresponsive. CT brain showed a massive right hemispheric hemorrhage with ballooning of the Sylvian fissure, hemorrhagic extension into the lateral and third ventricles, and midline shift. On initial examination she intermittently followed commands with her right side (GOS 3) but after ventriculostomy placement and emergency hematoma evacuation she was only able to localize with her right side. The following day, stage 1 embolization was performed and 5.5 ml of Onyx were injected through the proximal right posterior cerebral artery branches. Post-embolization, her neurologic status did not change. No change in her hemorrhage was found on CT brain. Her hospital course was complicated by fever and after discussion with the family, care was withdrawn on post-embolization day 11.

    Discussion

    Timing and treatment of ruptured bAVMs remains controversial.22 In contrast to the treatment of ruptured brain aneurysms exclusive of bAVMs,2 the management of ruptured bAVMs is usually conservative with delay of intervention.14 In the 1960s and 1970s, surgical clipping of ruptured aneurysms was also delayed for several weeks. As the physiology and complications of subarachnoid hemorrhage became better understood and technology and techniques improved, early surgical clipping became the standard of care.23 Although early endovascular treatment has been recently advocated when bAVMs and aneurysms are associated,15 ,24 a similar reticence for early intervention with ruptured bAVMs still applies. In light of improved occlusion rates, Onyx embolization may enable safe and definitive endovascular intervention in the acute phase and an expedited approach without bias toward the underlying etiology may be warranted.15

    Ruptured bAVMs cause significant morbidity and mortality2 ,4 and re-rupture at unknown rates. Our goal for acute embolization was to expeditiously initiate treatment with the aim of eliminating re-hemorrhage. Because the specific triggers leading to bAVM rupture are often unclear, rupture points were treated when identified, but acute stage I embolizations were also performed in their absence.

    Achieving complete occlusion during stage I embolization was not a prespecified goal; it was not expected that we would achieve complete occlusion after stage I embolizations in one-third of patients. Overall, complete occlusion occurred in 11 patients (52.4%) with Onyx alone.

    Most bAVM embolization studies with Onyx report complete occlusion in 10–30% of patients16 ,17 although 51% was recently achieved in a large series.15 Saatci et al achieved astonishing results in patients with Spetzler–Martin grades I and II bAVMs where 98% had endovascular cures, although Spetzler–Martin grades III and IV reached the same result in only 12.5%. Some of the embolizations were performed acutely but the number, precise timing and outcomes for this subgroup were not reported.15 In our patients with Spetzler–Martin grades I and II bAVMs, we less frequently achieved complete occlusion (4/7 cases; 57%) but in patients with Spetzler–Martin grades III and IV, we achieved complete occlusion more frequently (54% with complete occlusion and 15% with near complete occlusion). Saatci et al injected 2.4–18 ml of Onyx in a single pedicle15; we injected 0.7–5.5 ml of Onyx (mean 2.0 ml) in our stage I embolizations. It should be noted that Saatci et al used a definition of ‘definitive treatment status for embolization’ that included many things, as they reiterate in the editorial response to comments on their paper: “it indicates that the embolization treatment has reached its final result and that no additional embolization session is being planned (whether because of total obliteration, effective decrease in size or the decision to discontinue is made due to ineffective embolization)”.25 Our definition of occlusion entails complete angiographic obliteration of the bAVM on follow-up angiography.

    Case series of bAVM embolizations with Onyx show morbidity between 7% and 20% and mortality in 1–3%16 ,17 although it is difficult to compare studies that include both ruptured and unruptured bAVMs with our study in which all bAVMs were ruptured. In our series, 12 patients (57.1%) presented with a GOS <5, of whom seven (33.3%) were discharged after initial embolization with a GOS remaining <5. Of those seven patients, three had an improved GOS compared with admission, two were unchanged and two were worse. Two patients suffered asymptomatic technical complications during stage I embolization (9.5%). Two patients (9.5%) included in the study died; one had care withdrawn in the hospital and one died 5 months after embolization.

    Early resection for ruptured bAVMs has been reported in a few small series4 ,26 with most of those surgeries performed because of large hematomas and rapid clinical deterioration.27 Two studies of early surgery following ruptured bAVMs involved 27 and 49 patients with 62.9% and 55%, respectively, achieving good functional outcomes.4 ,26 Lawton et al reviewed ruptured bAVMs in 120 patients who underwent resection (time from ictus not documented); 53.3% made a complete or nearly complete recovery (GOS 5) and 37.5% remained moderately or severely disabled (GOS 3–4).28 In comparison with our study, 66.6% of patients were discharged with GOS 5 and on follow-up this number increased to 80.9%; 9.5% remained disabled (GOS 3–4) on follow-up and 9.5% died.

    Our primary goal for acute embolization was to prevent re-hemorrhage and, to date, re-hemorrhage has not occurred in any patient (mean follow-up 7.5 months). A direct safety comparison with other studies of acute endovascular embolization is not possible. Compared with studies of the natural history of ruptured bAVMs, acute surgery for ruptured bAVMs or embolization for ruptured or unruptured bAVMs, our results compare favorably (occlusion rates, complications rates, neurologic outcomes) although the number of patients included in this study is too small to make definitive conclusions. The optimal timing for treatment of ruptured bAVMs is not yet established and embolization is always a potentially dangerous intervention but our initial results suggest that acute endovascular embolization without bias to underlying etiology is feasible and can be performed safely.

