Secondary Logo

Journal Logo

Original Investigations

Dilated Superior Ophthalmic Vein: Clinical and Radiographic Features of 113 Cases

Adam, Christopher R. M.D.*; Shields, Carol L. M.D.; Gutman, Justin M.D.*; Kim, H. Joon M.D.; Hayek, Brent M.D.; Shore, John W. M.D.§; Braunstein, Alexandra M.D.; Levin, Flora M.D.; Winn, Bryan J. M.D.; Vrcek, Ivan M.D.#; Mancini, Ronald M.D.#; Linden, Craig M.D.**; Choe, Christina M.D.††; Gonzalez, Mithra M.D.‡‡; Altschul, David M.D.§§; Ortega-Gutierrez, Santiago M.D.‖‖; Paramasivam, Srinivasan M.D.¶¶; Fifi, Johanna T. M.D.##; Berenstein, Alejandro M.D.***; Durairaj, Vikram M.D.§; Shinder, Roman M.D., F.A.C.S.

Author Information
Ophthalmic Plastic and Reconstructive Surgery: January/February 2018 - Volume 34 - Issue 1 - p 68-73
doi: 10.1097/IOP.0000000000000872
  • Free

Advances in radiographic imaging modalities and techniques have revolutionized the work-up of ophthalmic disease and are an indispensable adjunct to the clinical examination. Dilated superior ophthalmic vein (SOV) is a rare radiologic finding that encompasses a wide variety of diseases with outcomes ranging from benign to sight-threatening and life-threatening endpoints. Dilated SOV has been previously reported in patients with carotid-cavernous fistula, cavernous sinus thrombosis,1 SOV thrombosis,2 thyroid eye disease,3 idiopathic orbital inflammation (IOI),4 increased intracranial pressure,5 malignancy,2 postoperative,6 and many others.7

Paramount to making a diagnosis and understanding the implications of a dilated SOV is a thorough knowledge of orbital and cerebral vascular anatomy. The venous anatomy of the orbit is complex, demonstrates considerable variability, and may impact the clinical presentation and surgical management of ophthalmic disease.8,9 The SOV is the largest venous channel in diameter and serves as the principal route of venous drainage of the orbit. Anteriorly, the SOV is embedded within the retrobulbar fat and supported by connective tissue septae. Posteriorly, the SOV is anchored by a fibrous sling at the superior orbital fissure. However, in the posterior orbit, in proximity to the lateral orbital wall, the SOV is unsupported and thus subject to displacement and compression.9 Any deviation of this course, in the direction of venous flow, outflow obstruction, mechanical compression, or hydrodynamic changes within the orbital vascular system or cerebrospinal fluid, may result in dilatation of the SOV as it becomes congested and engorged with blood.

Radiologic identification and characteristics of the SOV using thin-section high-resolution CT and MRI have been previously described.10–12 Using high-resolution orbital MRI, Tsutsumi et al. reported a normal mean SOV outer radiographic diameter of 1.7 mm, consistent with previous studies reporting similar mean diameters.10–14 The diameter of the SOV has been reported to be highly variable and asymmetric in up to 75% of individuals, with mean normal diameters ranging from 0.3 mm to 4.6 mm.10,15,16 Previous studies have demonstrated the SOV’s lumen widens as it tracks posteriorly receiving its multitude of tributaries and slightly narrows as the vein passes through the superior orbital fissure.15,17,18 In their study investigating variations in intracranial pressure and its correlation with SOV diameter, Lirng et al.5 reported no correlation between the average SOV diameter and patient age, sex, or body mass index.

Although a comprehensive history and clinical examination are the most important approaches a physician can take when evaluating a patient with a dilated SOV, a thorough understanding of the wide differential diagnosis and prompt recognition of this rare radiographic entity are essential in determining the underlying etiology and providing timely intervention. The authors herein review dilated SOV as a radiographic finding through their collaborative experience with 113 consecutive patients.


