J Neurosurg Pediatrics 12:344–350, 2013

©AANS, 2013

An institutional series and literature review of pial

arteriovenous fistulas in the pediatric population
Clinical article

Peter J. Madsen, M.D., M.B.E.,1 Shih-Shan Lang, M.D.,1

Jared M. Pisapia, M.D., M.T.R.,1 Phillip B. Storm, M.D.,1,3 Robert W. Hurst, M.D., 2
and Gregory G. Heuer, M.D., Ph.D.1,3

Departments of 1Neurosurgery and 2Radiology, Hospital of the University of Pennsylvania; and 3Division of
Neurosurgery, Children’s Hospital of Philadelphia, Pennsylvania

Object. Pial arteriovenous fistulas (PAVFs) are a rare form of cerebrovascular disease that tend to be overrepre-
sented in the pediatric population. There have been limited studies of the clinical features and outcomes in this group
of patients. Here, the authors attempt to better delineate this clinical entity with institutional cases and a review of
the literature.
Methods. A retrospective review of cases at our institution was performed to identify all pediatric patients treated
for a PAVF between 2000 and 2012.
Results. Five patients treated for a PAVF were identified. Patients had a mean age of 1.9 years at diagnosis,
and the most common presenting symptoms were seizure and macrocephaly. Patients were treated primarily with
embolization, and 3 patients required both N-butyl cyanoacrylate (NBCA) glue and coiling. Four of the patients had
complete obliteration of the PAVF and had a pediatric overall performance category score of either 1 (n = 3) or 2 (n
= 1) at follow-up. There was 1 death due to heart failure. Analysis of the literature review suggested that a younger
age or presence of intracerebral hemorrhage (ICH) or congestive heart failure (CHF) at presentation likely predicts
a worse prognosis. Older patients presented more often with ICH, whereas younger patients presented significantly
more often in CHF. The majority of pediatric patients reported on in the literature were treated with endovascular em-
bolization, most commonly with NBCA glue alone. Most patients (65.4%) in the literature had an excellent outcome
without neurological deficit.
Conclusions. Pial AVFs represent a serious yet rare form of cerebrovascular disease. Pediatric patients with ICH
or CHF at presentation or those who are very young are likely to have a worse prognosis. Endovascular management
of these patients has greatly changed the natural history of this disease, but the complication and mortality rates sug-
gest the need for continued insights and advances in treatment.
(http://thejns.org/doi/abs/10.3171/2013.6.PEDS13110)

Key Words      •      pial arteriovenous fistula      •      AVF      •      varix      •      surgery      •     

embolization      •      pediatric      •      vascular disorders

P
ial (or nongalenic) arteriovenous fistulas are a rare tion at an increased incidence of 17.2% of pial AVMs.52
form of cerebrovascular disease consisting of a The clinical features of PAVFs, especially in children,
high-flow connection between one or more pial ar- have been described in limited detail in the literature due
teries and the venous circulation.18,48,53 They are typically to the rarity of this entity. It has been suggested that there
classified as a subtype of pial AVMs, yet are differenti- is an increased incidence of venous varices, a higher rate
ated by their lack of an intervening nidus between the of seizures, and a lower rate of hemorrhage in children
arterial feeder(s) and venous drainage.24 Pial AVFs have than in adults.53
been found to account for 4.7% of pial AVMs;47 however, The management strategies for PAVFs have under-
they tend to be overrepresented in the pediatric popula- gone considerable evolution with the advancement and
refinement of endovascular techniques. Despite advances
Abbreviations used in this paper: AVM = arteriovenous malfor-
in treatment, PAVFs remain a potentially life-threatening
mation; CHF = congestive heart failure; CN = cranial nerve; GDC = condition, necessitating further evaluation and investiga-
Guglielmi Detachable Coil; ICH = intracerebral hemorrhage; MCA tion. In this study, we identify and delineate the clinical
= middle cerebral artery; NBCA = N-butyl cyanoacrylate; PAVF = course of pediatric patients with PAVFs at our tertiary
pial arteriovenous fistula; PCA = posterior cerebral artery. care children’s hospital. In addition, because of the rarity

