Article Text

Download PDFPDF

Original research
Vein of Galen aneurysmal malformation—clinical and angiographic spectrum with management perspective: an institutional experience
  1. Himanshu Agarwal,
  2. Leve Joseph Devarajan Sebastian,
  3. Shailesh B Gaikwad,
  4. Ajay Garg,
  5. Nalini K Mishra
  1. Department of Neuroimaging and Interventional Neuroradiology, CN Centre, All India Institute of Medical Sciences, New Delhi, India
  1. Correspondence to Dr Leve Joseph Devarajan Sebastian, Department of Neuroimaging and Interventional Neuroradiology, CN Centre, All India Institute of Medical Sciences, New Delhi110029, India; leve_s{at}


Background and purpose Vein of Galen aneurysmal malformation (VGAM) is a rare developmental intracranial vascular malformation. We analyzed the clinical presentations, imaging findings, angioarchitecture, management options, and outcome in a demographically heterogeneous set of VGAM patients.

Methods We retrospectively analyzed cases of VGAM from our departmental archive collected between 1988 and January 2015. Demographic, clinical, therapeutic, and follow-up details were obtained for each patient from the available records.

Results We identified 36 patients with VGAM including 6 neonates, 18 infants, 7 children aged 2–10 years, and 5 adults. Macrocrania was the commonest presenting feature. Type of fistulae was mural in 14 and choroidal in 18 patients while 4 had a thrombosed sac at presentation. In 3 cases the dilated venous sac had connection with the deep venous system. Bilateral jugular atresia and stenosis were seen in 9 and 6 patients, respectively. Giant venous sac (>4 cm) was significantly correlated with mural type (p=0.0001). Dural arterial recruitment was seen in 4 patients including 3 adults. Among the 23 patients treated by endovascular means, 14 had a good outcome, 5 had a poor outcome, and 4 died. A significant correlation was noted between jugular atresia and poor outcome (p=0.003).

Conclusions We encountered a wide range of demographic, clinical, and angiographic features in VGAM. Mural type malformations were associated with giant venous sacs. Good outcome after embolization was seen in selected neonates and in most of the infants, children, and adults. Jugular atresia was significantly associated with poor outcome.

  • Arteriovenous Malformation
  • Pediatrics
  • Congenital
  • Intervention

Statistics from


Vein of Galen aneurysmal malformation (VGAM) is a rare intracranial developmental vascular malformation.1 The insult to developing neurovasculature leading on to VGAM occurs between 6 and 11 weeks of embryonic life.2 These lesions are characterized by the presence of multiple abnormal arteriovenous communications in the developing choroid leading to dilation and persistence of the median vein of prosencephalon which is considered to be the precursor of the vein of Galen.2 Most of the patients reported in the literature are neonates, while only a few series have reported a sizeable number of older children or adults with VGAM. Optimal management of VGAM requires proper understanding of its embryology and pathophysiology. Knowledge of the angioarchitecture and hemodynamics in a given patient is also essential. Various treatment options including surgery, endovascular treatment, and radiosurgery have been tried worldwide. Of these, endovascular treatment has emerged as the treatment of choice with favorable results and fewer complications.3–6

Because of its rarity, most series of VGAM, except probably that of Lasjaunias et al, contain a relatively small number of patients. In this context, we would also like to share our experience with a small series of 36 patients with VGAM. More importantly, we wish to highlight the heterogeneous demography of our patient population with a significant proportion of adults. The range of angioarchitectural features accompanying this demographic heterogeneity could indicate the existence of a spectrum or continuum at various levels of the VGAM evolution—namely, the timing of the insult to developing neurovasculature and resultant angioarchitecture, the adaptations of the host and hence the clinical manifestations—and our study seeks to review these aspects.