    Previously, partial embolization has been cautioned against because of the concern that incompletely treated bAVMs tended to re-hemorrhage.18 A large retrospective long term survival study recently concluded that any treatment (partial or complete) may enhance survival.19 This may be due to the benefits from occluding high risk aneurysms, altering flow dynamics to decrease the risk of rebleeding, facilitating neurosurgical resection or promoting eventual complete occlusion. Fourteen patients (66.6%) in our series achieved only partial occlusion after stage I embolization; none suffered re-hemorrhage.

    A possible limitation of our study is that incomplete embolization can be caused by unseen bAVM compartments.29 Hidden compartments may not be identified despite careful evaluation, a problem that could be aggravated by hematoma and brain edema. The incidence of undetected hidden compartments during angiography is rare, although recurrence of lesions can also be seen after total surgical resection or embolization despite angiographic confirmation.15 ,29 ,30 Nonetheless, we found excellent correlation in all patients who underwent subsequent bAVM resection and in all patients in whom follow-up angiograms confirmed the initial findings. Follow-up angiograms were performed at a mean of 7.5 months after embolization, and a mean of 10.5 months in patients not undergoing resection.

    The most important limitations of our study were the small number of patients and relatively short follow-up period. Of 21 patients whose bAVMs were acutely embolized, six underwent subsequent surgical resection, two died and one was lost to follow-up. The remaining 12 patients were too few to draw definitive conclusions. Our mean follow-up period of 7.5 months adequately excludes acute re-hemorrhage but if long-term rebleeding occurs at 2–4% per year, this would likely not be seen in this time period.

    Several other limitations were present. An inherent selection bias exists as it was a retrospective observational study performed at a tertiary care center. Pediatric patients are different than adults but in this study we considered them together. The GOS, while commonly used to report surgical results, is not as sensitive as the modified Rankin Scale but this was not collected prospectively. Furthermore, our patients were treated by highly skilled and experienced operators. No direct causality can be proven between treatments and outcomes, and it remains to be seen if these results can be replicated by different operators. The safety, efficacy and impact on the natural history of ruptured bAVMs after acute embolization can only be evaluated in a larger controlled study.

    Conclusion

    Treatment of ruptured bAVMs is often delayed but new experience with Onyx shows that acute embolization is feasible. The optimal time to treat ruptured bAVMs is unknown but our results suggest acute embolization is safe and an expedited endovascular approach may be warranted.

