Summary
The most important issue when dealing with a patient with a brain AVM is the decision whether to treat or not. Only after this decision has been made, taking into consideration a number of factors depending on both the patient and the specific type of AVM, can the best option for treatment be chosen.
An operative classification of brain AVMs, previously adopted in the Department of Neuroradiology and Neurosurgery of Verona (Italy) and published in this journal, was subjected to validation in a consecutive group of 104 patients clinically followed for at least three years after completion of treatment.
This classification, slightly modified from the original version concerning the importance of some specific items, allowed us to assess the indication to treat in each case, whatever type of treatment was offered to the patient.
Key Words: arteriovenous malformations, brain, treatment
Introduction
Brain arteriovenous malformations (AVMs) are complex pathologies. Neither the real pathophysiology nor the risk associated with these intriguing lesions have yet been fully clarified1-3 so that not infrequently the correct therapeutic management for an individual patient may be cumbersome4.
The decision to treat a patient or not and, secondly, to select the best options for treatment must take a number of factors into consideration depending on both the patient and the specific type of AVM.
Thanks to the development of modern catheter angiography and MR techniques, over the years a number of classification systems have been proposed and refined5,6.
The different classifications proposed have considered various angio-architectural features such as size, site, number and distribution of afferent vessels, patterns of venous drainage, flow velocity and the amount of blood steal to the adjacent normal brain tissue in order to obtain a possible association with the bleeding risk7-14. Nevertheless, the available classifications are so different that comparisons between various series are extremely difficult.
Most relevant papers dealing with predictors of unfavourable outcome identified separately several factors for surgery15-17, radiosurgery18-21 or embolization22-26. In the case of endovascular treatment adverse factors linked to neurological complications were identified by means of clinical characteristics such as age or the absence of pre-treatment neurological deficits 22 or morphological AVM characteristics such as deep venous drainage and the size of the AVM25, basal ganglia location 24, eloquent areas or fistula 23.
Other adverse factors identified such as periprocedural hemorrhage 25, venous deposition of glue23 or number of embolizations22, do not appear useful for pre-treatment identification of possible risks.
Moreover, in literature morbi/mortality rates deal exclusively with a single type of treatment, scarcely considering the overall rate on completion (i.e. obliteration of the AVM) of combined treatments.
Relying on the data available in literature and on the opinions gained from our personal experience, we developed an operative classification of brain AVMs published in this journal27 with the main aim of helping the physician and the patient to reach informed consent on the treatability of his/her AVM. This classification was mainly based on a thorough review of the literature and partly on the authors' personal experience, but did not include a review of their own AVM cases.
The main purpose of the present study was to validate and/or refine our previous classification on the basis of the retrospective analysis of 104 patients treated in our institution.
Our Department has developed and utilized a cumulative score (CS) made up of the sum of an intention to treat score (ITS) and a treatment risk score (TRS)27,28. Even though these scores combine morphological AVM features with clinical/physiological factors, their use can be justified by the achievement of a single score, the "core" of the process of decisionmaking for the single patient (treatment or abstention, type of treatment). Lower ITS scores indicate the need for treatment due to an unfavourable natural history, while lower TRS scores indicate lower risk levels for each type of treatment.
In our previous classification, ITS was made up of the sum of scores deriving from both patient and AVM characteristics. Patient characteristics and relative score were: age (≥65: 2; ≥50 <65:1; <50: 0); history of hemorrhage (no: 2; yes: 0); neurological deficits not related to a previous hemorrhage (no: 1; yes: 0); patient's firm intention to be treated, important from a psychological point of view (no: 1; yes: 0).
AVM characteristics were: small size, i.e. volume less than 10 ml (no: 1; yes: 0); deep brain location (no: 1; yes: 0); exclusive deep venous drainage (no: 2; yes: 0); associated aneurysms or varix (no: 2; yes: 0). By totalling the patient score (0-6) and the AVM characteristics score (0-6), ITS was calculated ranging from 0 to 12.