Materials and methods

The study was approved by the institutional ethical committee of our hospital. The neurovascular disease database in our department was searched to identify patients with VGAM in the period between 1988 and June 2015. Data on demographics, clinical presentation, neurological deficits, and overall clinical condition at presentation, at the time of discharge from the hospital, and on follow-up were collected from medical records. Cross-sectional imaging (CT and MRI) were assessed for the presence of hydrocephalus, parenchymal calcification, and cerebral atrophy. Digital subtraction angiograms of each patient were reviewed independently by two neuroradiologists (NKM and LJDS) to characterize the morphology and functional angioarchitecture of the VGAM. Important points included in our analysis were type of malformation (choroidal or mural), arterial feeders (anterior choroidal artery (AchA), posterior choroidal artery (PchA), anterior cerebral artery (ACA), thalamoperforators, dural arterial recruitment), size of venous sac (enlarged venous sac draining the fistula was measured in the largest dimension), status of deep venous system (deep venous connection with venous sac, alternate drainage of deep venous system), status of jugular veins (patent, stenosed, atresia), status of fetal sinuses (falcine, occipital sinus), and cavernous sinus.

Records were reviewed for type of endovascular therapy instituted, route used (arterial, venous or both), embolic material used, percentage of occlusion achieved, angiographic outcome of procedure, and post-embolization clinical status.


Of the 36 patients with VGAM identified (27 male, 9 female), 6 were neonates, 18 were infants, 7 were aged 2–10 years (children), 1 was in the second decade, and 4 were aged >20 years. The mean age was 5 years and the median age was 1 year. The oldest case was aged 38 years. The only patient in the second decade was 18 years, hence there were 5 adult patients in the series. Two neonates had been diagnosed antenatally.

The most common presentation in our series of VGAM was increasing head size (n=13), followed by developmental delay (n=10) and seizures (n=8). These symptoms were seen beyond neonatal age, predominantly in infants and early childhood. The mean age for macrocrania was 12 months. Four neonates presented with congestive heart failure. One apparently asymptomatic antenatally diagnosed neonate was found to have isolated severe pulmonary arterial hypertension on a routine follow-up evaluation at 3 weeks in a private hospital and was referred to us for further management. Later on the child's mother acknowledged mild feeding difficulty in the initial weeks compared with the longer feeding times after treatment. One neonate presented with feeding difficulties and had melting brain syndrome. Headache and vomiting were the most common presenting symptoms among adults. One pre-teen child had vision loss and another child of the same age had visual blurring. Epistaxis, diplopia, and proptosis were usually seen beyond the infant age group (table 1).

Table 1

Clinical presentation in different age groups

On cross-sectional imaging, hydrocephalus seen in 28 (76%) patients was the most common finding across all the age groups. Cerebral atrophy and parenchymal calcification were seen in 6 (16%) and 10 (27%) cases, respectively. Intraventricular hemorrhage (IVH) at presentation was seen in two cases. Diffuse white matter signal change was noted in four cases, three of whom were found to have jugular atresia also. Four patients had thrombosed venous sac on first referral. This was confirmed by transcranial sonography in two infants; the other two underwent cerebral angiography. The thrombosed venous sac measured >4 cm in diameter in all of these cases.

Thirty-four patients underwent cerebral angiography; VGAM was choroidal type in 18 cases and mural type in 14. Both types were almost equally distributed in the different age groups, except in patients aged >20 years in whom only the choroidal type was seen. Three patients with both large mural type single hole fistula and multiple choroidal type fistulae were finally classified as choroidal. Two patients with a thrombosed venous sac had no evidence of fistulae.

Among the different groups of arterial feeders, all patients had the posterior choroidal artery as one of the feeding arteries. The anterior choroidal artery was the feeder in 39% of cases (n=13), ACA in 12% (n=4), and thalamoperforators in 12% (n=4). Dural feeders were seen in 12% of cases (n=4), which included three adults.

The largest dimension of aneurysmal venous sac was measured in each case and a sac of ≥4 cm was defined as ‘giant’. There was a significant association between the mural type of nidus (12/14) and giant venous sac (p=0.0001, Fisher exact test).

Persistence of the falcine sinus was seen in 27 cases (82%) while eight (24%) had a persistent occipital sinus. Only seven patients had a straight sinus, four of whom were aged >10 years. Two patients had both a persistent falcine sinus and straight sinus, forming a falcine loop. Cavernous sinus re-routing was seen in 20 cases (65%). No neonate had cavernous sinus drainage. Deep venous connection with the aneurysmal sac was seen in three cases. Jugular venous stenosis and atresia were seen in six and nine cases, respectively. All adult cases (>10 years) had patent jugular veins. None of the patients with patent jugular veins had parenchymal calcifications (p<0.05).

