Neurosurg Rev (2014) 37:685–691
DOI 10.1007/s10143-014-0557-3

CASE REPORT

Case report of a de novo brainstem arteriovenous malformation
in an 18-year-old male and review of the literature
Kent J. Kilbourn & Gary Spiegel & Brendan D. Killory &
Inam Kureshi

Received: 20 July 2013 / Revised: 29 March 2014 / Accepted: 13 April 2014 / Published online: 4 July 2014
# Springer-Verlag Berlin Heidelberg 2014

Abstract De novo intracerebral arteriovenous malformations
(AVMs) are exceedingly rare with only seven reported cases
in the literature. Although generally considered congenital by
nature, the lesions do not manifest themselves clinically until
the third or fourth decades of life. However, with the advent of
improved imaging modalities and more frequent surveillance,
an increasing number of de novo cases are being found
challenging the concept AVMs develop in the perinatal/
antenatal period. Alternatively, this phenomenon could represent a distinct entity in which lesion development occurs after
birth. A PubMed search of “de novo cerebral arteriovenous
malformation” was performed in which seven reported cases
were found. The mean age at diagnosis was 14.7 years with a
mean follow-up imaging study of 5.8 years. Lesion location
was supratentorial in all previously described cases. This case
involves an 18-year-old male with congenital hydrocephalus
and seizures diagnosed at 7 months of age. The patient
underwent a ventriculoperitoneal shunt and was followed
frequently by a neurologist. The last diagnostic imaging was
an unremarkable MRI of the brain at age 12. Seven years later,
the patient presented with an intracerebral hemorrhage. A CT
angiogram demonstrated a large brainstem AVM with an
intraparenchymal hemorrhage and intraventricular extension.
This case is unique in that it is the first infratentorial de novo
AVM. The congenital nature of AVMs is challenged with the
increasingly described series of patients with previously documented normal radiographic imaging. This suggests there

K. J. Kilbourn (*) : G. Spiegel : B. D. Killory : I. Kureshi
Department of Neurosurgery Hartford Hospital,
85 Seymour Street, Hartford, CT 06102, USA
e-mail: kentkilbourn@yahoo.com
G. Spiegel
Department of Radiology Hartford Hospital,
85 Seymour Street, Hartford, CT 06102, USA

may be a subset of patients genetically predisposed to postnatal development of AVMs.
Keywords Cerebral arteriovenous malformation . De novo .
Brainstem . Angiogenesis

Introduction
Intracranial arteriovenous malformations (AVMs) are rare but
often devastating cause of stroke and disability. AVMs, which
are an abnormal collection of arterial and venous vessels with
no intermediary capillary system, are usually found in the
third and fourth decades of life [7, 13]. Hemorrhage, seizures,
and headaches are common presenting symptoms. De novo
AVMs are exceptionally rare with only seven reported cases in
the literature [21, 17, 9, 20, 22, 4, 5]. Herein, we present the
case of an 18-year-old male with a history of congenital
hydrocephalus with a ventriculoperitoneal (VP) shunt who
suffered an intracerebral hemorrhage. Imaging at presentation
revealed a large brainstem AVM which was not present on an
MRI done 7 years earlier.

Case report
This 18-year-old male with a past medical history significant
for autism was diagnosed with hydrocephalus at 7 months of
age. A VP shunt was placed at that time. At age 7, the patient
was diagnosed with absence seizures and was referred to a
pediatric neurologist. A non-contrast head CT scan was performed at that time which was normal (Fig. 1). Four years
later, a non-contrast MRI was performed which revealed
stable ventricular size and no abnormal flow voids (Fig. 2).
There is no family history of AVM. The patient was in his
usual state of health until immediately prior to admission

686

Neurosurg Rev (2014) 37:685–691

Fig. 1 Non-contrast axial CT
images (a) and (b) done 11 years
prior to hemorrhage

Fig. 2 Unremarkable non-contrast MRI done 7 years prior to the intracerebral hemorrhage depicting a T2 coronal image (a), axial T2 image (b), and
sagittal T1 image (c)

when he developed a severe headache with vomiting followed
by a period of unresponsiveness. The patient was intubated at
an outside hospital, stabilized, and transported to our facility
Fig. 3 Post hemorrhage noncontrast head CT scan depicting
blood in the brainstem (a) and
fourth ventricle (b)

by air ambulance. A non-contrast head CT revealed a 13-cc
hemorrhage involving the left pons extending into the fourth
ventricle (Fig. 3). A subsequent CTA demonstrated a