This is a retrospective, multi-institutional case series wherein the records of 113 patients with a documented dilated SOV present on radiographic imaging were evaluated between January 2001 and April 2014. Cases contributed by the participating 11 institutions were from ophthalmology, neurosurgery, otolaryngology, ocular oncology, and radiology services. All patient records and imaging were initially evaluated at the host institution and reanalyzed independently by 3 masked neuroradiologists. Inclusion criteria included diagnosis of a dilated SOV by the neuroradiologists and confirmation of the underlying etiology. For purposes of this study, a diagnosis of a dilated SOV was made if the SOV measured greater than 3.0 mm in diameter on CT or MRI orbital imaging and was present in 2 or more contiguous coronal slices imaged at any point along the SOV’s intraorbital course. Only patients that had scans of 1 mm thickness or less were included. Implementing this method provided the reviewing radiologists with high-resolution scans and a degree of standardization across institutions. Additional advanced radiographic modalities performed including CT angiography, magnetic resonance angiography, magnetic resonance venography, digital subtraction angiography (DSA), and ultrasound were used to further assess the dilated SOV and underlying disease-related pathology. Exclusion criteria excluded any patient with enlarged SOV that was found to be a normal variant without evidence of identifiable disease. Patient demographics, presenting symptoms, ophthalmic findings, radiographic modality, diagnosis, treatment, outcomes, and visual acuity (Va) at last follow up were recorded. Visual impairment at initial presentation and last follow up was defined for purposes of this study as best corrected Snellen Va of <20/40 and did not include cases of children less than 5 years of age in whom accurate measurement may not be reliable. Research procedures were approved by the Institutional Review Boards of all collaborating centers and SUNY Downstate Medical Center served as the central IRB institution. The treatment of all patients and data conformed with the tenets set forth in the Declaration of Helsinki, and was performed in accordance with the Health Insurance Portability and Accountability Act of 1996.


There were 113 patients presenting with a dilated SOV on radiographic imaging. Patients consisted of 75 women (66%) and 38 men (34%), and had a mean age of 49.2 ± 24 years (range, 0.4–90 years). The most frequent etiology to cause a dilated SOV was found to be cerebral vascular malformations (n = 92, 81%). Within this group, the most common specific diagnoses were dural-cavernous fistula (DCF, n = 50, 44%) and carotid-cavernous fistula (CCF, n = 21, 19%). Patient demographics and underlying etiologies are represented in Table 1. Patient disease status at last follow up, mean follow-up time, and visual impairment at presentation and last follow up across all disease processes are represented in Table 2. The imaging modalities utilized to diagnose a dilated SOV included MRI (n = 98, 87%), DSA (n = 77, 68%), CT (n = 29, 26%), and ultrasound (n = 4, 4%). A dilated SOV was most frequently initially diagnosed by MRI in cases of vascular malformations and with either CT or MRI in remaining cases. Treatment and follow up were tailored to the underlying disease process. Disease status at last follow up included no evidence of disease (NED, n = 57, 50%), alive with persistent disease (AWD, n = 53, 47%), and expired from disease (EXD, n = 3, 3%), with a mean follow up of 18 months (range, 1 day to 180 months). Visual impairment observed at presentation and last follow up across all cases was 26% and 22%, respectively.

Dilated superior ophthalmic vein in 113 patients: related conditions
Dilated superior ophthalmic vein in 113 patients: disease and vision status

Vascular Malformations.

Dural-Cavernous Fistula.

There were 50 cases (44%) presenting with a dilated SOV that were diagnosed with a DCF. Most of these subjects had known risk factors for the development of a DCF, including hypertension (n = 24, 48%), atherosclerotic vascular disease, pregnancy, and minor trauma. Initial clinical presentations ranged on a spectrum from 2 asymptomatic patients to complaints of decreased vision (n = 15), severe headache (n = 10), diplopia (n = 7), pulsatile tinnitus (n = 5), eye redness (n = 4), eye pain (n = 4), tearing (n = 1), and nausea/vomiting (n = 1). Ophthalmic findings consisted of proptosis (n = 28), chemosis (n = 28), ophthalmoplegia (n = 18), decreased Va (n = 15), ocular hypertension (n = 4), conjunctival injection (n = 4), periorbital swelling (n = 2), and ptosis (n = 1). A dilated SOV was most commonly initially diagnosed on MRI in this subset of cases with additional radiographic imaging consisting of CT, DSA, and ultrasound, which were used to further characterize the SOV and cerebral vasculature. Representative images from this cohort are demonstrated in Figures 1 and 2. The most common treatment method was embolization of the fistula (n = 34, 68%). Additional treatment options included observation (n = 7, 14%), compression maneuvers (n = 4, 8%), and topical glaucoma medication (n = 3, 12%). One case, a 42-year-old woman with history of a large right middle cerebral artery cerebrovascular accident presented with proptosis, dilated episcleral vessels, ocular hypertension, and tortuous retinal vessels, and was found to have bilateral dilated SOVs on CT imaging. No visual impairment was noted on presentation or last follow up. She was treated with embolization and currently shows NED at 9 months.

