Original Investigation
Carotid Cavernous Sinus Fistulas Without Superior Ophthalmic Vein Enlargement Sarah M. Jacobs, M.D.*, Eric J. Arias, M.D.†, Colin P. Derdeyn, M.D.†‡, Steven M. Couch, M.D.*, and Philip L. Custer, M.D.* *Department of Ophthalmology and Visual Sciences, †Department of Neurosurgery, and ‡Departments of Neurology and Radiology, Washington University, St. Louis, Missouri, U.S.A.
Purpose: Diagnosis of carotid cavernous fistula (CCF) relies on clinical findings, such as proptosis, chemosis, and pulsatile tinnitus, plus imaging features including enlargement of the superior ophthalmic vein (SOV). This study reviewed patients with CCF, with a focus on those who were clinically symptomatic but had a normal-appearing SOV on routine scans. Methods: Retrospective review was conducted on the clinical records of patients with CCF seen by ophthalmology or interventional neuroradiology, with attention to clinical and imaging features, angiography findings, management, and outcomes. Results: Forty patients presented with CCF. History of head trauma was present in 13 (average age 43.8 years; all direct or complex), while the remainder occurred spontaneously (average 66 years; 85% indirect). The most common presenting ophthalmologic signs or symptoms were proptosis (65%), binocular diplopia (60%), redness (57.5%), and chemosis (47.5%). After diagnosis, 36 underwent endovascular treatment, with successful occlusion achieved in 90% of cases for whom follow-up data was available (n = 21). Notably, 3 patients with CCF did not have SOV enlargement on any imaging modality including catheter angiography. Conclusions: In this series of patients with clinical signs of CCF, there was no radiologic evidence of enlarged SOV in 26% of patients on noninvasive imaging and in 8% on catheter angiography. To avoid inappropriate interventions or delays in diagnosis and care, it is important to recognize that CCF can exist without SOV enlargement. Patients with clinical features suspicious for CCF should undergo catheter angiography if treatment is being considered. Endovascular treatment can produce clinical improvement or resolution. (Ophthal Plast Reconstr Surg 2015;31:191–196)
low-pressure venous space, resulting in elevated venous pressure. The resulting congestion within the orbital venous system gives rise to clinical signs and symptoms including diplopia and the classic triad of conjunctival chemosis, proptosis, and pulsatile tinnitus.1–5 In addition to clinical findings, practitioners often rely on noninvasive imaging methods such as CT or MRI to identify superior ophthalmic vein (SOV) enlargement which supports the diagnosis of CCF.2,6 Invasive catheter angiography is currently the most sensitive imaging modality for diagnosis and characterization of CCF.3,7 Timely diagnosis and endovascular treatment of CCF can produce clinical improvement or resolution.3,8,9 Unusual clinical or imaging characteristics of CCF can cause delays in diagnosis and management. This retrospective case series reviewed the clinical presentation, noninvasive imaging results, catheter angiography findings, management, and outcomes of 40 patients with CCF, with attention to features leading to initial misdiagnosis or delayed diagnosis.
MATERIALS AND METHODS The Washington University Institutional Review Board approved this study. A retrospective chart review was performed to identify patients with CCF who had presented to ophthalmology and/or interventional neuroradiology between January 2002 and April 2013. Age, gender, history (including trauma), time of onset and evolution of signs and symptoms, clinical course leading to diagnosis, management, and follow-up outcomes were collected. Imaging studies (CT, CT angiogram, MRI, MR angiogram, and catheter angiogram) were reviewed by an interventional neuroradiologist (author C.P.D.) to establish the presence or absence of SOV enlargement, to determine the CCF subtype (direct, indirect, or complex), and to characterize the venous drainage pathway of each lesion.