For each brain AVM to be treated, TRS ranging from 1 to 5 was calculated, irrespective of the choice of surgery, radiosurgery, embolization or combined treatment. In our Department the decision on which option would be most suitable for the patient is made after assessment by all members of the AVM team. In addition, the patient's preference, which is of no less importance, is taken into consideration.
The most popular grading scheme for predicting surgical risk is the system described by Spetzler and Martin in 198616 which divides patients into five risk categories on the basis of three AVM features: size (1-3), eloquence of location (0,1) and pattern of venous drainage (0,1). The grade of any particular lesion is the sum of its score for each of these three characteristics, ranging from grade I (simplest and lowest risk) to grade V (most complex and highest risk). Using this classification, permanent morbidity and mortality ranged from 0 for patients with grade I AVMs to 17% for grade V AVMs15. As in the grading scheme of Spetzler and Martin, by far the most popular and widely accepted, two other grading schemes scoring 1 to 5 both for radiosurgery and embolization are used respectively in our institution.
The main feature in predicting radiosurgical risk both for radionecrosis and/or failure of treatment is the volume of the AVM nidus. On the other hand, in our experience based on more than 250 treated AVMs18,19, reduced nidal flow (both spontaneous or decreased by embolization) is associated with a higher post-RS obliteration rate, also occurring in a shorter time. After analyzing the results of our personal experience, we assigned a higher score to the larger AVMs (<5 cm3:1; >5 <10 cm3: 2; >10 <20 cm3: 3; >20 <30 cm3: 4; >30 cm3: 5). One point is subtracted if the AVM has a low flow (for example: a 13 cm3 AVM is scored 2 - not 3 - if it is a low-flow AVM).
In our experience 28, the three main factors predicting embolization risk were: volume (<10 cm3: 1; > 10 < 20 cm3: 2; >20 cm3: 3), eloquence (not eloquent: 0; eloquent: 1) and the presence of perforators as feeders (no perforators: 0; perforators: 1).
The cumulative score (CS) was calculated according to the total ITS (0-12) and TRS (1-5) scores-According to this procedure, with a score from 1 to 10, treatment was highly recommended, with a score from 11 to 12 treatment was offered but with a significant risk and with a score from 13 to 17 no treatment at all was advised.
Materials and Methods
104 patients with brain AVMs treated in the Department of Neuroradiology and Neurosurgery of Verona University Hospital from 1998 to 2005 were retrospectively analyzed.
The main goal was to verify - and/or improve - the predictive potential of our proposed classification 27 which considered both the specific clinical features of a patient together with an individual AVM and the risk associated with the specific morphology of that AVM. After weighing up these factors, we devised a classification system yielding an individual score, tailored for every single patient.
This approach was based mainly on a thorough review of the literature and partly on personal experience derived from our opinions built up over more than 15 years of daily involvement in the field of brain AVMs.
The 104 patients included in the study represent a subgroup of the whole series of patients treated in the same period, fulfilling precise criteria of homogeneity for the following reasons:
- all patients had complete clinical/radiological documentation
- all patients had an adequate clinical/radiological follow-up of at least three years after treatment (mean four years)
- all patients were treated in a similar manner (same neurosurgical/endovascular/ radiosurgical team; same devices; same materials). Morphological characteristics of all AVMs were evaluated by means of MR and angiographic data, and in particular the precise topography of each lesion was assessed on MR scans while the exact volume was calculated on angiographic films taking into account the magnification factor, using the system described by Pasqualin et Al.29.
A modified ranking outcome score (ROS) was used30 for the precise evaluation of neurological disability (Table 1): compared with the original version characterized by a score 0 to 5 according to increasing severity, the modified version includes the score 6 in the event of patient death.