344 J Neurosurg: Pediatrics / Volume 12 / October 2013

Pial arteriovenous fistulas in the pediatric population

of pediatric PAVFs, we performed a literature review of ing head circumference, and delay of motor development
published cases. was found to have angiographic evidence of a mixed
AVM and PAVF with venous varix fed by the right angu-
Methods lar artery that drained into the right sigmoid sinus. Em-
bolization of both the AVM and PAVF was performed
Study Design and Clinical Data using NBCA glue and resulted in complete occlusion.
A retrospective review of cases at a single tertiary The residual nidus of the AVM was resected. There were
referral children’s hospital was performed to identify no complications from either procedure, and the patient
patients who demonstrated angiographic evidence of a remained in excellent condition and seizure free during 4
PAVF between 2000 and 2012. Patients older than 18 years of follow-up.
years and those with an isolated AVM or vein of Galen
malformation were excluded. Patient demographics, clin- Case 3. This 6-month-old boy who presented with
ical presentation, radiographic images, angioarchitecture, atonic seizures and macrocephaly was found on CT to
treatment details, complications, and outcomes were ex- have a thrombosed varix. Magnetic resonance angiog-
tracted from inpatient hospital records and neurosurgical raphy, MR venography, and cerebral angiography con-
outpatient clinic charts. Only patients with angiographic firmed the presence of an occipital PAVF fed by the right
features of a PAVF such as rapid venous filling, large ar- MCA and draining into the straight and right transverse
terial feeder(s), and direct filling of a venous varix were sinuses via the vein of Labbé (Fig. 1A–C). Embolization
included. Neurological outcomes were assessed using the of the fistula was performed using detachable coils and
pediatric overall performance category.11,12 NBCA glue, with postembolization angiography showing
minimal flow through the PAVF. Radiographic follow-up
Literature Review over the next 3 months demonstrated complete oblitera-
tion of the fistula (Fig. 1D). After the procedure, the pa-
We performed a MEDLINE and a Cochrane data-
tient neglected stimuli presented in his left visual field,
base search of articles published in the English-language
literature before December 2012 containing the terms but this finding resolved over 3 months. At 3 years follow-
“pial arteriovenous fistula” and “nongalenic arterio- up the patient was neurologically stable and seizure free.
venous fistula.” We supplemented our search using the Case 4. This 1-week-old girl presented with CHF and
“Related Citations” feature of PubMed and performing associated pulmonary hypertension. Magnetic resonance
manual reviews of the bibliographies of selected articles. imaging, MR angiography, MR venography, and cerebral
Only studies in which individual patient data could be angiography demonstrated a large PAVF adjacent to the
clearly extracted were included in our review. Patients midbrain comprising numerous dilated arterial feed-
with AVMs, vein of Galen malformations, or with unclear ers arising from the right posterior choroidal artery and
angiographic characterization were excluded. right posterior inferior cerebellar artery with drainage to
the deep cerebral veins. The malformation was associ-
Results ated with mild hydrocephalus. An attempt at emboliza-
Summary of Cases tion was performed using NBCA and detachable coils
with incomplete obliteration and residual arterial feeding
Between 2000 and 2012, 5 patients were identified as vessels. Postembolization ultrasonography of the head
meeting our inclusion criteria for a PAVF (Tables 1 and showed evidence of a possible thalamic infarct. The pa-
2). Individual patient cases are described. tient’s heart failure continued to worsen, and she died at 3
Case 1. This 3-month-old boy, who was previously weeks of age of multisystem organ failure.
reported on in the literature,45 presented with evidence of Case 5. This 8-year-old girl was found to have a
right CN III palsy, emesis, and a cranial bruit. Magnetic left parietooccipital vascular anomaly on MRI during
resonance imaging and MR angiography and convention- workup for multiple cutaneous hemangiomas present
al cerebral angiography demonstrated a large (6.6 × 4.4 × since birth. Genetic testing uncovered a RASA1 muta-
4.7–cm) dilation of the M2 branch of the right MCA, caus- tion passed along by her father who harbored the same
ing mass effect on the midbrain, pons, and CN III. There mutation. RASA1-related disorders are characterized by
was direct communication of this dilation with enlarged face and appendicular capillary malformations as well
venous drainage into the junction of the right transverse as AVMs and/or AVFs in the skin, muscle, bone, spine,
and sigmoid sinuses. and brain.10 The patient was neurologically asymptomatic
Embolization of the fistula was performed using from the vascular lesion. On cerebral angiography, she
GDCs and NBCA glue, achieving minimal residual fill- was noted to have a multichannel PAVF that was fed by
ing. The procedure was complicated by ischemic stroke the left pericallosal and posterior cerebral arteries and
of the right MCA territory, resulting in mild hemipare- drained via the straight sinus with an associated venous
sis. At 1 year of follow-up, complete occlusion of the fis- varix (Fig. 2A–C). Embolization was performed using
tula was achieved. The patient’s hemiparesis improved;
multiple GDCs, with the achievement of immediate com-
however, he ultimately required medial and lateral rectus
plete obliteration (Fig. 2D). The patient had no complica-
muscle resection to correct persistent CN III dysfunction.
tions from the procedure and continued to be neurologi-
Case 2. This 9-month-old boy with seizures, increas- cally intact at 1 month of follow-up.