RESULTS
C
arotid cavernous fistula (CCF) is a connection between the cavernous sinus and the internal carotid artery (ICA), its branches, or branches of the external carotid artery (ECA) without an intervening capillary bed. This connection allows blood flow from the high-pressure arterial system to enter the
Accepted for publication June 2, 2014. Presented at 2013 ASOPRS Fall Scientific Symposium, New Orleans, LA, U.S.A. The authors have no financial or conflicts of interest to disclose. Address correspondence and reprint requests to Sarah M. Jacobs, M.D., Department of Ophthalmology and Visual Sciences, Washington University, 660 South Euclid Avenue, Campus Box 8096, St. Louis, MO 63110. E-mail: [email protected] DOI: 10.1097/IOP.0000000000000241
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
Forty patients were identified with CCF. The average age at diagnosis was 58.4 years (range 15–99 years), and 42.5% were male (n = 17). Direct fistulas between the ICA and the cavernous sinus were present in 37.5% (n = 15), while indirect fistulas involving branches of the ICA and/or ECA were present in 57.5% (n = 23), and complex or combined direct/indirect fistulas were found in the remaining 5% (n = 2). Subgroup analysis found that 32.5% of the CCF patients had a history of preceding head trauma (n = 13, 46% male, 11 direct and 2 complex, average age 43.8 years, age range 15–99 years). The remaining 67.5% appeared to have arisen spontaneously (n = 27, 37% male, 4 direct and 23 indirect, average age 66 years, age range 39–96 years). The average time from symptom onset to correct diagnosis was 2.9 months (data available for n = 37; range 1 day to 11 months), with the trauma subgroup diagnosed more quickly (average 1.5 months, range 1 day to 9 months) compared with the subgroup with spontaneous onset (average 3.7 months, range 3 weeks to 11 months).
191
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
S. M. Jacobs et al.
Presenting signs and symptoms included proptosis (65%, n = 26), binocular diplopia (60%, n = 24), red eye due to episcleral venous engorgement (57.5%, n = 23), conjunctival chemosis (47.5%, n = 19), retrobulbar or periocular pain (37.5%, n = 15), pulsatile tinnitus (17.5%, n = 7), eyelid ptosis (10%, n = 4), and seizures (5%, n = 2). Unilaterally elevated intraocular pressure was present in 53.3% (n = 16) of the 30 patients in whom measurements were taken. The complete triad of conjunctival chemosis, proptosis, and pulsatile tinnitus was observed in only 2 cases (Table 1). Most patients in the series were seen by an ophthalmologist at some point during their management (n = 32). Sixteen patients sought out an ophthalmologist as the first healthcare provider after the onset of symptoms. Of the 27 patients who were seen by an ophthalmologist prior to CCF diagnosis, 48% were correctly diagnosed at the first evaluation and 41% on a subsequent visit or after referral to a different ophthalmologist, and in 11% the diagnosis was made by another specialist after being missed by ophthalmology. Overall, 18 patients (45%) were initially misdiagnosed by their first provider (primary care, neurology, ophthalmology), with causes ranging from conjunctivitis to Bell’s palsy (Table 2). Early misdiagnosis
TABLE 1. Demographics and clinical and radiologic features in 40 patients with carotid cavernous fistula Number of cases (%)* Demographics Male Female Traumatic Spontaneous Average age: 58.5 (range 15–99) years Indirect CCF: 70.3 (range 47–96) years Direct CCF: 43.6 (range 15–99) years Complex CCF: 53.5 (range 30–77) years Average time from symptoms to diagnosis: 2.9 months (range 1 day–11 months) Traumatic: 1.5 months (range 1 day–9 months) Spontaneous 3.7 months (range 3 weeks–11 months) Clinical signs and symptoms Proptosis Binocular diplopia Red eye/episcleral venous engorgement Elevated intraocular pressure Conjunctival chemosis Retrobulbar or periocular pain Pulsatile tinnitus Eyelid ptosis New-onset seizures Imaging Noninvasive imaging done prior to catheter angiography Positive for SOVE Negative SOVE on noninvasive imaging, but positive SOVE on CA Negative SOVE on noninvasive imaging and on CA Catheter angiography Positive for CCF Positive for SOVE Negative for SOVE
17 (42.5) 23 (57.5) 13 (32.5) 27 (67.5) 22 (55) 16 (40) 2 (5)
26 (65) 24 (60) 23 (57.5) 16 of 30 (53.3) 19 (47.5) 15 (37.5) 7 (17.5) 4 (10) 2 (5) 30 (75) 22 of 30 (73.3) 6 of 30 (20) 2 of 30 (6.7) 39 of 39 (100) 36 of 39 (92) 3 of 39 (8)
*Percentage calculated as number of cases out of 40, unless otherwise noted. CA, catheter angiography; CCF, carotid cavernous fistula; CTA, CT angiogram; MRA, MR angiogram; PVA, polyvinyl alcohol; SOVE, superior ophthalmic vein enlargement.