Table 1.
| Modified Rankin Outcome Scale TOTAL (0-6) | |
|---|---|
| Score | Description |
| 0 | No symptoms at all |
| 1 | No significant disability despite symptoms; able to carry out all usual duties and activities |
| 2 | Slight disability; unable to carry out all previous activities, but able to look after own affairs without assistance |
| 3 | Moderate disability; requiring some help, but able to walk without assistance |
| 4 | Moderately severe disability; unable to walk without assistance and unable to attend to own bodily needs without assistance |
| 5 | Severe disability; bedridden, incontinent and requiring constant nursing care and attention |
| 6 | Dead |
|
ROS 0-2 are considered Independent, ROS 3-6 are considered disabled. | |
Patients were classified as non autonomous when ROS ranged from 3 to 6 (score 6, death, was included); they were classified as autonomous (ROS ≤2) when they were able to work and to carry out a majority of tasks.
Results
In the 104 patients treated, 29 were treated by direct radiosurgery, 27 by direct surgery, four by embolization alone and 44 by combined treatment in which embolization was followed by surgery (24), radiosurgery (15), or both (5) (Table 2).
Table 2.
Brain AVMs treated in the Department of Neurosurgery of Verona (period 1998-2005)
| Treatment Modality | N° of Patients |
Percentage of Total |
|||
|---|---|---|---|---|---|
| Direct surgery | 27 | 26% | |||
| Direct radiosurgery | 29 | 28% | |||
| Embolization | 4 | 3.5% | |||
| Embolization + radiosurgery | 15 | 42.5% | |||
| Embolization + surgery | 24 | ||||
| Embolization + radiosurgery + surgery |
5 | ||||
| Total | 104 | 100% | |||
In the 104 treated patients complete obliteration of the lesion was achieved in 103 (99%). In one patient treated by radiosurgery the AVM was found to be incompletely obliterated (70%) at three year follow-up.
In the group of 104, 16 patients with a ROS score 3-6 were distributed as follows: seven patients with a ROS score of 3, eight patients with ROS 4 and one patient with ROS 6 (deceased). Therefore global mortality was 0.9 % and global morbidity was 14% (Table 3).
Table 3.
Mortality and morbidity rates according to type of treatment.
| Direct surgery |
Direct radiosurgery |
Embolization + Combined treat |
Total | |
|---|---|---|---|---|
| Mortality (ROS 6) | 1/27 (3.7%) | 0/29 | 0/48 | 1/104 (0.9%) |
| Morbidity (ROS 3-5) | 5/27 (18.5%) | 4/29(14%) | 6/48 (12.5%) | 15/104 (14%) |
As regards distribution according to the type of treatment (Table 3), direct surgery resulted in a mortality rate of 3.7% (1/27 patients) and a 18.5% morbidity rate (5/27 patients). The deceased patient was the oldest of the entire series, was admitted to hospital with a bleeding small left temporal AVM and the cause of death must be attributed to pulmonary embolism. All five of the non autonomous patients treated by direct surgery were admitted to hospital with intracerebral hematoma and their clinical condition was poor.
Radiosurgery, performed in small volume AVMs (<5 ml), did not result in any mortality, with a morbidity rate of 14% (4/29 patients).
Embolization alone or followed by surgery or radiosurgery did not result in any fatality and presented a morbidity rate of 12.5% (6/48 patients).
To sum up, while the only death was due to surgery, as far as mortality rates are concerned there was no statistical difference between the various types of treatments.
Each item of the intention to treat score was analyzed on the basis of the recorded data (Table 4):
Table 4.
Distribution of Cumulative Score (CS) among 104 treated patients.
| Cumulative Score |
PATIENTS | |||
|---|---|---|---|---|
| Total | Surgery | Radio- surgery |
Embo & Combined |
|
| 1 - 10 Treatment Recommended |
99 | 27 | 29 | 43 |
| 11 - 12 Treatment Offered With Significant Risk |
5 | - | - | 5 |
| 13 - 17 Treatment Not Recommended |
0 | - | - | - |
- AGE was a predictor of unfavourable outcome by univariate analysis: older patients (>40 yrs) were more likely to show significant neurological deficits (ROS ≥3; Graph I).