Neurosurg Rev (2014) 37:685–691

687

Fig. 4 Coronal CTA recon image (a) demonstrating two large posterior communicating arteries. Axial images (b) and (c) show a collection of irregular
vessels adjacent to the brainstem

Spetzler–Martin grade 4 AVM. A tangle of abnormal vessels
was seen along with an area of hemorrhage in the brainstem
(Fig. 4). Digital subtraction angiography (DSA) demonstrated
direct communication of the vertebrobasilar vessels with the
posterior communicating arteries. Venous drainage was
through diencephalic veins into the retroclival venous plexus
(Fig. 5). After being admitted to the neuroscience ICU, the
patient received an external ventricular catheter which was
placed adjacent to the VP shunt. The patient was sent to the
interventional neurosurgery suite for treatment. A 3-mm perinidal aneurysm arising from the anterior inferior cerebellar
artery (AICA) was felt to have been the source of the bleeding
and was subsequently embolized with 0.1 cc of N-butyl cyanoacrylate liquid. Definitive treatment of the AVM was deferred until the patient could recover from his initial
hemorrhage.

The patient had a prolonged hospitalization ultimately
requiring a tracheostomy and PEG tube placement. Upon
improvement in his neurological status, he was discharged
to an acute rehabilitation facility. At the time of discharge,
he was awake, alert, and moving his right leg and arm. An
MRI performed (Fig. 6) after the initial hemorrhage demonstrated resolving hemorrhage with persistent abnormal
flow voids. The AVM appeared stable in size. The patient
has since been discharged from rehab and regained some
level of pre-hemorrhage function with a modified Rankin
Scale (mRS) [23] of 4. He is able to eat and answers some
questions appropriately. The patient has been scheduled
for a repeat DSA in order to assess the remaining AVM.
Once this has been performed, treatment including further
embolization and/or stereotactic radiosurgery will be
considered.

Fig. 5 Digital subtraction angiography of vertebral/basilar arteries included an early (a) phase showing supplying vessels, including one
directly from the basilar artery. An intermediate supply vessel was

selected and angiographically examined (b) demonstrating venous drainage through the prepontine vein to the right Table 1. Previous case reports
of de novo AVMs

688

Neurosurg Rev (2014) 37:685–691

Fig. 6 MRI with gadolinium of the brain demonstrating the partially
embolized AVM. AT1-weighted gadolinium enhanced coronal image (a)
shows numerous abnormal vessels in the pontomedullary cistern. An

axial T2 (b) image shows abnormal flow voids within the brainstem.
Sagittal T1 (c) image shows magnetic susceptibility within the brainstem
reveals resolving hemorrhage

Discussion

The process of spontaneous AVM formation is not fully
understood. Mutations of two identified genes have been
implicated in patients with type 1 (endoglin (ENG)) and type
2 (activin-like kinase receptor 1 (ALK-1)) forms of hereditary
hemorrhagic telangiectasia (HHT) [18]. HHT is an autosomal
dominant vascular disorder which manifests with pulmonary
and cerebral AVMs among other anatomical locations. A third
gene, MADH4, which is part of the TGF-β and bone morphogenic protein (BMP) superfamily signaling pathway, has
been implicated in the third subtype of HHT [1]. The exact
role each gene defect plays in AVM development is unclear;
however, ALK-1, ENG, and MADH4 are all found in the
vascular endothelium and may play a role in smooth muscle
recruitment.
Mutations in these same genes have been demonstrated in
cases of sporadic AVM; however, it is also possible that a
“second hit” occurs in which the phenotypic expression of one
gene causes activation of other genes setting pathological
angiogenesis into motion. An overexpression of vascular endothelial growth factor (VEGF) may lead to downstream
signaling promoting abnormal angiogenesis and the inadequate recruitment of cells important in the formation of normal
vessel structure [14]. Numerous other signaling pathways
have been linked to AVM development including interleukin