J Neurosurg: Pediatrics / Volume 12 / October 2013 345

 P. J. Madsen et al.
TABLE 1: Patient characteristics*

Case
No. Age at Dx, Sex Presentation Location Feeding Vessel Draining Vessel Varix
1 3 mos, M CN III palsy, cranial bruit, rt frontotemporal M2 branch of MCA rt transverse sinus yes
 emesis
2 9 mos, M seizure, macrocephaly, rt temporoparietal rt MCA rt sigmoid sinus yes
  developmental delay
3 6 mos, M seizure, macrocephaly rt occipital rt MCA rt vein of Labbé yes
4 1 wk, F heart failure, pulmonary interpeduncular fossa rt posterior choroidal deep cerebral yes
 HTN   artery, rt PICA  veins
5 8 yrs, F cutaneous hemangiomas lt parietooccipital lt pericallosal artery, straight sinus yes
lt PCA

*  Dx = diagnosis; HTN = hypertension; PICA = posterior inferior cerebellar artery.

Literature Review and interpret the data reported from our literature review
Sixty-two papers were reviewed; 40 of which were of pediatric PAVFs.
considered suitable for the current study.1–9,13–17,20–23,26–28, Clinical Presentation
30–33,35–38,40–44,50–52,54–56
These studies included 27 case re-
ports or series and 13 retrospective studies and yielded Similar to vein of Galen aneurysmal malformations,
extractable data on 147 cases of intracranial PAVF in pa- the clinical presentation in pediatric patients with PAVFs
tients younger than 18 years old (Table 3). The mean age of tends to vary by age. Studies have suggested that neonates
patients in the literature was 5.02 years with a slight male with a PAVF are more likely to present in CHF than older
predominance, and the most common presenting symp- patients.18,40 This is further supported by the fact that 22
tom was seizure, followed by CHF. The most common of the 28 patients in our literature review who presented
first-line treatment modality performed was embolization with cardiac failure were in the neonatal period (< 28
with NBCA glue, and the vast majority of patients had no days). The average age of patients presenting in CHF was
complications and excellent neurological outcomes. 0.21 ± 0.52 years compared with 6.2 ± 5.4 years for those
patients presenting without CHF (p < 0.0001, unpaired t-
test). Our series also supported this association, as our pa-
Discussion tient who presented with CHF was only 1 week old (Table
1). The other 4 patients were all 3 months of age or older
Pial AVF is a rare form of cerebrovascular disease, and did not have symptoms of heart failure. Historically,
with a reported incidence of 3%–7.3% in clinical series of newborns with CHF due to a cranial arteriovenous shunt
AVMs in all age groups.40 This increases to a reportedly have a very poor prognosis that approached 100%, but
higher incidence of 17.2% of AVMs in the pediatric popu- advances in endovascular techniques in this population
lation.52 Limited distinction has been made between the have improved outcomes significantly.13,40
clinical features of this disease in pediatric and adult pa- Intracerebral hemorrhage at presentation is also sug-
tients. To better define the pathogenesis, clinical elements, gested to be age dependent, in that it is typically seen as
therapeutic strategies, and outcomes of this disease in the a presenting sign in older patients, especially those older
pediatric population, we describe our experience with than 15 years.53 In our literature review, the mean age of
this clinical entity at our tertiary care children’s hospital patients presenting with ICH was 6.4 ± 5.5 years versus a
TABLE 2: Treatment and patient outcomes*

Case Embolization Outcome Follow-Up

No. Treatment Modality Material Complication Fistula Occlusion (POPC)† (yrs)
1 embol NBCA, GDCs MCA territory infarct, mild 100% 2 0.9
 hemiparesis
2 embol, op‡ NBCA none 100% 1 4.2
3 embol NBCA, GDCs transient visual field neglect 100% 1 2.8
4 embol NBCA, GDCs thalamic infarct incomplete 6 NA
5 embol GDCs none 100% 1 0.1

*  embol = embolization; NA = not applicable; POPC = pediatric overall performance category.

†  Based on the system of Fiser11 and Fiser and colleagues:12 1 = good overall performance, 2 = mild overall disability, 6 = brain
death.
‡  Surgery was performed only on the nidal portion of the patient’s associated AVM.