Younger patients (<40 yrs) had a more favourable outcome (63/88, 72%, were autonomous patients vs 5/16,31%, non autonomous patients) than older patients (25/88, 28%, were autonomous vs 11/16,63%, non autonomous).
We concluded that the division into three subgroups adopted in the previous version of the classification works better when the three subgroups are distributed as follows: ≥60 yrs, ≥ 40 <60 yrs, <40 yrs, as a higher statistical significance (p <0,001) is reached.
- Previous hemorrhage was a predictor of unfavourable outcome by univariate analysis (p <0,001): hemorrhages were recorded in 13/16, 81% of the non autonomous (ROS ?3) patients vs 35/88,40% of the autonomous (ROS ≤2) patients. No previous hemorrhages were recorded for 53/88, 60% of the autonomous patients vs. 3/16, 19% of the non autonomous patients (Graph II).
- Firm intention of the patient to be treated was a factor which cannot be considered in our retrospective evaluation but, in our opinion, is of paramount importance in the prospective evaluation of the complex process of decision-making shared by the patient and his/her relatives.
- The remaining items (neurological deficits, small size, exclusive deep venous drainage, associated aneurysm or varix), did not reach statistical significance, even if their importance is evident in current literature.
Graph I.
Distribution of ROS according to age.
Graph II.
Distribution of ROS according to previous hemorrhages.
The factors contributing to the treatment risk score (TRS) were all considered separately as regards surgery, radiosurgery and embolization. All the items evaluated for surgery, (eloquence of adjacent brain, size, pattern of venous drainage) radiosurgery (volume) and embolization (eloquence of adjacent brain, volume, presence of perforating arteries as feeders, unfavourable angio-architecture) were considered important, even if they did not reach statistical significance. The only exception was the eloquence of adjacent brain which was significant by univariate analysis for surgery (p < 0.05). There were 14/21 (67%) autonomous patients with an AVM not adjacent to eloquent brain vs 7/22 (33%) patients with an AVM adjacent to eloquent brain. On the other hand, there were 1/6 (17%) non autonomous patients with an AVM not adjacent to eloquent brain vs 5/6 (83%) patients with an AVM adjacent to eloquent brain (Graph III).
Graph III.
Distribution of ROS according to eloquent/non eloquent adjacent brain.
As regards CS score, all patients had a score below 12 whatever treatment was selected (score 1-10: 99 patients; Score 11-12: five patients) (Table 4).
This means that the CS score formula, based on personal experience and literature data, was accurate enough to allow an adequate decision for treatment to be made in all the cases: no patient, even if evaluated retrospectively, scored more than 12, in which case treatment would not have been recommended.
Mean ITS was 5, TRS was 2.1 and CS was 7.1. There were small differences between the two groups of patients (autonomous and non autonomous) not reaching statistical significance.
This is not surprising since all the patients were actually considered treatable by our brain vascular team. The CS must be considered an important factor in the decision whether to treat a patient or not. On the other hand, the outcome may not be strictly correlated to CS since other variables (i.e. complications following surgery or anesthesia) may be more important than the patient and AVM features.
Conclusions
As regards the decision to proceed or abstain from treatment, our operative classification was efficient enough for us to verify that the advice for treatment was justified, as none of the patients had a CS score above 12 (the maximum limit for patient treatability). Among the 104 patients treated, only five had a score above 10 (treatment offered although at a significant risk): in all these patients embolization was performed as the first treatment to decrease the risks associated with surgery and radiosurgery (Figure 1).
Figure 1.
A 52-year-old woman with epileptic seizures. Left parietal post-central 5 ml AVM. ITS= 9 (Age ≥ 40 ≤60 yrs = 1; Previous hemorrhages NO= 1; Neurological deficits NO =1; Firm patient willing YES = 0; Deep brain location NO = 1; Exclusive deep venous drainage NO = 2; Associated aneurysm/varix NO = 2), TRS embolization = 2 (Volume <10 ml = 1; Eloquence YES = 1; Perforators NO = 0; Unfavourable angio-architecture NO = 0). CS 9+2 = 11. A) Pre-embolization coronal T2-weighted image. B) Preembolization left parasagittal MIP reconstruction from 2D-TOF MRA. C) Preembolization AP and LL angiography. D) AP and LL angiography after second embolization session. E) CT scan after second embolization. F) AP and LL view two and a half years after radiosurgery and three years after first embolization.