Twenty-four patients received endovascular treatment; 11 had a choroidal type of nidus and 13 had a mural type of nidus. All except one were embolized using the arterial route, with glue (N-butyl cyanoacrylate) being the embolic agent. One patient was embolized through both the arterial and venous routes using coils and glue. Fifteen patients had complete occlusion of fistulae, 10 of which were of mural type. In one infant, who presented with IVH, the major feeder went into spasm during superselective navigation with a microcatheter so the procedure was abandoned. The patient later improved spontaneously, has been achieving milestones and is stable except for minimal hemiparesis on 3-year follow-up. Retreatment is being deferred since the parents are apprehensive.

Endovascular treatment was withheld in four patients due to poor clinical condition, presence of bilateral jugular atresia, and/or extensive brain parenchymal changes. Five patients were conservatively managed, four of whom had thrombosed sac at presentation. The other one was a neonate who presented with mild heart failure and was treated successfully by medical means and followed up until 2 years of age when his first angiogram was done. The angiogram showed choroidal type of malformation with patent jugular veins. No hydrocephalus or brain parenchymal changes were seen on imaging. It was therefore decided to follow the child conservatively. However, further long-term follow-up records are not available for this patient. No treatment or follow-up records were available for three patients including one adult patient who declined treatment (table 2).

Table 2

Management summary of cases

Five patients underwent ventriculoperitoneal (VP) shunt. In two patients the shunt was done before the embolization and in two after the embolization. One patient with spontaneous thrombosis underwent VP shunt.

Among the 23 embolized cases, four patients died and the mean follow-up period in the remainder was 31 months, ranging from 6 months to 16 years. Fourteen patients had normal development with improvement in clinical symptoms or were stable with no new neurological symptoms. One patient with vision loss, headache and vomiting had improvement in headache and vomiting with no improvement in vision. One infant with mural type malformation had persistent hypertension after embolization which eventually settled spontaneously. One neonate with isolated pulmonary arterial hypertension and feeding difficulties had a significant drop in her pulmonary arterial pressure with improved feeding after embolization. Five patients did not improve despite good angiographic occlusion of fistulae. Four of them had bilateral jugular atresia and one had bilateral jugular stenosis. Post-procedural mortality was either due to IVH or subarachnoid hemorrhage. Three out of five neonates developed post-embolization IVH and died. All of them had complete occlusion of the sac at embolization. One adult patient in whom the venous sac was occluded by coils and glue developed IVH 8 h after embolization and died (figure 1). Out of nine patients with poor outcome (died or not improved), five had bilateral jugular atresia (p=0.003). One infant developed femoral arterial thrombosis. Two patients developed intraventricular bleeding after VP shunt, which resolved without clinical consequences (tables 3 and 4).

Table 3

Clinical, angiographic, outcome and follow-up data of embolized cases

Table 4

Outcome of embolized cases

Figure 1

An adult patient who presented with headache. (A) Plain CT showing dilated venous sac in the quadrigeminal cistern (black star). (B,C) Vertebral angiogram showing choroidal malformation opening in the venous sac, draining into the falcine sinus (black arrow) with reflux into the deep venous system (double arrows). (D) The venous sac was occluded with coils and glue (not shown). (E) Post-embolization angiogram showing no evidence of fistulae. (F) Plain CT scan taken 8 h after the procedure showing massive intraventricular bleed.

All five conservatively managed patients were clinically stable with normal development until their last follow-up ranging from 6 months to 4 years. In two cases of thrombosed sac with longer follow-up, imaging demonstrated shrinkage of the venous sac (figure 2).

Figure 2

An infant. (A) Flair MR and (B) plain CT images showing thrombosed aneurysmal venous sac in the quadrigeminal cistern draining into the falcine sinus. (C,D) MRI after 2 years showing completely thrombosed and shrunken venous sac.