346 J Neurosurg: Pediatrics / Volume 12 / October 2013

Pial arteriovenous fistulas in the pediatric population

Fig. 2.  Case 5. Preembolization digital subtraction angiograms of

the lateral left internal carotid artery injection (A), lateral left vertebral
artery injection (B), and superimposed, separate carotid and vertebral
injections (C) showing a left parietooccipital PAVF arising from the left
pericallosal and PCA with drainage into the straight sinus and an asso-
ciated venous varix (circle). Embolization was performed through a mi-
Fig. 1.  Case 3. Preembolization lateral (A) and anteroposterior (B) crocatheter threaded from the vertebral artery through the PCA across
digital subtraction angiograms and MR angiogram (C) showing a right the fistula and into the distal anterior cerebral artery. Postembolization
temporooccipital PAVF arising from an enlarged, distal MCA branch angiogram demonstrating obliteration of the PAVF with no filling from
with deep (vein of Labbé) and superficial drainage and an associated either the internal carotid artery or vertebral artery on separate injec-
venous varix. Postembolization MR angiogram (D) demonstrating oblit- tions (circle, D).
eration of the PAVF.
(GDCs and NBCA glue) to achieve complete obliteration
of the PAVF.
mean of 4.8 ± 5.4 years for those patients presenting with- Notably, there were very few patients (4 [3.5%] of 113)
out ICH (p = 0.19, unpaired t-test), although this finding in the literature review whose PAVFs were embolized with
was not statistically significant. Onyx, which contrasts with the increasing usage of this
agent for treating other cerebrovascular malformations, es-
Treatment of PAVFs pecially in the adult population. At our institution, NBCA
The vast majority of pediatric patients with PAVF in glue is preferred over Onyx, because Onyx utilization typi-
the literature were treated with endovascular embolization cally extends the embolization29 and requires more fluoro-
(113 patients [83.7%]) rather than surgical intervention (23 scopic time,49 which potentially increases the detrimental
patients [17.0%]) (Table 3). Of the 23 reports involving sur- effects of ionizing radiation on children’s developing ner-
gical interventions, 19 of these cases were reported prior to vous systems. Furthermore, the potentially large doses of
2002, which represents a significant shift in management Onyx required for embolization of these fistulas increas-
strategies with the advancement of the endovascular field. es the risk for the rare complication of DMSO (dimethyl
In terms of the type of embolic material used, NBCA glue sulfoxide)–related pulmonary toxicity39,46 associated with
alone was the most common (53 [46.9%] of 113) followed Onyx. This pulmonary toxicity may exacerbate the preex-
by coiling alone (26 [23.0%] of 113) and combined coil- isting cardiovascular and pulmonary comorbidities often
ing and NBCA glue (21 [18.6%] of 113; Table 3). Of the seen in this group of children.
108 patients who underwent embolization as their initial
treatment and had data reported on fistula obliteration, 74 Complications and Outcomes
(68.5%) had angiographic evidence of complete oblitera- Procedure-related death was infrequent in the litera-
tion of their PAVF. ture review (3.0%), representing 4 patients who died of
Of the 34 patients who continued to have residual ICH after embolization.30,40 The most common procedure-
filling of the PAVF, 24 (70.6%) had initially received related complication reported in the literature was the de-
NBCA glue alone, 7 (20.6%) received coiling and NBCA velopment of new intracranial hemorrhage occurring in
glue, 2 (5.9%) received coiling alone, and 1 (2.9%) re- 17 (12.6%) of 135 patients, followed by the development
ceived balloon occlusion. In our series, all patients were of new neurological deficit (10 [7.4%] of 135 patients; Ta-
treated with endovascular embolization, with one patient ble 3). Complication rates in series of patients with related
subsequently undergoing open resection of an associated malformations, such as vein of Galen aneurysmal malfor-
AVM (Table 2). Three of the patients from our institu- mations, have been reported at 17%–36%.19,25,34
tional series required combined embolization material Among cases in the literature in which outcome was

J Neurosurg: Pediatrics / Volume 12 / October 2013 347

 P. J. Madsen et al.
TABLE 3: Composite data from a literature review of 147 pediatric TABLE 3: Composite data from a literature review of 147 pediatric
patients presenting with PAVFs* patients presenting with PAVFs* (continued)