This case represents a rare occurrence of a de novo AVM and
describes the first reported de novo infratentorial AVM. Two
prior imaging studies, including a high-resolution MRI, failed
to capture any abnormality, indicating lesion development
within 7 years of presentation. This case challenges the assertion that all AVMs develop in utero and instead may also form
after birth [19]. Improved imaging techniques and more frequent surveillance have resulted in the increasing number of
reported cases of de novo AVMs (Table 1). AVMs do not
appear to undergo linear growth, but instead change in size in
a bidirectional fashion. During a particular time interval,
spontaneous thrombosis may lead to angiographic disappearance [2] while angiogenesis and vessel recruitment can result
in AVM growth. It is possible that there was an occult lesion
not seen on either prior imaging study as we did not have a
dedicated vessel study and catheter angiography may have
revealed a small arteriovenous fistula without a nidus visible
on MRI. AVM reoccurrence has been reported despite complete angiographic obliteration of AVMs after surgery [11, 15,
10, 3] suggesting the natural history of the disease is a dynamic process in which phenotypic expression may evolve
over time.
Table 1 Previous reported cases of de novo AVMs
Reference

Demographic

AVM location

Time from negative exam (y)

Associated disease/condition

Gonzalez et.al [9]
O’Shaughnessy et. al [20]
Alvarez et. al [4]
Stevens et. al [22]
Schmit et. al [21]
Bulsara et. al [5]
Mahajan et. al [17]

7-year-old female
6-year-old-female
8-year-old-male
9-year-old-female
11-year-old male
32-year-old female
30-year-old female

Rt posterior temporal
Rt sylvian
Vein of Galen
Lt parieto-occipital
Lt parietal
Lt posttemporal
Rt posttemporal

4
3
2
3
9
6
14

Trauma
Sickle cell
CCM/angioma
Seizures/band heterotopia
Moyamoya
Inflammation/demyelination
Bell’s palsy/seizure

Neurosurg Rev (2014) 37:685–691

6 (IL-6), myeloperoxidase (MPO), and transforming growth
factor alpha (TGF-alpha) [14, 12]. Studies with inflammatory
markers from surgical specimens of un ruptured AVMs indicate that high levels of neutrophils and macrophages are found
in the walls of vessels within the AVM nidus and may play a
role in disease progression [6]. As described in some of the
previous case reports of de novo AVMs, trauma and inflammation may be predisposing factors. Inflammatory markers
have long been implicated in numerous disorders of the central nervous system, and in fact, this may be the common
thread among this unique group of patients. It has been suggested another important factor in AVM development may
arise from a “response to injury” [14] from an insult such as
inflammation, compression, infection, or other process. This,
coupled with an abnormal dysplastic response, may very well
set the stage for AVM development.
Hydrocephalus is a common presenting symptom in children with AVM, especially when located in the posterior fossa
[8]. In these instances, hydrocephalus may be caused by either
direct blockage of CSF flow from the AVM nidus, or more
commonly by venous hypertension. It is unclear whether this
patient’s VP shunt placement played any role in the formation
of this AVM. It is conceivable that changes in CSF flow
dynamics contributed to AVM development from an unrecognized congenital vascular anomaly or irregular capillary bed
which had been previously dormant; however, this mechanism has never been previously established. It has been hypothesized that cerebral cavernous malformation (CCM) and
AVM formation are secondary to changes occurring at the
capillary level leading to vessel fusion and AVM nidus formation [16].
De novo AVMs are rare with only seven reported cases in
the literature, with all occurrences in the supratentorial compartment. This is the first reported infratentorial lesion. It adds
to the growing number of this unique subset of vascular
malformations, further challenging the long-held belief that
all AVMs are congenital lesions.

References
1. Abdalla SA, Letarte M (2006) Hereditary haemorrhagic telangiectasia: current views on genetics and mechanisms of disease. J Med
Genet 43(2):97–110. doi:10.1136/jmg.2005.030833
2. Abdulrauf SI, Malik GM, Awad IA (1999) Spontaneous angiographic obliteration of cerebral arteriovenous malformations.
Neurosurgery 44(2):280–287, discussion 287–288
3. Akimoto H, Komatsu K, Kubota Y (2003) Symptomatic de novo
arteriovenous malformation appearing 17 years after the resection of
two other arteriovenous malformations in childhood: case report.
Neurosurgery 52(1):228–231, discussion 231–222
4. Alvarez H, Perry V, Solle M, Castillo M (2012) De novo cerebral
arteriovenous malformation in a child with previous cavernous