FIG. 1.
FIG. 1.:
Clinical presentation and imaging characteristics of a dilated superior ophthalmic vein presenting in a patient with a left dural-cavernous fistula. A, Worms-eye clinical photograph showing left proptosis. B, Clinical photograph with digital eyelid retraction demonstrating diffuse dilation and tortuosity of the left episcleral vessels. Contrast enhanced axial (C) and coronal (D) orbital CT demonstrating a dilated left SOV (arrows).
FIG. 2.
FIG. 2.:
Axial (A) and coronal (B) orbital CT demonstrating bilateral dilation of the dilated superior ophthalmic veins (arrows) in a patient with a dural-cavernous fistula.

Carotid-Cavernous Fistula.

There were 21 cases (19%) of study patients diagnosed with a CCF. Many of these subjects had known risk factors for the development of a CCF, including trauma (n = 8, 38%) and dissection or spontaneous rupture of a carotid artery aneurysm. Mechanisms of traumatic injuries were most commonly caused by either direct blunt force to the ipsilateral periocular region or recent history of a motor vehicle collision. All patients were symptomatic at presentation with findings similar to and more pronounced than DCF. Initial clinical presentations included complaints of pulsatile tinnitus (n = 6), eye pain (n = 6), eye redness (n = 5), decreased vision (n = 5), diplopia (n = 4), severe headache (n = 3), nausea/vomiting (n = 3), tearing (n = 1), and ataxia (n = 1). Ophthalmic findings consisted of proptosis (n = 11), ophthalmoplegia (n = 11), periorbital swelling (n = 7), decreased Va (n = 5), conjunctival injection (n = 5), chemosis (n = 3), papilledema (n = 1), and ptosis (n = 1). The mean duration of disease before presentation was 11 months. The imaging modalities utilized to diagnose a dilated SOV and CCF included CT, MRI, ultrasound, and DSA. Representative images from this cohort are demonstrated in Figure 3. The most common treatment was embolization of the fistula (n = 15, 71%). Additional treatment options included observation (n = 5, 24%) and topical glaucoma medication (n = 1, 5%).

FIG. 3.
FIG. 3.:
A, T2-weighted MRI of the orbits demonstrating dilation of the right superior ophthalmic vein (arrow). B, Digital subtraction angiography documenting an anomalous direct connection between the cavernous portion of the internal carotid artery and cavernous sinus (arrow), resulting in a carotid-cavernous fistula.

Orbital Arteriovenous Malformation.

There were 7 cases (6%) diagnosed with an orbital arteriovenous malformation (AVM). All patients were symptomatic at presentation. Initial clinical presentations included complaints of periorbital swelling (n = 6), decreased vision (n = 4), eye pain (n = 2), and eye redness (n = 1). Ophthalmic findings consisted of periorbital swelling (n = 6), decreased Va (n = 4), proptosis (n = 3), chemosis (n = 1), ophthalmoplegia (n = 1), and conjunctival injection (n = 1). A dilated SOV was diagnosed in all cases with MRI with additional DSA confirming the dilated SOV and orbital AVM. The most common treatment was embolization of the AVM (n = 5, 71%), with 2 subjects managed conservatively by observation. One case, a 31-year-old woman who was treated with embolization presented with an initial Va of 20/40 that declined to 20/200 during 9-year follow up.

Facial AVM.

There were 5 cases (4%) diagnosed with a facial AVM. All patients were symptomatic at presentation. Initial clinical presentations included complaints of periorbital swelling (n = 4), oral/nasal bleeding (n = 3), and headache (n = 1). Ophthalmic findings consisted of periorbital swelling (n = 4) and proptosis (n = 1). All cases were diagnosed with MRI, with additional DSA confirming the dilated SOV and facial AVM. All cases underwent treatment by embolization of the AVM. In the single subject who initially presented with visual impairment, visual improvement at last follow up was not observed. This case, a 45-year-old man known to have a facial AVM for over 6 years presented with proptosis, oral bleeding, and an initial Va of 20/60. He experienced progressively worsening Va to 20/400 at 32-month follow up despite undergoing embolization of the AVM.

Venous Thrombosis.

Superior Ophthalmic Vein Thrombosis.