Parameter Value Parameter Value

patient demographics (n = 147) complications (n = 135)§,¶ (continued)
 age  seizure 1 (0.7)
  mean 5.02 yrs   mental retardation 1 (0.7)
  range 0 days–17 yrs   glued microcatheter 1 (0.7)
  male/female ratio 1.5:1   pedicle embolization 1 (0.7)
  male 88 (59.9) outcome (n = 130)¶
  female 59 (40.1)   excellent (no deficit) 85 (65.4)
common presenting signs/symptoms (n = 147)   good (minimal deficit) 21 (16.2)
 seizure 36 (24.5)  poor 13 (10.0)
 CHF 28 (19.0)  death 11 (8.5)
  neurological deficit 27 (18.4) average LOF (n = 49)¶ 17.4 mos
 headache 24 (16.3)
 hemorrhage 22 (15.0) *  Unless otherwise noted, values represent the number of patients (%).
DAVF = dural arteriovenous fistula; IVH = intraventricular hemorrhage;
 macrocephaly 21 (14.3)
LOF = length of follow-up; SAH = subarachnoid hemorrhage; SDH =
  genetic syndrome 19 (12.9) subdural hemorrhage; SRS = stereotactic radiosurgery.
  developmental delay 12 (8.2) †  Includes only cases in which the presence or absence of varix was
presence of varix (n = 119)† noted or could be determined from images provided.
 present 110 (92.4) ‡  The sum of individual treatment numbers is greater than the total
 absent 9 (7.6) number of treated patients because some patients had multiple inter-
ventions.
treatment modality (n = 135)‡
§  The sum of individual complications is greater than the total number
  surgical clipping/resection 23 (17.0) of patients with complications because some patients had multiple com-
 embolization 113 (83.7) plications.
  w/ glue 53 (46.9) ¶  Includes only cases in which individual complications, outcomes, or
  w/ coiling 26 (23.0) length of follow-up was reported.
   w/ glue & coiling 21 (18.6)
  balloon occlusion 11 (9.7) clearly reported, the majority of patients (85 [65.4%] of
130) achieved an excellent outcome, defined as the ab-
  Onyx 4 (3.5)
sence of any neurological deficit at follow-up. Twenty-one
  unknown material 2 (1.8) patients (16.2%) were noted to have a minimal neurologi-
 SRS 1 (0.7) cal deficit at follow-up, considered a good outcome, and
 observation 7 (5.2) 13 (10.0%) had a poor outcome, representing a significant
complications (n = 135)§,¶ neurological deficit at follow-up (for example, persistent
 none 102 (75.6) hemiparesis, mental retardation, and persistent visual field
 any 34 (25.2)
deficit). In the literature review, 11 patients died (8.5%);
most deaths were due to either ICH (n = 5) or CHF (n = 4).
  procedure-related death 4 (3.0) The 1 death reported in our series was attributable to the
  any intracranial hemorrhage 17 (12.6) patient’s CHF rather than any neurological compromise.
  ICH 8 (5.9) Patients with a clearly reported outcome in the lit-
  IVH 4 (3.0) erature who presented with ICH had a higher mortality
  SAH 3 (2.2) rate than those without ICH on presentation (4 [18.2%] of
  SDH 2 (1.5)
22 vs 7 [6.5%] of 108, respectively; p = 0.09, Fisher exact
test). In addition to ICH, age at diagnosis also seemed
  new neurological deficit 10 (7.4)
to be a potential indicator of increased mortality. In the
  coil/balloon migration 6 (4.4) reviewed literature the mean age at diagnosis of those pa-
  secondary DAVF 4 (3.0) tients who died was 2.3 ± 3.7 years compared with 5.6 ±
  CSF diversion required 3 (2.2) 5.5 years for those who survived (p = 0.05, unpaired t-
  venous sinus occlusion 3 (2.2) test). The observation of a worse prognosis in younger pa-
  venous infarct 1 (0.7) tients (especially those < 2 years) has also been described
by Hetts et al., who attributed it to the high rate of severe,
 sepsis 1 (0.7)
life-threatening CHF seen in these patients.18 In our lit-
  heart failure 1 (0.7) erature review, 6 (26.1%) of 23 patients with a clear out-
come reported, and who presented with CHF, died, which
(continued)
represents a significantly higher mortality rate compared