As regards the items taken into consideration for the ITS score, age and the presence or absence of previous hemorrhages were statistically significant also as independent factors, thus justifying the higher score attributed. As far as age is concerned, data obtained by the retrospective analysis of our 104 patients suggest the previous subdivisions should be modified. Considering that the two main subgroups of patients belonged to the age groups <40 or ≥ 40 and that only a few patients were over 60, the new age subdivisions are as follows: <40 (score 0); ≥ 40 < 60 (score 1); ≥ 60 (score 2).
Previous TRS, in our opinion, cannot be improved further, as all the items considered proved to be adequate, given that "eloquence of adjacent brain" was statistically significant.
Our attitude is based on a "soft" approach to AVMs at risk, without the aim of achieving 100% occlusion of these malformations by embolization alone. This is justified by the relative low global morbidity/mortality rates in our cases in comparison with other series 31. Our morbidity and mortality rates refer to a group of patients in which various treatments (mainly combined ones) achieved a near 100% obliteration of the AVM.
In conclusion, our classification (Tables 5 and 6) proved simple and efficient even though it includes many different parameters that were validated on the basis of a retrospective analysis of a significant number of treated patients.
Table 5 (A,B).
Brain AVMs intention to treat score (ITS: 0-12)
| A) Patient features (0-6) | ||
| Score | ||
|---|---|---|
| Age (yrs) | ≥ 60 | 2 |
| ≥ 40 < 60 | 1 | |
| < 40 | 0 | |
| Previous hemorrhages |
No | 2 |
| Yes | 0 | |
| Neurological deficits |
No | 1 |
| Yes | 0 | |
| Patient's firm intention to be treated |
No | 1 |
| Yes | 0 | |
| B) AVM characteristics (0-6) | ||
| Score | ||
| Small size |
No | 1 |
| Yes | 0 | |
| Deep brain Location |
No | 1 |
| Yes | 0 | |
| Exclusive deep Venous drainage |
No | 2 |
| Yes | 0 | |
| Associated Aneurysm/Variz |
No | 2 |
| Yes | 0 | |
Table 6.
BRAIN AVMs Treatment Risk Score (TRS: 1-5).
| A) Surgery (1-5) | B) Radiosurgery (1-5) | C) Embolization (1-5) | |||||||
| Grade | Score | Volume cm3 | Score | Score | |||||
|---|---|---|---|---|---|---|---|---|---|
| SPETZLER- MARTIN Modified |
I | 1 | < 5 | 1 | Volume cm3 |
< 10 | 1 | ||
| II | 2 | > 5 < 10 | 2 | > 10 < 20 | 2 | ||||
| III | 3 | > 10 < 20 | 3 | > 20 | 3 | ||||
| IV | 4 | > 20 < 30 | 4 | Eloquence | No | 0 | |||
| V | 5 | > 30 | 5 | Yes | 1 | ||||
| One point minus if low flow | Perforators | No | 0 | ||||||
| Yes | 1 | ||||||||
Since the beginning of 2008 this classification has been adopted as a routine protocol for the brain AVMs team in our Department to assess the indication to treat in every single case, whatever type of treatment is offered to the patient.
Data collected from the prospective application of this classification may be useful in the future for its further validation.