689
malformation and developmental venous anomaly. J Neurosurg
Pediatr 9(3):327–330. doi:10.3171/2011.12.PEDS11312
5. Bulsara KR, Alexander MJ, Villavicencio AT, Graffagnino C
(2002) De novo cerebral arteriovenous malformation: case
report. Neurosurgery 50(5):1137–1140, discussion 1140–1131
6. Chen Y, Zhu W, Bollen AW, Lawton MT, Barbaro NM, Dowd CF,
Hashimoto T, Yang GY, Young WL (2008) Evidence of inflammatory cell involvement in brain arteriovenous malformations.
Neurosurgery 62(6):1340–1349. doi:10.1227/01.neu.0000333306.
64683.b5, discussion 1349–1350
7. Garcin B, Houdart E, Porcher R, Manchon E, Saint-Maurice JP,
Bresson D, Stapf C (2012) Epileptic seizures at initial presentation
in patients with brain arteriovenous malformation. Neurology 78(9):
626–631. doi:10.1212/WNL.0b013e3182494d40
8. Geibprasert S, Pereira V, Krings T, Jiarakongmun P, Lasjaunias P,
Pongpech S (2009) Hydrocephalus in unruptured brain arteriovenous
malformations: pathomechanical considerations, therapeutic implications, and clinical course. J Neurosurg 110(3):500–507. doi:10.3171/
2008.7.JNS0815
9. Gonzalez LF, Bristol RE, Porter RW, Spetzler RF (2005) De novo
presentation of an arteriovenous malformation. Case report and review of the literature. J Neurosurg 102(4):726–729. doi:10.3171/jns.
2005.102.4.0726
10. Hashimoto N, Nozaki K (1999) Do cerebral arteriovenous
malformations recur after angiographically confirmed total extirpation? Crit Rev Neurosurg: CR 9(3):141–146
11. Kader A, Goodrich JT, Sonstein WJ, Stein BM, Carmel PW,
Michelsen WJ (1996) Recurrent cerebral arteriovenous
malformations after negative postoperative angiograms. J
Neurosurg 85(1):14–18. doi:10.3171/jns.1996.85.1.0014
12. Kilic T, Pamir MN, Kullu S, Eren F, Ozek MM, Black PM (2000)
Expression of structural proteins and angiogenic factors in cerebrovascular anomalies. Neurosurgery 46(5):1179–1191, discussion
1191–1172
13. Kim H, Sidney S, McCulloch CE, Poon KY, Singh V, Johnston SC,
Ko NU, Achrol AS, Lawton MT, Higashida RT, Young WL, Project
UBS (2007) Racial/ethnic differences in longitudinal risk of intracranial hemorrhage in brain arteriovenous malformation patients.
Stroke: J Cereb Circ 38(9):2430–2437. doi:10.1161/STROKEAHA.
107.485573
14. Kim H, Su H, Weinsheimer S, Pawlikowska L, Young WL (2011)
Brain arteriovenous malformation pathogenesis: a response-to-injury
paradigm. Acta neurochir Suppl 111:83–92. doi:10.1007/978-37091-0693-8_14
15. Kondziolka D, Humphreys RP, Hoffman HJ, Hendrick EB, Drake
JM (1992) Arteriovenous malformations of the brain in children: a
forty year experience. Can J Neurol Sci 19(1):40–45
16. Leblanc GG, Golanov E, Awad IA, Young WL, Biology of Vascular
Malformations of the Brain NWC (2009) Biology of vascular
malformations of the brain. Stroke: J Cereb Circ 40(12):e694–e702.
doi:10.1161/STROKEAHA.109.563692
17. Mahajan A, Manchandia TC, Gould G, Bulsara KR (2010) De
novo arteriovenous malformations: case report and review of
the literature. Neurosurg Rev 33(1):115–119. doi:10.1007/
s10143-009-0227-z
18. Marchuk DA, Srinivasan S, Squire TL, Zawistowski JS (2003)
Vascular morphogenesis: tales of two syndromes. Hum Mol Genet
12(1):R97–R112
19. Mullan S, Mojtahedi S, Johnson DL, Macdonald RL (1996)
Embryological basis of some aspects of cerebral vascular fistulas and
malformations. J Neurosurg 85(1):1–8. doi:10.3171/jns.1996.85.1.0001
20. O'Shaughnessy BA, DiPatri AJ Jr, Parkinson RJ, Batjer HH (2005)
Development of a de novo cerebral arteriovenous malformation in a
child with sickle cell disease and moyamoya arteriopathy. Case
report. J Neurosurg 102(2 Suppl):238–243. doi:10.3171/jns.2005.
102.2.0238