348 J Neurosurg: Pediatrics / Volume 12 / October 2013

Pial arteriovenous fistulas in the pediatric population

with the 5 (4.7%) of 107 who died but presented without parenchymal varix in childhood: case reports. Neurosurgery
CHF symptoms (p = 0.004, Fisher exact test). This find- 26:122–125, 1990
ing would support the notion that CHF at presentation is   6.  Batista LL, Mahadevan J, Sachet M, Husson B, Rasmussen J,
Alvarez H, et al: Encephalocraniocutaneous lipomatosis syn-
predictive of a worse prognosis in pediatric patients with drome in a child: association with multiple high flow cerebral
PAVFs. arteriovenous fistulae. Case report and review. Interv Neuro-
Our study is limited by the number of cases and its radiol 8:273–283, 2002
retrospective nature, which makes it difficult to form   7.  Beltramello A: Endovascular treatment of cerebrovascular pa-
definitive conclusions about the pathogenesis, clinical thology, in Raimondi AJ, Choux M, DiRocco C (eds): Cere-
course, and optimal treatment strategy for pediatric pa- brovascular Diseases in Children. New York: Springer-Ver-
tients with PAVF. Our review of the literature is subject to lag, 1991, pp 132–158
  8.  Carrillo R, Carreira LM, Prada J, Rosas C, Egas G: Giant an-
reporting bias and potentially represents a more historical eurysm arising from a single arteriovenous fistula in a child.
perspective rather than a current account of this disease Case report. J Neurosurg 60:1085–1088, 1984
and its treatment.   9.  Coubes P, Humbertclaude V, Rodesch G, Lasjaunias P, Echenne
B, Frerebeau P: Total endovascular occlusion of a giant direct
arteriovenous fistula in the posterior fossa in a case of Rendu-
Conclusions Osler-Weber disease. Childs Nerv Syst 12:785–788, 1996
10.  Eerola I, Boon LM, Mulliken JB, Burrows PE, Dompmartin
This study demonstrates both the serious nature of A, Watanabe S, et al: Capillary malformation-arteriovenous
PAVF and its heterogeneous presentation. Aggregated malformation, a new clinical and genetic disorder caused by
data from the literature suggest that a younger age or pres- RASA1 mutations. Am J Hum Genet 73:1240–1249, 2003
ence of ICH or CHF at presentation may signify a worse 11.  Fiser DH: Assessing the outcome of pediatric intensive care. J
prognosis. Furthermore, older patients seem to present Pediatr 121:68–74, 1992
with more frequent ICH, whereas younger patients pre- 12.  Fiser DH, Long N, Roberson PK, Hefley G, Zolten K, Brodie-
sented significantly more often in CHF. The treatment of Fowler M: Relationship of pediatric overall performance cat-
PAVFs has greatly changed with advanced endovascular egory and pediatric cerebral performance category scores at
pediatric intensive care unit discharge with outcome measures
techniques. As shown in our series and with a review of collected at hospital discharge and 1- and 6-month follow-up
the literature, PAVFs are typically amenable to endovas- assessments. Crit Care Med 28:2616–2620, 2000
cular intervention; however, they may require more than 13.  García-Monaco R, De Victor D, Mann C, Hannedouche A,
one type of embolization material to achieve complete Terbrugge K, Lasjaunias P: Congestive cardiac manifestations
obliteration. As with many complex medical conditions, from cerebrocranial arteriovenous shunts. Endovascular man-
pediatric patients with PAVFs require the input of a team agement in 30 children. Childs Nerv Syst 7:48–52, 1991
of physicians in varying disciplines—neurosurgery, inter- 14.  García-Mónaco R, Taylor W, Rodesch G, Alvarez H, Burrows
P, Coubes P, et al: Pial arteriovenous fistula in children as pre-
ventional radiology, neurology, and pediatrics—to devel- senting manifestation of Rendu-Osler-Weber disease. Neuro-
op and implement an effective treatment plan. radiology 37:60–64, 1995
15.  Guerra LR, Barbosa L de A, Barbosa LG de A, Pimentel D de
Disclosure P, Leite FIS: Pial arteriovenous fistula in the posterior fossa.
Arq Neuropsiquiatr 69:718–719, 2011
The authors report no conflict of interest concerning the mate- 16.  Halbach VV, Higashida RT, Hieshima GB, Hardin CW, Dowd
rials or methods used in this study or the findings specified in this CF, Barnwell SL: Transarterial occlusion of solitary intracere-
paper. bral arteriovenous fistulas. AJNR Am J Neuroradiol 10:747–
Author contributions to the study and manuscript preparation 752, 1989
include the following. Conception and design: Heuer, Madsen, Lang, 17.  Hermier M, Turjman F, Bozio A, Duquesnel J, Lapras C: En-
Storm, Hurst. Acquisition of data: Madsen, Lang, Pisapia, Hurst. dovascular treatment of an infantile nongalenic cerebral ar-
Analysis and interpretation of data: Heuer, Madsen, Lang, Pisapia, teriovenous fistula with cyanoacrylate. Childs Nerv Syst 11:
Hurst. Drafting the article: Heuer, Madsen, Lang, Storm, Hurst. Crit- 494–498, 1995
ically revising the article: all authors. Reviewed submitted version of 18.  Hetts SW, Keenan K, Fullerton HJ, Young WL, English JD,
manuscript: all authors. Approved the final version of the manuscript Gupta N, et al: Pediatric intracranial nongalenic pial arterio-
on behalf of all authors: Heuer. Statistical analysis: Madsen, Lang. venous fistulas: clinical features, angioarchitecture, and out-
Study supervision: Heuer, Storm, Hurst. comes. AJNR Am J Neuroradiol 33:1710–1719, 2012
19.  Heuer GG, Gabel B, Beslow LA, Stiefel MF, Schwartz ES,
References Storm PB, et al: Diagnosis and treatment of vein of Galen an-
eurysmal malformations. Childs Nerv Syst 26:879–887, 2010
  1.  Aguilar M, González A, López A, Gutiérrez I, Durand F, May- 20.  Hoh BL, Putman CM, Budzik RF, Ogilvy CS: Surgical and
ol A: Endovascular treatment of a pial arteriovenous fistula endovascular flow disconnection of intracranial pial single-
with occipital remodeling secondary to giant torcular dilation. channel arteriovenous fistulae. Neurosurgery 49:1351–1364,
J Child Neurol 26:1015–1020, 2011 2001
  2.  Almeida GM, Shibata MK: Hemispheric arteriovenous fistulae 21.  Hung PC, Wang HS: Successful endovascular treatment of ce-
with giant venous dilation. Childs Nerv Syst 6:216–219, 1990 rebral arteriovenous fistula. Pediatr Neurol 27:300–302, 2002
  3.  Antunes JL, DiGiacinto GV, Michelsen WJ: Giant hemispheric 22.  Izzo R, Diano AA, Lavanga A, Vassallo P, Muto M: Posterior
arteriovenous fistula in an infant. Surg Neurol 7:45–48, 1977 fossa arteriovenous pial fistula: diagnostic and endovascular
  4.  Aoki N, Sakai T, Oikawa A: Intracranial arteriovenous fistula therapeutic features. Neuroradiolog J 19:783–786, 2006
manifesting as progressive neurological deterioration in an in- 23.  Kikuchi K, Kowada M, Sasajima H: Vascular malformations
fant: case report. Neurosurgery 28:619–623, 1991 of the brain in hereditary hemorrhagic telangiectasia (Rendu-
  5.  Barnwell SL, Ciricillo SF, Halbach VV, Edwards MS, Cogen Osler-Weber disease). Surg Neurol 41:374–380, 1994
PH: Intracerebral arteriovenous fistulas associated with intra- 24.  Lasjaunias P, Manelfe C, Chiu M: Angiographic architecture