192
was more common with indirect (13 of 23, 56%) than direct fistulas (5 of 15, 33%). Three patients underwent surgery based on an incorrect diagnosis: one for “traumatic strabismus,” the second for “aponeurotic ptosis,” and the third for “unilateral glaucoma.” Prior to proceeding with diagnostic catheter angiography, 30 patients had one (n = 22) or more (n = 8) noninvasive imaging studies. Noninvasive imaging modalities included CT (n = 12), CT angiogram (n = 7), MRI (n = 13), and MR angiogram (n = 6). Among patients with noninvasive imaging, SOV enlargement was present in 73.3% (n = 22). The remaining 26.7% (n = 8) had no appreciable SOV enlargement on noninvasive studies. Clinical findings were similar in patients with and without radiographic evidence of SOV enlargement (Fig. 1). Thirty-nine of 40 patients underwent catheter angiography, which confirmed the presence of CCF. One patient’s health and age precluded catheter angiogram, but her indirect CCF was apparent on CT angiogram, despite a normal-appearing SOV on all her imaging studies. Of the 8 patients who did not show SOV enlargement on noninvasive imaging, 6 did show SOV enlargement on subsequent catheter angiogram (false negatives: 3 by CTA, 2 by postcontrast MRI, and 1 by MRA), while 2 did not (true negatives: 2 by MRA). Although clinical signs and symptoms of orbital venous congestion were present, 3 of 39 patients (8%) had no SOV enlargement on catheter angiography (Table 3). Among this SOV-negative subset:
• Patient
1 was an 85-year-old female with 6 to 8 months of gradually progressive symptoms including diplopia in right- and
TABLE 2. Initial misdiagnosis in patients presenting with clinically symptomatic CCF Number of cases (%)* Overview Initially misdiagnosed Incidence of initial misdiagnosis in patients with indirect vs. direct fistula Indirect Direct Incidence of initial misdiagnosis in patients with spontaneous vs. traumatic etiology Spontaneous Traumatic Evaluation by ophthalmology Saw ophthalmologist during CCF diagnostic process Saw ophthalmologist as first provider after symptom onset Result of ophthalmology exam Diagnosed CCF at first evaluation Diagnosed CCF at subsequent evaluation CCF undiagnosed by ophthalmologist Initial misdiagnoses Infectious conjunctivitis Allergic conjunctivitis Nonspecific ocular irritation Bell’s palsy Scleritis Sinusitis Orbital cellulitis Headache syndrome Tolosa hunt Unilateral glaucoma Traumatic strabismus Aponeurotic ptosis
18 (45) 13 of 23 (56) 5 of 15 (33) 15 of 27 (55.6) 3 of 13 (23.1) 27 (67.5) 16 (40) 13 of 27 (48) 11 of 27 (41) 3 of 27 (11) 6 3 2 2 1 1 1 1 1 1† 1† 1†
*Percentage calculated as number of cases out of 40, unless otherwise noted. †Underwent surgery for the misdiagnosed entity prior to diagnosis of CCF.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
Carotid Cavernous Fistula
FIG. 1. Clinical appearance in CCF patients with or without radiographic evidence of SOV enlargement. A, 61-year-old male with marked chemosis, episcleral venous injection, and proptosis clinically, and asymmetric enlargement of the right SOV on postcontrast fat-saturated T1 coronal MR. B, 59-year-old female with mild chemosis, episcleral venous injection, and diplopia clinically, but no apparent SOV asymmetry on postcontrast fat-saturated T1 coronal MR. upgaze, episcleral injection, conjunctival chemosis, proptosis, pain, and decreasing visual acuity. Unilaterally elevated intraocular pressure (12 mm Hg higher on the affected side) led to glaucoma surgery. When symptoms persisted postoperatively, CCF was suspected. Catheter angiography showed an indirect left CCF draining through the circular sinus to the right cavernous and transverse sinuses via a network of cortical veins, without enlargement or retrograde filling of either SOV (Fig. 2). The inferior petrosal sinuses were thrombosed. The patient opted against treatment of her fistula. A follow-up angiogram 14 months later showed persistent CCF, still without SOV enlargement.