690
21. Schmit BP, Burrows PE, Kuban K, Goumnerova L, Scott RM (1996)
Acquired cerebral arteriovenous malformation in a child with
moyamoya disease. Case report. J Neurosurg 84(4):677–680. doi:
10.3171/jns.1996.84.4.0677
22. Stevens J, Leach JL, Abruzzo T, Jones BV (2009) De novo cerebral
arteriovenous malformation: case report and literature review. AJNR
Am J Neuroradiol 30(1):111–112. doi:10.3174/ajnr.A1255
23. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J
(1988) Interobserver agreement for the assessment of handicap in
stroke patients. Stroke:J Cereb Circ 19(5):604–607

Comments
Martin M. Mortazavi and Louis J. Kim, Seattle, USA
With an average annual risk of bleeding of 2–4 %, arteriovenous
malformations (AVM) account for 3 % of hemorrhagic strokes and 9 %
of subarachnoid hemorrhage [1]. They represent the most common cause
of childhood hemorrhagic stroke [2]. Infratentorial AVMs are uncommon,
comprise less than 15 % of all cerebral AVMs, and are associated with
higher annual risk of hemorrhage, reaching up to 11 % [3, 4].
Up to our knowledge, eight cases of acquired AVM have been
reported in the literature thus far [5–12]. More cases are expected due
to the advancement of neuroimaging. In all of these cases, the malformation has developed in the supratentorial area. The interesting finding of
infratentorial de novo AVM reported in the current article is therefore
unique.
The etiology of AVM is still unknown. The early theory of a static
congenital malformation of solely genetic origin has been questioned
with the reported cases of de novo AVM and AVM recurrence after
successful treatment. Therefore, later studies have suggested that acquired
somatic conditions including trauma, inflammation, and radiation are
potentially involved, and they may represent a second hit for de novo
AVM development [2, 13]. This may also explain the occurrence of AVM
in certain vascular territory and absence in another, in a genetically
predisposed patient. However, this theory was only tested on mice, and
no genetic studies were performed on patients with de novo AVM to date.
Certain genes have been associated with AVM and well documented in
this current manuscript. The incidence of a “second hit” is interesting as
there is a potential parallel in hereditary hemorrhagic telangiectasia
(HHT). Park et al. [13] reported de novo subdermal AVM formation
following wounding of an activin receptor-like kinase-1 (ALK-1)-deficient mouse. ALK-1 is a transforming growth factor-β (TGF-β) receptor
protein that inhibits the TGF-β-induced vascular endothelial growth
factor (VEGF) expression in endothelial cells, and its deficiency is one
of the main underlying causes of a “second hit,” in this case, the wound
itself that could induce the release of multiple angiogenic factors, including VEGF, which potentially could lead to local AVM formation [13].
It must again be noted that none of reported cases of de novo AVM
were for patients with HHT, and unique molecular mechanism is, therefore, sought after.
At our institution, the treatment approach to brainstem AVMs is
dictated by the accessibility and size of the lesion. In general, Spetzler–
Martin grade 3 lesions are suitable for resection. For grade 4 and 5 AVMs,
treatment is conservative or radiosurgery. In our experience, the specific
location within the brainstem and degree of surgical accessibility greatly
influences the likelihood of surgical success without permanent neurological morbidity.
References
1. Mortazavi MM et al (2013) Long-term control of large pontine
arteriovenous malformation using gamma knife therapy: a review with
illustrative case. Brain Behav 3(4): 329–334