J Neurosurg: Pediatrics / Volume 12 / October 2013 349

 P. J. Madsen et al.

of intracranial vascular malformations and fistulas—prethera- 43.  Tabatabai SAF, Zadeh MZ, Habibi Z, Meybodi AT, Hashemi
peutic aspects. Neurosurg Rev 9:253–263, 1986 M: Intracerebral atypical calcification in nongalenic pial arte-
25.  Lasjaunias PL, Chng SM, Sachet M, Alvarez H, Rodesch G, riovenous fistula: a case report. Cases J 1:335, 2008
Garcia-Monaco R: The management of vein of Galen aneu- 44.  Talamonti G, Versari PP, D’Aliberti G, Villa F, Fontana RA,
rysmal malformations. Neurosurgery 59 (3 Suppl 3):S3184– Collice M: Complex arteriovenous fistula of the brain in an
S3194, 2006 infant. Case report. J Neurosurg Sci 41:337–341, 1997
26.  Lee JS, Oh CW, Bang JS, Kwon OK, Hwang G: Intracranial pial 45.  Tamhankar MA, Liu GT, Young TL, Sutton LN, Hurst RW:
arteriovenous fistula presenting with hemorrhage: a case report. Acquired, isolated third nerve palsies in infants with cerebro-
J Cerebrovasc Endovasc Neurosurg 14:305–308, 2012 vascular malformations. Am J Ophthalmol 138:484–486,
27.  Lee JY, Son YJ, Kim JE: Intracranial pial arteriovenous fistu- 2004
las. J Korean Neurosurg Soc 44:101–104, 2008 46.  Tawil I, Carlson AP, Taylor CL: Acute respiratory distress
28.  Limaye US, Siddhartha W, Shrivastav M, Anand S, Ghatge S: syndrome after onyx embolization of arteriovenous malfor-
Endovascular management of intracranial pial arterio-venous mation. Crit Care Res Pract 2011:918185, 2011
fistulas. Neurol India 52:87–90, 2004 47.  Tomlinson FH, Rüfenacht DA, Sundt TM Jr, Nichols DA, Fode
29.  Loh Y, Duckwiler GR: A prospective, multicenter, random- NC: Arteriovenous fistulas of the brain and the spinal cord. J
ized trial of the Onyx liquid embolic system and N-butyl cya- Neurosurg 79:16–27, 1993
noacrylate embolization of cerebral arteriovenous malforma- 48.  Upchurch K, Feng L, Duckwiler GR, Frazee JG, Martin NA,
tions. Clinical article. J Neurosurg 113:733–741, 2010 Viñuela F: Nongalenic arteriovenous fistulas: history of treat-
30.  Lownie SP, Duckwiler GR, Fox AJ, Drake CG: Endovascular ment and technology. Neurosurg Focus 20(6):E8, 2006
therapy of nongalenic cerebral arteriovenous fistulas, in Vi- 49.  Velat GJ, Reavey-Cantwell JF, Sistrom C, Smullen D, Fau­
ñuela F, Halbach VV, Dion JE (eds): Interventional Neuro- theree GL, Whiting J, et al: Comparison of N-butyl cyanoac-
radiology: Endovascular Therapy of the Central Nervous rylate and onyx for the embolization of intracranial arterio-
System. New York: Raven Press, 1992, pp 87–106 venous malformations: analysis of fluoroscopy and procedure
31.  Lv X, Jiang C, Li Y, Yang X, Wu Z: Clinical outcomes of endo- times. Neurosurgery 63 (1 Suppl 1):ONS73–ONS80, 2008
vascular treatment for intracranial pial arteriovenous fistulas. 50.  Viñuela F, Drake CG, Fox AJ, Pelz DM: Giant intracranial
World Neurosurg 73:385–390, 2010 varices secondary to high-flow arteriovenous fistulae. J Neu-
32.  Lv X, Li Y, Jiang C, Wu Z: Endovascular treatment of brain rosurg 66:198–203, 1987
arteriovenous fistulas. AJNR Am J Neuroradiol 30:851–856, 51.  Wang YC, Wong HF, Yeh YS: Intracranial pial arteriovenous
2009 fistulas with single-vein drainage. Report of three cases and