A dearth of neonatal cases and hence the apparent clustering of cases in infancy and early childhood were the peculiar features in our series, in contrast to most of the western series5–7 which comprise mostly neonates. Another series from India by Gupta et al8 also reported a similar trend of clustering of cases in infants and early childhood. Factors which may account for this phenomenon include a poor rate of antenatal diagnosis and neonates with VGAM rarely surviving until they reach a specialized center such as ours. Adults constituted 13.9% of our cases, which is relatively high compared with other series available in the literature. Only some anecdotal case reports have discussed adult patients with VGAM, most of them prior to 2000. Angioarchitectural features (discussed later) differing from the classical VGAM described in children might be responsible for underreporting of this malformation in adults. On the other hand, Alvarez et al9 excluded cases above 16 years of age from their review, for they might have thought adult VGAM as a different entity.

Clinical presentation

Clinical presentation in various age groups in this series was in accordance with the known natural history of VGAM, which is largely determined by the changing or evolving venous angioarchitecture. Cardiac manifestations were seen exclusively in neonates. One of the neonates in our series presented with isolated pulmonary hypertension and mild feeding difficulties which improved after the treatment of VGAM. Pulmonary hypertension without overt congestive heart failure has seldom been reported in the literature.

In infants and children, the brain bears the brunt of the hemodynamic effects of VGAM. The ensuing hydrovenous disorder results in macrocrania, developmental delay, and seizures.9 Accordingly, these rather than cardiac failure were the most common presentations in our series due to its different demographic composition (as mentioned above). Other manifestations such as epistaxis and proptosis could be explained by venous re-routing. Interestingly, one pre-teen child with a choroidal type of malformation suffered rapid visual deterioration in the 2-month waiting period for a planned embolization. Ophthalmologic examination showed bilateral secondary optic atrophy. Longstanding venous hypertension could probably explain our case, although such vision loss is rarely reported in patients with VGAM.

Headache and vomiting were the most common symptoms in adults. Headache was the commonest and sometimes the only presenting complaint in the occasional case reports of adult VGAM found in the literature. Vomiting was also reported in some cases.10 ,11 Xu et al12 reported an adult patient with VGAM who presented with vertigo and dizziness.


Yasargil classified VGAM into four types according to angioarchitecture.13 Later Lasjaunias and colleagues divided VGAM into two main types—choroidal and mural—which is widely accepted.14 They did not consider Yasargil type 4 as true VGAM. Three of our cases which we classified as choroidal had one large feeder much like mural malformation as well as myriad small feeders as seen in the choroidal type. These types can be labeled as mixed types (Yasargil type 3).13

Venous sac size

We explored the relation between venous sac size and type of malformation and found a significant association. Twelve (86%) out of 14 patients with mural type VGAM had sac size ≥4 cm (p=0.0001). No case series in the available literature has mentioned such an association. Alvarez et al9 stated that dilation of the venous sac is variable and is unrelated to the architecture of a choroidal or mural VGAM. Although the number of cases in our series is smaller, we feel this observation is significant and needs further enquiry. One possible mechanism is the occurrence of early venous outlet stenosis secondary to high flow jet from the usually enlarged arteries in the mural type of malformation.

Deep venous connection

The fact that we noted deep venous reflux in three cases indicates the presence of a deep venous connection with the venous sac in VGAM. Similar findings were also reported by Levrier et al15 and Iizuka et al.16 In fact, this connection could be the reason for the fatal IVH in our cases where the venous sac was rapidly occluded. Lasjaunias et al4 denied the possibility of a deep venous connection of the venous sac. However, Raybaud in his description of the embryology of VGAM emphasized that deep venous connections could occur in certain cases in which timing of the insult is relatively delayed.3 Thus, our findings support the notion that insult to the developing neurovasculature can lead on to a spectrum of morphologic features depending on the timing of the insult.