    References

    1. ↵
      1. Fullerton HJ,
      2. Achrol AS,
      3. Johnston SC,
      4. et al
      . Long-term hemorrhage risk in children versus adults with brain arteriovenous malformations. Stroke 2005;36:2099–104.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Crawford PM,
      2. West CR,
      3. Chadwick DW,
      4. et al
      . Arteriovenous malformations of the brain: natural history in unoperated patients. J Neurol Neurosurg Psychiatry 1986;49:1–10.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. van Beijnum J,
      2. Lovelock CE,
      3. Cordonnier C,
      4. et al
      . Outcome after spontaneous and arteriovenous malformation-related intracerebral haemorrhage: population-based studies. Brain 2009;132:537–43.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Pavesi G,
      2. Rustemi O,
      3. Berlucchi S,
      4. et al
      . Acute surgical removal of low-grade (Spetzler–Martin I–II) bleeding arteriovenous malformations. Surg Neurol 2009;72:662–7.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Perret G,
      2. Nishioka H
      . Report on the cooperative study of intracranial aneurysms and subarachnoid hemorrhage. Section VI. Arteriovenous malformations. An analysis of 545 cases of cranio-cerebral arteriovenous malformations and fistulae reported to the cooperative study. J Neurosurg 1966;25:467–90.
      OpenUrlPubMedWeb of Science
    6. ↵
      1. Forster DM,
      2. Steiner L,
      3. Hakanson S
      . Arteriovenous malformations of the brain. A long-term clinical study. J Neurosurg 1972;37:562–70.
      OpenUrlPubMedWeb of Science
    7. ↵
      1. Ondra SL,
      2. Troupp H,
      3. George ED,
      4. et al
      . The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg 1990;73:387–91.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Stapf C,
      2. Mast H,
      3. Sciacca RR,
      4. et al
      . Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology 2006;66:1350–5.
      OpenUrlCrossRef
    9. ↵
      1. Hademenos GJ,
      2. Massoud TF
      . Risk of intracranial arteriovenous malformation rupture due to venous drainage impairment. A theoretical analysis. Stroke 1996;27:1072–83.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Hernesniemi JA,
      2. Dashti R,
      3. Juvela S,
      4. et al
      . Natural history of brain arteriovenous malformations: a long-term follow-up study of risk of hemorrhage in 238 patients. Neurosurgery 2008;63:823–9.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Mast H,
      2. Young WL,
      3. Koennecke HC,
      4. et al
      . Risk of spontaneous haemorrhage after diagnosis of cerebral arteriovenous malformation. Lancet 1997;350:1065–8.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Inagawa T,
      2. Kamiya K,
      3. Ogasawara H,
      4. et al
      . Rebleeding of ruptured intracranial aneurysms in the acute stage. Surg Neurol 1987;28:93–9.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Giles MF,
      2. Rothwell PM
      . Risk of stroke early after transient ischaemic attack: a systematic review and meta-analysis. Lancet Neurol 2007;6:1063–72.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Luessenhop AJ,
      2. Rosa L
      . Cerebral arteriovenous malformations. Indications for and results of surgery, and the role of intravascular techniques. J Neurosurg 1984;60:14–22.
      OpenUrlPubMedWeb of Science
    15. ↵
      1. Saatci I,
      2. Geyik S,
      3. Yavuz K,
      4. et al
      . Endovascular treatment of brain arteriovenous malformations with prolonged intranidal Onyx injection technique: long-term results in 350 consecutive patients with completed endovascular treatment course. J Neurosurg 2011;115:78–88.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Weber W,
      2. Kis B,
      3. Siekmann R,
      4. et al
      . Preoperative embolization of intracranial arteriovenous malformations with Onyx. Neurosurgery 2007;61:244–52.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Panagiotopoulos V,
      2. Gizewski E,
      3. Asgari S,
      4. et al
      . Embolization of intracranial arteriovenous malformations with ethylene-vinyl alcohol copolymer (Onyx). AJNR Am J Neuroradiol 2009;30:99–106.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Miyamoto S,
      2. Hashimoto N,
      3. Nagata I,
      4. et al
      . Posttreatment sequelae of palliatively treated cerebral arteriovenous malformations. Neurosurgery 2000;46:589–94.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Laakso A,
      2. Dashti R,
      3. Seppanen J,
      4. et al
      . Long-term excess mortality in 623 patients with brain arteriovenous malformations. Neurosurgery 2008;63:244–53.
      OpenUrlPubMed
    20. ↵
      1. Halim AX,
      2. Johnston SC,
      3. Singh V,
      4. et al
      . Longitudinal risk of intracranial hemorrhage in patients with arteriovenous malformation of the brain within a defined population. Stroke 2004;35:1697–702.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Spetzler RF,
      2. Martin NA
      . A proposed grading system for arteriovenous malformations. J Neurosurg 1986;65:476–83.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Ogilvy CS,
      2. Stieg PE,
      3. Awad I,
      4. et al
      . AHA scientific statement: recommendations for the management of intracranial arteriovenous malformations: a statement for healthcare professionals from a special writing group of the Stroke Council, American Stroke Association. Stroke 2001;32:1458–71.
      OpenUrlFREE Full Text
    23. ↵
      1. Bederson JB,
      2. Connolly ES Jr,
      3. Batjer HH,
      4. et al
      . Guidelines for the management of aneurysmal subarachnoid hemorrhage: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke 2009;40:994–1025.
      OpenUrlFREE Full Text
    24. ↵
      1. Meisel HJ,
      2. Mansmann U,
      3. Alvarez H,
      4. et al
      . Cerebral arteriovenous malformations and associated aneurysms: analysis of 305 cases from a series of 662 patients. Neurosurgery 2000;46:793–800.
      OpenUrlPubMedWeb of Science
    25. ↵
      1. Elhammady MS,
      2. Heros RC
      . Onyx and arteriovenous malformations. J Neurosurg 2011;115:75–6.
      OpenUrlPubMedWeb of Science
    26. ↵
      1. Kuhmonen J,
      2. Piippo A,
      3. Vaart K,
      4. et al
      . Early surgery for ruptured cerebral arteriovenous malformations. Acta Neurochir Suppl 2005;94:111–14.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Jafar JJ,
      2. Rezai AR
      . Acute surgical management of intracranial arteriovenous malformations. Neurosurgery 1994;34:8–12.
      OpenUrlPubMedWeb of Science
    28. ↵
      1. Lawton MT,
      2. Du R,
      3. Tran MN,
      4. et al
      . Effect of presenting hemorrhage on outcome after microsurgical resection of brain arteriovenous malformations. Neurosurgery 2005;56:485–93.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Pellettieri L,
      2. Svendsen P,
      3. Wikholm G,
      4. et al
      . Hidden compartments in AVMs—a new concept. Acta Radiol 1997;38:2–7.
      OpenUrlFREE Full Text
    30. ↵
      1. Kader A,
      2. Goodrich JT,
      3. Sonstein WJ,
      4. et al
      . Recurrent cerebral arteriovenous malformations after negative postoperative angiograms. J Neurosurg 1996;85:14–18.
      OpenUrlCrossRefPubMedWeb of Science

    Footnotes

    • Competing interests None.

    • Ethics approval This study was approved by the institutional review board.

    • Provenance and peer review Not commissioned; externally peer reviewed.