33.  Martin NA, Macagba-Crain CL, Geffner M, Peacock W: Iso- review of the literature. J Neurosurg 100 (2 Suppl Pediat-
lated growth hormone deficiency associated with a giant arte- rics):201–205, 2004
riovenous varix. Neurosurgery 27:295–299, 1990 52.  Weon YC, Yoshida Y, Sachet M, Mahadevan J, Alvarez H,
34.  McSweeney N, Brew S, Bhate S, Cox T, Roebuck DJ, Ganesan Rodesch G, et al: Supratentorial cerebral arteriovenous fistu-
V: Management and outcome of vein of Galen malformation. las (AVFs) in children: review of 41 cases with 63 non cho-
Arch Dis Child 95:903–909, 2010 roidal single-hole AVFs. Acta Neurochir (Wien) 147:17–31,
35.  Nakiri GS, Abud TG, Oliveira RS, Santos AC, Machado HR, 2005
Abud DG: Endovascular treatment of intracranial pial arterio- 53.  Yang WH, Lu MS, Cheng YK, Wang TC: Pial arteriovenous
venous fistula. Arq Neuropsiquiatr 68:463–465, 2010 fistula: a review of literature. Br J Neurosurg 25:580–585,
36.  Nelson PK, Niimi Y, Lasjaunias P, Berenstein A: Endovascu- 2011
lar embolization of congenital intracranial pial arteriovenous 54.  Yoshida Y, Weon YC, Sachet M, Mahadevan J, Alvarez H,
fistulas. Neuroimaging Clin N Am 2:309–317, 1992 Rodesch G, et al: Posterior cranial fossa single-hole arteriove-
37.  Newman CB, Hu YC, McDougall CG, Albuquerque FC: Bal- nous fistulae in children: 14 consecutive cases. Neuroradiol-
loon-assisted Onyx embolization of cerebral single-channel ogy 46:474–481, 2004
pial arteriovenous fistulas. Technical note. J Neurosurg Pe- 55.  Youn SW, Han MH, Kwon BJ, Kang HS, Chang HW, Kim
diatr 7:637–642, 2011 BS: Coil-based endovascular treatment of single-hole cerebral
38.  Oyesiku NM, Gahm NH, Goldman RL: Cerebral arteriove- arteriovenous fistulae: experiences in 11 patients. World Neu-
nous fistula in the Klippel-Trenaunay-Weber syndrome. Dev rosurg 73:2–10, 2010
Med Child Neurol 30:245–248, 1988 56.  Zuccaro G, Argañaraz R, Villasante F, Ceciliano A: Neuro-
39.  Pamuk AG, Saatci I, Cekirge HS, Aypar U: A contribution to surgical vascular malformations in children under 1 year of
the controversy over dimethyl sulfoxide toxicity: anesthesia age. Childs Nerv Syst 26:1381–1394, 2010
monitoring results in patients treated with Onyx embolization
for intracranial aneurysms. Neuroradiology 47:380–386, 2005
40.  Paramasivam S, Toma N, Niimi Y, Berenstein A: Develop-
ment, clinical presentation and endovascular management of
congenital intracranial pial arteriovenous fistulas. J Neuroin-
terv Surg 5:184–190, 2013 Manuscript submitted March 6, 2013.
41.  Pillai A, Rajeev K, Unnikrishnan M: Surgical management Accepted June 20, 2013.
of a pial arteriovenous fistula with giant varix in an infant. Please include this information when citing this paper: pub-
Neurol India 54:434–436, 2006 lished online August 2, 2013; DOI: 10.3171/2013.6.PEDS13110.
42.  Rivera R, Blanc R, Piotin M, Spelle L, Moret J: Single hole Address correspondence to: Gregory G. Heuer, M.D., Ph.D.,
cerebral arteriovenous fistula between the anterior choroidal Division of Neurosurgery, The Children’s Hospital of Philadel-
artery and the basal vein of Rosenthal in a child. Childs Nerv phia, 34th St. & Civic Center Blvd., Philadelphia, PA 19104.
Syst 25:1521–1523, 2009 email: [email protected].

350 J Neurosurg: Pediatrics / Volume 12 / October 2013