References
- 1.Brown RD, Jr, Wiebers DO, et al. The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg. 1988;68:352–357. doi: 10.3171/jns.1988.68.3.0352. [DOI] [PubMed] [Google Scholar]
- 2.Mast H, Young WL, et al. Risk of spontaneous haemorrhage after diagnosis of cerebral arteriovenous malformation. Lancet. 1997;350:1065–1068. doi: 10.1016/s0140-6736(97)05390-7. [DOI] [PubMed] [Google Scholar]
- 3.Ondra SL, Troupp H, et al. The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg. 1990;73:387–391. doi: 10.3171/jns.1990.73.3.0387. [DOI] [PubMed] [Google Scholar]
- 4.Choi JH, Mast H, et al. Clinical outcome after first and recurrent hemorrhage in patients with untreated brain arteriovenous malformation. Stroke. 2006;37:1243–1247. doi: 10.1161/01.STR.0000217970.18319.7d. [DOI] [PubMed] [Google Scholar]
- 5.Houdart E, Gobin YP, et al. A proposed angiographic classification of intracranial arteriovenous fistulae and malformations. Neuroradiology. 1993;35:381–385. doi: 10.1007/BF00588376. [DOI] [PubMed] [Google Scholar]
- 6.Valavanis A, Schubiger O, Wichmann W. Classification of brain arteriovenous malformation nidus by magnetic resonance imaging. Acta Radiol Suppl. 1986;369:86–89. [PubMed] [Google Scholar]
- 7.Berenstein A, Lasjaunias PL, Brugge K. Clinical and Endovascular Treatment Aspects in Adults, inBook. Heidelberg: Springer-Verlag; 2004. [Google Scholar]
- 8.Brown RD, Jr, Wiebers DO, et al. Frequency of intracranial hemorrhage as a presenting symptom and subtype analysis: a population-based study of intracranial vascular malformations in Olmsted Country, Minnesota. J Neurosurg. 1996;85:29–32. doi: 10.3171/jns.1996.85.1.0029. [DOI] [PubMed] [Google Scholar]
- 9.Graf CJ, Perret GE, Torner JC. Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg. 1983;58:331–337. doi: 10.3171/jns.1983.58.3.0331. [DOI] [PubMed] [Google Scholar]
- 10.Kondziolka D, McLaughlin MR, Kestle JR. Simple risk predictions for arteriovenous malformation hemorrhage. Neurosurgery. 1995;37:851–855. doi: 10.1227/00006123-199511000-00001. [DOI] [PubMed] [Google Scholar]
- 11.Marks MP, Lane B, et al. Hemorrhage in intracerebral arteriovenous malformations: angiographic determinants. Radiology. 1990;176:807–813. doi: 10.1148/radiology.176.3.2389040. [DOI] [PubMed] [Google Scholar]
- 12.Stapf C, Mast H, et al. Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology. 2006;66:1350–1355. doi: 10.1212/01.wnl.0000210524.68507.87. [DOI] [PubMed] [Google Scholar]
- 13.Miyasaka Y, Yada K, et al. An analysis of the venous drainage system as a factor in hemorrhage from arteriovenous malformations. J Neurosurg. 1992;76:239–243. doi: 10.3171/jns.1992.76.2.0239. [DOI] [PubMed] [Google Scholar]
- 14.Turjman F, Massoud TF, et al. Aneurysms related to cerebral arteriovenous malformations: superselective angiographic assessment in 58 patients. Am J Neuroradiol. 1994;15:1601–1605. [PMC free article] [PubMed] [Google Scholar]
- 15.Hamilton MG, Spetzler RF. The prospective application of a grading system for arteriovenous malformations. Neurosurgery. 1994;34:2–6. discussion 6-7. [PubMed] [Google Scholar]
- 16.Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg. 1986;65:476–483. doi: 10.3171/jns.1986.65.4.0476. [DOI] [PubMed] [Google Scholar]
- 17.Fisher WSI. Concomitant intracranial aneurysms and arteriovenous malformations. In: Wilkins Rh, Rengachary Ss., editors. Book. New York: McGraw-Hill; 1996. pp. 2429–2432. [Google Scholar]