There were 6 cases (5%) diagnosed with an SOV thrombosis. Past medical histories included myelodysplastic syndrome, chronic myelomonocytic leukemia, systemic lupus erythematosus, orbital cellulitis, and suprasellar meningioma. Initial clinical presentations ranged from asymptomatic (n = 1) to complaints of eye redness (n = 3), nausea/vomiting (n = 2), eye pain (n = 1), decreased vision (n = 1), and diplopia (n = 1). Ophthalmic findings consisted of proptosis (n = 4), conjunctival injection (n = 3), chemosis (n = 3), ophthalmoplegia (n = 2), periorbital swelling (n = 2), decreased Va (n = 1), and relative afferent pupillary defect (n = 1). An SOV thrombosis was detected on CT in all cases with confirmation and further evaluation by MRI. Treatment was tailored to the underlying disease process and included chemotherapy, anticoagulation therapy, intravenous antibiotics, systemic steroids, functional endoscopic sinus surgery, and observation. In the 2 subjects who initially presented with visual impairment, neither experienced visual improvement with both subjects having no light perception (NLP) vision at last follow up.

Cavernous Sinus Thrombosis.

There were 5 cases (4%) diagnosed with a cavernous sinus thrombosis. Past medical histories included multiple abscesses, sinusitis with orbital cellulitis, sepsis, and neurosurgery after trauma. All patients were symptomatic at presentation. Initial clinical presentations ranged from complaints of diplopia (n = 2), eye redness (n = 1), eye pain (n = 1), and decreased vision (n = 1). Ophthalmic findings consisted of proptosis (n = 5), ophthalmoplegia (n = 4), periorbital swelling (n = 3), chemosis (n = 2), conjunctival injection (n = 1), ptosis (n = 1), and decreased Va (n = 1). A dilated SOV was diagnosed in all cases by CT and with additional MRI and magnetic resonance venography confirming the dilated SOV and cavernous sinus thrombosis. All cases received intravenous anticoagulation therapy and a majority additionally received intravenous antibiotic therapy. In the 2 subjects who initially presented with visual impairment, neither experienced visual improvement at last follow up. One subject, a 45-year-old man who recently underwent cervical spine surgery for jumped facets after a traumatic motor vehicle collision presented with unilateral no light perception vision. After stabilization, he was ultimately treated with full-dose intravenous heparin for the cavernous sinus thrombosis; however, his no light perception vision remained unchanged at last follow up.


Idiopathic Orbital Inflammation.

There were 5 cases (4%) diagnosed with a dilated SOV and IOI. All patients were symptomatic at presentation. Initial clinical presentations ranged from complaints of eye pain (n = 5), eye redness (n = 3), and diplopia (n = 2). Ophthalmic findings consisted of ophthalmoplegia (n = 5), periorbital swelling (n = 5), conjunctival injection (n = 3), proptosis (n = 2), and ptosis (n = 1). A dilated SOV in each case was diagnosed on MRI with additional radiographic imaging consisting of CT and magnetic resonance venography confirming the dilated SOV and IOI. Unique cases included one subject with marked superior rectus myositis and adjacent dilated SOV, and a case of Tolosa–Hunt syndrome with inflammatory changes noted in the cavernous sinus on MRI with an ipsilateral dilated SOV. All cases were treated with systemic steroid therapy with some additionally requiring intravenous anticoagulation.


The remaining cases consisted of patients presenting with rare underlying etiologies responsible for causing a dilated SOV. Each of these occurred in less than 3.5% of total cases. Causative etiologies included various cerebrovascular malformations of which a vein of galen malformation was most common, orbital hemorrhage, intracranial hemorrhage, orbital lymphoma, thyroid eye disease, and orbital cellulitis.


To the authors’ knowledge, this is the largest case series describing dilated SOVs, including etiologies, clinical implications, treatment, and visual outcomes. Although previous reports have described dilated SOVs, these investigations consisted of either smaller cohorts, limited etiologies, or did not report ophthalmic findings, visual prognosis, treatment, and long-term follow-up status. In agreement with previous reports, the etiology of a dilated SOV in the authors’ cohort fell within the major categories of vascular malformation, venous thrombosis, and inflammatory. A careful examination of the numerous underlying etiologies of a dilated SOV can be explained and classified by 3 basic overall mechanisms. The principle mechanisms of SOV dilation include 1) alterations in orbital vascular hydrodynamics including high-pressure retrograde flow caused by an arteriovenous fistula; 2) impaired venous outflow with congestion from intraluminal obstruction; or 3) external mechanical compression of the SOV and/or orbital vasculature by either a mass or secondary to local inflammatory changes.