Neurosurg Rev (2014) 37:685–691
2. Achrol AS et al (2009) Pathogenesis and radiobiology of brain
arteriovenous malformations: implications for risk stratification in natural
history and posttreatment course. Neurosurg Focus 26(5): E9
3. Arnaout OM et al (2009) Posterior fossa arteriovenous
malformations. Neurosurg Focus 26(5): E12
4. Stapf C et al (2006) Predictors of hemorrhage in patients with
untreated brain arteriovenous malformation. Neurology. 66(9): 1350–5
5. Alvarez H et al (2012) De novo cerebral arteriovenous malformation in a child with previous cavernous malformation and developmental
venous anomaly. J Neurosurg Pediatr 9(3): 327–30
6. Bulsara KR et al (2002) De novo cerebral arteriovenous malformation: case report. Neurosurgery 50(5): 1137–40; discussion 1140–1
7. Gonzalez LF et al (2005) De novo presentation of an arteriovenous
malformation. Case report and review of the literature. J Neurosurg
102(4): 726–9
8. Mathon B et al (2013) De novo arteriovenous malformation
after brain radiotherapy for medulloblastoma in a child. Neurology
81(4): 398–9
9. O’Shaughnessy BA et al (2005) Development of a de novo cerebral
arteriovenous malformation in a child with sickle cell disease and
moyamoya arteriopathy. Case report. J Neurosurg 102(2 Suppl): 238–43
10. Schmit BP et al (1996) Acquired cerebral arteriovenous malformation in a child with moyamoya disease. Case report. J Neurosurg
84(4): 677–80
11. Stevens J et al (2009) De novo cerebral arteriovenous malformation:
case report and literature review. AJNR Am J Neuroradiol 30(1): 111–2
12. Mahajan A et al (2010) De novo arteriovenous malformations:
case report and review of the literature. Neurosurg Rev 33(1): 115–9
13. Park SO et al (2009) Real-time imaging of de novo arteriovenous
malformation in a mouse model of hereditary hemorrhagic telangiectasia.
J Clin Invest 119(11): 3487–96
Oliver Müller and Ulrich Sure, Essen, Germany
Kilbourn et al. report on the first de novo AVM of the posterior fossa,
so far. De novo AVMs are indeed exceptional occurrences with respect to
this latest review. Scarce information is available on the incidence of brain
AVMs. Large prospective studies estimated an incidence of brain AVMs
of 1.3 per 100,000 person-years with a cumulative bleeding risk of 0.51
[1]. On the other hand, a large study on incidental MRI findings did not
detect any brain AVMs among 2,000 scans [2]. With this in mind, the
present report should inspire us to dig even deeper into the potential
biology of brain AVMs.
Are AVMs hereditary, connatal, or acquired lesions? Several studies
have uncovered proliferative and neoangiogenetic abilities within the
AVM vasculature [3–6] that might even be induced by previous partial
treatment. Are AVMs benign vascular tumors rather than true
malformations?
ARUBA has told us that medical treatment is superior to medical
treatment and intervention over a (too short for a benign tumor or
malformation) 33-month period in unruptured AVMs [7]. However, this
conclusion must be discussed highly controversially since 87 % of the
malformations (assessed for eligibility) were excluded from statistical
analysis in this prospective study. In the treatment arm group, the vast
majority of patients was not treated surgically (more than 80 %), suggesting that these AVMs were only occluded partially with all potential risk of
a remaining nidus. If future research reveals a more aggressive growth
behavior as shown in de novo AVMs, this finding will furthermore
downsize the significance of the results of the ARUBA study which
was designed under the assumption that AVMs are rather stable nonneoplastic non-growing lesions.
Therefore, the present well-documented case is an important contribution to the literature.
References
1. Stapf C et al (2003) The New York Islands AVM Study: design,
study progress, and initial results. Stroke 34(5): e29–33

Neurosurg Rev (2014) 37:685–691
2. Vernooij MW et al (2007) Incidental findings on brain MRI in the
general population. N Engl J Med 357(18): 1821–8
3. Rothbart D et al (1996) Expression of angiogenic factors and
structural proteins in central nervous system vascular malformations.
Neurosurgery 38(5): 915–24; discussion 924–5
4. Kilic T et al (2000) Expression of structural proteins and angiogenic
factors in cerebrovascular anomalies. Neurosurgery 46(5): 1179–91; discussion 1191–2

691
5. Sure U et al (2001) Treatment-induced neoangiogenesis in cerebral
arteriovenous malformations. Clin Neurol Neurosurg 103(1): 29–32
6. Sandalcioglu IE et al (2010) Proliferation activity is significantly
elevated in partially embolized cerebral arteriovenous malformations.
Cerebrovasc Dis 30(4): 396–401
7. Mohr JP et al (2014) Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA):
a multicentre, non-blinded, randomised trial. Lancet 383(9917): 614–21