Adult angioarchitecture

Straight sinus and patent jugular veins were found more frequently in the adult patients in our series. These could, in fact, be the very factors which permitted their uneventful survival to adulthood. The significance of predominantly choroidal type in these patients needs to be explored because, according to Lasjaunias et al, choroidal is the more severe malformation found commonly in neonates; however, they did not include patients above 16 years of age in their large series.9 Dural arterial recruitment was another interesting feature in our adult patients. Berenstein et al also found that seven out of 71 patients with VGAM had dural arterial feeders and they were older than the other patients at the time of their initial angiogram. They attribute the dural supply to non-sprouting angiogenesis due to sump effect.17


Endovascular treatment is the mainstay of VGAM treatment. The therapeutic goal differs in different age groups. In neonates the goal is to reduce the shunt to such a level that heart failure becomes medically manageable. However, it is difficult to identify patients who will benefit from endovascular intervention. In our series all three neonates who died in the immediate post-embolization period had presented with severe clinical features. Complete rapid occlusion of the malformation in a single sitting in each of these cases might also have precipitated IVH. In two other neonates, embolization played a decisive role in altering the clinical course favorably. In contrast, medical control of cardiac failure alone enabled the other neonate in our series to reach childhood with normal growth and milestones. The three different scenarios described above indicate the necessity of a robust scoring system like the Bicetre score of Lasjaunias et al to triage such cases. The utility and feasibility of such a scoring system for prospective application in different clinical settings needs to be further studied. Yet the neonatal VGAM referrals to our center were so few that no such criteria for intervention could be examined or evolved. One of the neonates in our series who presented with severe isolated pulmonary hypertension needs special mention because, to the best of our knowledge, no management precedence for a similar case is available in the literature. Applying the Bicetre score would have mandated a conservative approach for this case, while Berenstein et al18 indicated that early echocardiographic identification of high pulmonary artery pressure could necessitate early intervention. After discussion with the pediatric cardiologist, consensus was reached to treat this patient by embolization. The patient's pulmonary artery pressure dropped significantly after the procedure with improved feeding time according to the mother.

In the infant age group the primary aim is effective management of the hydrovenous disorder secondary to VGAM so that the infant attains normal neurologic development and clinical milestones. Embolization benefited 55% of children in this age group in our series. Significantly, all the infants who deteriorated had bilateral jugular atresia or jugular stenosis. Jugular atresia is an independent poor prognostic factor which reached statistical significance in the infant age group and also in the whole study group. Lasjaunias et al4 have also stressed the role of jugular patency in the prognosis of treated as well as untreated cases of VGAM and indicated that evolving significant jugular stenosis called for urgent endovascular intervention. However, Geibprasert et al19 did not find any significant correlation between jugular stenosis and clinical outcome. This is probably because their series constituted mainly neonates. This also implies that jugular stenosis perhaps prevents congestive heart failure and allows the child to survive to infancy.

In children, management of hydrocephalus and symptoms such as epistaxis secondary to venous re-routing are of major concern. Sometimes, life-threatening IVH can also occur in these children. The strategy and risks of endovascular management is similar to other non-Galenic malformations. All the children in this age group in our series did relatively well after embolization as they had relatively good brain parenchymal status and none had jugular atresia.

In adults, multiple sessions may be required to eliminate the fistulae since dural arteries are also recruited. In one of our adult patients an overzealous attempt at eliminating the malformation in one go by packing the sac with coils through the venous route and injecting glue from one of the dural feeders resulted in a disastrous IVH hours after the embolization. We therefore strongly discourage rapid complete occlusion of the venous sac.


Our study demonstrates that there could be a wide range of demographic, clinical, and angiographic spectrum in VGAM. The mural type of malformation tends to have giant venous sacs. Management of VGAM should be individualized based on clinical, brain parenchymal, and angiographic features. Good outcome after embolization can be seen in selected neonates and in most infants, children, and adults. Jugular atresia is significantly associated with a poor outcome. Rapid occlusion of the venous sac should be avoided.


Sachin A Borkar and Pankaj Kumar Singh, Assistant Professor, Department of Neurosurgery, All India Institute of Medical Sciences.



  • Contributors HA: acquisition, analysis and interpretation of data; drafting the work and revising it critically for important intellectual content. LJDS: conception and design of the work, analysis and interpretation of data; drafting the work and revising it critically for important intellectual content. SBG: design of the work and revising it critically for important intellectual content. AG: design of the work and revising it critically for important intellectual content. NKM: conception of the work, analysis and interpretation of data; drafting the work and revising it critically for important intellectual content. All authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

  • Competing interests None declared.

  • Ethics approval The study was approved by the Ethics Committee for Post Graduate Research, All India Institute of Medical Sciences.

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

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.