The most common specific diagnoses in decreasing frequency included cerebrovascular malformations, orbital AVM, SOV thrombosis, facial AVM, cavernous sinus thrombosis, and IOI. Clinical implications of the various underlying disease processes causing a dilated SOV ranged from benign to sight- and life-threatening. Visual impairment was observed at presentation in 26% of subjects among all patients. Visual impairment on presentation was more likely in patients with orbital cellulitis, orbital hemorrhage, orbital AVM, cavernous sinus thrombosis, and SOV thrombosis. These etiologies were also found to have the greatest visual deficit at last follow up. Etiologies that demonstrated the least visual impairment at presentation included IOI, thyroid eye disease, intracranial hemorrhage, orbital lymphoma, and various cerebrovascular malformations.

Among diagnostic categories, cerebrovascular malformations were most common. Within this subgroup, DCF represented most cases followed by CCF. Among these 2 subgroups, cases with a DCF presented with greater visual impairment when compared with cases of CCF. Consistent with previous reports, a DCF in this study was diagnosed predominantly in postmenopausal women.19 However, those patients found to have a CCF were largely composed of women, with a mean age of 54 years, and most commonly appeared to arise idiopathically with less than half reporting a history of trauma. This finding is in contrast to most studies reporting CCF occurring more frequently in young men and typically diagnosed shortly after head trauma with associated skull base fractures.

While a minority of patients with an intracranial arteriovenous fistula were asymptomatic, most of them presented with ophthalmic signs including cranial neuropathies, changes in pupil reactivity, proptosis, chemosis, optic nerve edema, and neurologic manifestations such as seizures or chronic headaches, which were found to be vital diagnostic clues. The wide-ranging ophthalmic findings and clinical presentations documented in this series are believed to be largely dependent on the unique underlying disease process, location, and severity.20 Recognizing these presenting signs and symptoms stresses the importance of taking a comprehensive ophthalmic and neurologic history and performing a detailed physical examination. Furthermore, a thorough understanding of the orbital and cerebral vascular anatomy, wide differential diagnosis of a dilated SOV, and clinical implications of this rare radiologic finding are crucial for timely diagnosis and management. The appropriate diagnosis and treatment often relied heavily on clinical suspicion and subsequent radiography. Consistent with previous reports, the authors found a dilated SOV was readily demonstrated and well characterized on CT, MRI, and orbital ultrasound imaging.21 Treatment within the cohort ranged from observation and close follow up to advanced endovascular intervention and was tailored to both the underlying etiologies, complexity of the vascular abnormality, and patient preference.

This study is limited by its retrospective nature, lack of standardized data across etiologies and imaging modalities, wide ranging patient demographics and clinical presentations, potentially unrecognized confounding variables that may have contributed to vision status and outcomes, and a relatively small sample size among various etiologies due to the rarity of a dilated SOV. Referral bias must also be considered as many of the described cases were contributed by neurosurgeons and interventional neuroradiologists potentially leading to a predominance of cases with cerebral vascular malformations, which may not accurately represent the actual distribution of dilated SOVs in the general population. Referral by these subspecialties also provided a limited analysis of potentially useful ophthalmic data such as intraocular pressure recordings and fundus exams. Additionally, the authors acknowledge that some of the previously described etiologies, such as dural arteriovenous fistulas, may not be associated with or initially present with a significantly dilated SOV as described by Kawaguchi et al.22 This fact potentially limits the diagnostic value of detecting a dilated SOV early in some of these disease processes. Lastly, the authors’ ability to study outcomes was limited by lack of standardized Va testing across institutions, follow-up time, and recording of ophthalmic data among all cases within a particular subgroup.

Though this study does have limitations, it nevertheless represents the largest report of dilated SOVs to date. When evaluating patients with a dilated SOV, physicians must keep the broad differential diagnosis in mind and use patient characteristics, clinical presentations, and radiographic features to distinguish between competing possibilities and allow for timely and proper treatment. The aggressive nature and potential visual and neurologic implications many of these etiologies possess stress the need for early diagnosis and treatment, which may lead to more favorable long-term outcomes.


The author Carlos Bianciotto, M.D., is deceased.