• Patient 2 was a 30-year-old male with a history of multiple skull fractures 6 years prior to symptom onset. After 6 weeks of gradually worsening bilateral eyelid edema and conjunctival chemosis, he experienced new-onset seizures that prompted diagnostic work-up. Catheter angiogram showed a complex direct CCF with brisk flow from the left internal carotid artery into the left cavernous sinus (Fig. 2). The fistula drained through the left superior petrosal vein into the transverse sinus, as well as to the left jugular bulb via the inferior petrosal sinus. The patient underwent transarterial embolization with coils and embolic glue, followed by additional coils and placement of 2 Amplatzer vascular closure
TABLE 3. Carotid cavernous fistula patients with symptomatic orbital congestion but no superior ophthalmic vein enlargement on catheter angiography
Patient
Sex
Fistula type
Trauma
1
85
F
Indirect
No
2
30
M
Complex
Yes
3
77
F
Complex
Yes
Signs, symptoms, and clinical course
Imaging performed (all negative for SOVE)
Fistula flow
6–8 months of binocular diplopia in rightand upgaze, with increasing episcleral injection, conjunctival chemosis, proptosis, retrobulbar pain, headache, and decreasing visual acuity. >10 mm Hg of IOP asymmetry prompted a unilateral glaucoma surgery. History of significant blunt head trauma 6 years prior to symptom onset, with multiple skull fractures. 6 weeks of mild bilateral eyelid edema with conjunctival chemosis. New-onset seizures prompted diagnostic work-up.
1. CA
Indirect flow into circular sinus which then drains through cortical veins into the right cavernous and transverse sinuses
1. MRI and MRA 2. CA
Fall down stairs 4 months prior to symptom onset, with skull base fracture through petrous apex and carotid canal. 3 weeks of unilateral right eyelid edema, conjunctival chemosis, and retrobulbar pain, with unilaterally elevated IOP. Gradually worsening binocular diplopia with bilateral extraocular motility deficits (bilateral CN6 palsy, left CN3 palsy). New-onset seizures prompted repeat angiography.
The day of her fall: 1. CT 2. MRA 3. CA After seizure onset: 4. CT 5. CA
Aneurysmally dilated ICA directly flows into enlarged cavernous sinus, which drains through the superior petrosal vein to the transverse sinus, and through inferior petrosal sinus to the jugular bulb Initial CA showed small CCF with drainage to pterygoid plexus. Subsequent CA showed direct flow from the right ICA with complex drainage to MCV, SSS, ipsilateral spinal and mesencephalic veins, as well as indirect routes from the ethmoidal arteries into SSS
CA, Catheter angiography; CCF, carotid cavernous fistula; CN3, cranial nerve III; CN6, cranial nerve VI; CTA, CT angiogram; ICA, internal carotid artery; IOP, intraocular pressure; MCV, middle cerebral veins; MRA, MR angiogram; SOVE, superior ophthalmic vein enlargement; SSS, superior sagittal sinus.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
193
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
S. M. Jacobs et al.
FIG. 2. Catheter angiography in CCF patients without SOV enlargement. A, 85-year-old female with indirect right CCF draining from the circular sinus through the cavernous and transverse sinuses, without SOV filling or enlargement. B, 30-year-old male demonstrating absence of SOV filling or enlargement in the setting of a direct left CCF with complex drainage through the petrosal vein, transverse sinus, and internal jugular veins. (C) 77-year-old female with a slow-flowing direct right CCF draining to the superior sagittal sinus and bilateral pterygoid plexuses.
•
devices (St. Jude Medical). A repeat angiogram 4 months after treatment showed complete occlusion of the CCF. The patient experienced complete resolution of his signs and symptoms. Patient 3 was a 77-year-old female who experienced skull base fractures through the petrous apex and carotid canal. Angiogram at the time of her injury showed a small direct CCF without SOV enlargement, draining to the right pterygoid plexus. No intervention was performed at that time. Four months later, she developed unilateral eyelid edema, conjunctival chemosis, pain, left third nerve palsy and bilateral sixth nerve palsies, unilateral intraocular pressure elevation, and new-onset seizures. Repeat angiogram demonstrated a complex direct CCF arising from the right internal carotid artery just proximal to the ophthalmic artery, with a relatively slow arteriovenous shunt draining bilaterally through the pterygoid plexuses and superior sagittal sinus (Fig. 2). Indirect fistula routes were also present from the ethmoidal arteries into the superior sagittal sinus. The patient underwent transvenous coil embolization. Despite complete occlusion of the CCF, she still demonstrated residual cranial nerve palsies and symptomatic diplopia at 12-month follow up.
In the entire series, after CCF diagnosis, 36 of the 40 patients underwent endovascular treatment once (n = 31) or twice (n = 5), with coils (n = 15), embolization with Onyx copolymer (Micro Therapeutics Inc., Irvine, CA) (n = 3), or a combination thereof (n = 17). One patient required subsequent placement of Amplatzer vascular plugs (AGA Medical Corp., Golden Valley, MN). Among the 20 patients for whom follow-up data are available, successful complete occlusion was achieved in 18, while partial occlusion was achieved in 2. Follow-up data were unavailable for 16 treated patients due to death unrelated to CCF or its treatment (n = 2) or because subsequent care was provided elsewhere (n = 14). Seven of the patients who demonstrated complete post-treatment occlusion of the CCF still reported ophthalmologic symptoms on follow up, including diplopia related to persistent cranial nerve palsy (n = 6) and episodic chemosis plus retrobulbar pain (n = 1). Two patients experienced complications of endovascular treatment: one with new cranial nerve VI palsy, and the other with transient diplopia for 3 weeks postprocedure. Of the 4 patients managed by observation only, 1 spontaneously resolved, 2 had a persistent symptomatic CCF, and 1 was lost to follow up.