- 18.Nicolato A, Lupidi F, et al. Gamma knife radiosurgery for cerebral arteriovenous malformations in children/ adolescents and adults. Part I: Differences in epidemiologic, morphologic, and clinical characteristics, permanent complications, and bleeding in the latency period. Int J Radiat Oncol Biol Phys. 2006;64:904–913. doi: 10.1016/j.ijrobp.2005.07.983. [DOI] [PubMed] [Google Scholar]
- 19.Nicolato A, Lupidi F, et al. Gamma knife radiosurgery for cerebral arteriovenous malformations in children/ adolescents and adults. Part II: Differences in obliteration rates, treatment-obliteration intervals, and prognostic factors. Int J Radiat Oncol Biol Phys. 2006;64:914–921. doi: 10.1016/j.ijrobp.2005.09.013. [DOI] [PubMed] [Google Scholar]
- 20.Liscak R, Vladyka V, et al. Arteriovenous malformations after Leksell gamma knife radiosurgery: rate of obliteration and complications. Neurosurgery. 2007;60:1005–1014. doi: 10.1227/01.NEU.0000255474.60505.4A. discussion 1015-1016. [DOI] [PubMed] [Google Scholar]
- 21.Andrade-Souza YM, Zadeh G, et al. Radiosurgery for basal ganglia, internal capsule, and thalamus arteriovenous malformation: clinical outcome. Neurosurgery. 2005;56:56–63. doi: 10.1227/01.neu.0000145797.35968.ed. discussion 63-54. [DOI] [PubMed] [Google Scholar]
- 22.Hartmann A, Pile-Spellman J, et al. Risk of endovascular treatment of brain arteriovenous malformations. Stroke. 2002;33:1816–1820. doi: 10.1161/01.str.0000020123.80940.b2. [DOI] [PubMed] [Google Scholar]
- 23.Haw CS, terBrugge K, et al. Complications of embolization of arteriovenous malformations of the brain. J Neurosurg. 2006;104:226–232. doi: 10.3171/jns.2006.104.2.226. [DOI] [PubMed] [Google Scholar]
- 24.Jayaraman MV, Marcellus ML, et al. Neurologic complications of arteriovenous malformation embolization using liquid embolic agents. Am J Neuroradiol. 2008;29:242–246. doi: 10.3174/ajnr.A0793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ledezma CJ, Hoh BL, et al. Complications of cerebral arteriovenous malformation embolization: multivariate analysis of predictive factors. Neurosurgery. 2006;58:602–611. doi: 10.1227/01.NEU.0000204103.91793.77. discussion 602-611. [DOI] [PubMed] [Google Scholar]
- 26.Hartmann A, Mast H, et al. Determinants of staged endovascular and surgical treatment outcome of brain arteriovenous malformations. Stroke. 2005;36:2431–2435. doi: 10.1161/01.STR.0000185723.98111.75. [DOI] [PubMed] [Google Scholar]
- 27.Beltramello A, Zampieri P, et al. Operative classification of brain arteriovenous malformations. Interventional Neuroradiology. 2008;14:9–19. doi: 10.1177/159101990801400102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Beltramello A, Zampieri P, et al. Combined Treatment of Brain AVMs: Analysis of Five Years (2000-2004) in the Verona Experience. Interventional Neuroradiology. 2005;11:63–72. doi: 10.1177/15910199050110S111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Pasqualin A, Scienza R, et al. Treatment of cerebral arteriovenous malformations with a combination of preoperative embolization and surgery. Neurosurgery. 1991;29:358–368. doi: 10.1097/00006123-199109000-00004. [DOI] [PubMed] [Google Scholar]
- 30.Pollock BE, Flickinger JC, et al. Factors that predict the bleeding risk of cerebral arteriovenous malformations. Stroke. 1996;27:1–6. doi: 10.1161/01.str.27.1.1. [DOI] [PubMed] [Google Scholar]
- 31.Jayaraman M, Cloft HJ. Embolization of brain arteriovenous malformations for cure: because we could or because we should? Am J Neuroradiol. 2009;30:107–108. doi: 10.3174/ajnr.A1394. [DOI] [PMC free article] [PubMed] [Google Scholar]