1. Komatsu H, Matsumoto F, Kasai M, et al.Cavernous sinus thrombosis caused by contralateral sphenoid sinusitis: a case report. Head Face Med 2013;9:9.
2. Shinder R, Oellers P, Esmaeli B, et al.Superior ophthalmic vein thrombosis in a patient with chronic myeloid leukemia receiving antifibrinolytic and thrombopoietin receptor agonist therapy. J Ocul Pharmacol Ther 2010;26:293–6.
3. Monteiro ML, Angotti-Neto H, Benabou JE, et al.Color Doppler imaging of the superior ophthalmic vein in different clinical forms of Graves’ orbitopathy. Jpn J Ophthalmol 2008;52:483–8.
4. Carrim ZI, Ahmed TY, Wykes WNIsolated superior ophthalmic vein thrombosis with orbital congestion: a variant of idiopathic orbital inflammatory disease? Eye (Lond) 2007;21:665–6.
5. Lirng JF, Fuh JL, Wu ZA, et al.Diameter of the superior ophthalmic vein in relation to intracranial pressure. AJNR Am J Neuroradiol 2003;24:700–3.
6. Reddy A, Foroozan R, Edmond JC, et al.Dilated superior ophthalmic veins and posterior ischemic optic neuropathy after prolonged spine surgery. J Neuroophthalmol 2008;28:327–8.
7. Peyster RG, Savino PJ, Hoover ED, Schatz NJDifferential diagnosis of the enlarged superior ophthalmic vein. J Comput Assist Tomogr 1984;8:103–7.
8. Dutton J, Waldrop TDutton JThe venous system of the orbit. In: Atlas of Surgical and Clinical Orbital Anatomy. 1994:Philadelphia: WB Saunders, 81–92.
9. Cheung N, McNab AAVenous anatomy of the orbit. Invest Ophthalmol Vis Sci 2003;44:988–95.
10. Tsutsumi S, Nakamura M, Tabuchi T, et al.The superior ophthalmic vein: delineation with high-resolution magnetic resonance imaging. Surg Radiol Anat 2015;37:75–80.
11. Bacon KT, Duchesneau PM, Weinstein MADemonstration of the superior ophthalmic vein by high resolution computed tomography. Radiology 1977;124:129–31.
12. Ozgen A, Aydingöz UNormative measurements of orbital structures using MRI. J Comput Assist Tomogr 2000;24:493–6.
13. Brightbill TC, Martin SB, Bracer RThe diagnostic significance of large superior ophthalmic veins in patients with normal and increased intracranial pressure: CT and MR evaluation. Neuro-Ophthalmology. 2001;26:93–101.
14. Ahmadi J, Teal JS, Segall HD, et al.Computed tomography of carotid-cavernous fistula. AJNR Am J Neuroradiol 1983;4:131–6.
15. Brismar JOrbital phlebography. II. Anatomy of superior ophthalmic vein and its tributaries. Acta Radiol Diagn (Stockh) 1974;15:481–96.
16. Reis CV, Gonzalez FL, Zabramski JM, et al.Anatomy of the superior ophthalmic vein approach for direct endovascular access to vascular lesions of the orbit and cavernous sinus. Neurosurgery 2009;64(5 suppl 2):318–23.
17. Doyon D, Aron-Rosa D, Ram‘ee ANewton T, Potts DOrbital veins and cavernous sinus. In: Radiology of Skull and Brain: Angiography. 1974:Saint Louis: The C.V. Mosby Company, 2220–54.
18. Hanafee WN, Shiu PC, Dayton GOOrbital venography. Am J Roentgenol Radium Ther Nucl Med 1968;104:29–35.
19. Kurata A, Miyasaka Y, Oka H, et al.Spontaneous carotid cavernous fistulas with special reference to the influence of estradiol decrease. Neurol Res 1999;21:631–9.
20. Cognard C, Gobin YP, Pierot L, et al.Cerebral dural arteriovenous fistulas: clinical and angiographic correlation with a revised classification of venous drainage. Radiology 1995;194:671–80.
21. Wei R, Cai J, Ma X, et al.Imaging diagnosis of enlarged superior ophthalmic vein. Zhonghua Yan Ke Za Zhi 2002;38:402–4.
22. Kawaguchi S, Sakaki T, Uranishi RColor Doppler flow imaging of the superior ophthalmic vein in dural arteriovenous fistulas. Stroke 2002;33:2009–13.
© 2017 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.