DISCUSSION CCF classically presents with various combinations of diplopia, proptosis, conjunctival chemosis, episcleral venous engorgement, and pulsatile tinnitus and with enlargement of the SOV on CT or MRI. However, CCF is often misdiagnosed, especially when typical clinical features or imaging findings are not present. One series reported that patients presenting
194
without orbital bruit were more likely to pose a diagnostic challenge.10 Barry et al.8 reported that initial misdiagnosis is more common for patients with indirect fistula (58% misdiagnosed) than direct (25% misdiagnosed), similar to the findings in this study. Patients with other traumatic deficits that confound the diagnostic picture, or with unusual demographics for CCF (such as pediatric age), are also more likely to be misdiagnosed initially.11,12 Patients with CCF are commonly evaluated by ophthalmology prior to diagnosis, and 40% of those in this series initially presented to their ophthalmologist. Nearly half of these patients were initially misdiagnosed. Delayed diagnosis and treatment may result in higher likelihood of persistent ophthalmologic deficits.8 In a 76-patient series, the average time from symptom onset to diagnosis was 16.1 weeks, with longer diagnostic delays occurring in patients who initially presented with cranial nerve palsy before the onset of other orbital congestive symptoms.13 Fistulas with a higher proportion of drainage via cortical veins were more likely to have an atypical presentation pattern resulting in diagnostic delays.13 This is consistent with other case reports in which CCF presented atypically with venous infarction, intracranial hemorrhage, dural venous sinus thrombosis, posterior ischemic optic neuropathy, isolated cranial nerve palsies, or elevated intracranial pressure mimicking pseudotumor cerebri, leading to initial misdiagnosis.14–20 It is standard teaching that a CCF with orbital clinical findings will demonstrate SOV enlargement on imaging.1,2,21–23 However, in this series, 26.7% (8 of 30) did not have SOV enlargement when studied with various noninvasive modalities (CT, CTA, MRI, MRA). There are several inherent sources of error in the interpretation of noninvasive images, including subtle venous asymmetry that can be undetectable, or bilateral venous drainage from the CCF leading to mild symmetric SOV enlargement that is difficult to appreciate.24 Three of the patients (8%) had no SOV enlargement on any imaging study, including catheter angiography. The question arises how the orbital venous system can be pressurized enough to cause clinical symptoms without distending the SOV. In all 3 cases, an alternate venous drainage pathway was present, apparently decompressing the SOV sufficiently to render it normal in caliber radiologically. Furthermore, vessel caliber is a function of not only pressure but also flow volume. As all 3 patients had drainage back through intracranial venous channels rather than forward through the orbit to the angular and facial veins, fistula flow through the orbit was limited. This lower flow state may have allowed the SOV to remain normal in caliber, despite relative pressurization of the orbital venous system. Another possible mechanism may relate to the presence of competent valves, decreasing the volume of retrograde flow from
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
the fistula into the SOV in these 3 patients. Although classically believed to be devoid of valves, recent anatomical studies have shown that valves are present in up to 75% of ophthalmic veins.25 Finally, SOV stenosis or occlusion could potentially prevent enlargement of the SOV despite elevated orbital venous pressure, but this was not seen in this series. There have been other reports of patients with ophthalmologic symptoms but no SOV enlargement on noninvasive imaging, including a case report of a 75-year-old woman with diplopia, episcleral venous injection, ophthalmoplegia, and ptosis, but a normal CT and MRI.26 Kurata et al.13 found that 3 of 51 patients with symptomatic orbital congestion had no appreciable drainage through the SOV on angiography. Conversely, a patient with radiologic evidence of CCF with SOV enlargement but without any ophthalmologic symptoms has also been reported.27 Several publications have suggested alternate means for assessing elevated venous pressure within the orbit to help diagnose CCF. One study found that the ocular pulse amplitude (difference between minimum and maximum intraocular pressure during the cardiac cycle, as measured by pneumotonometry) was significantly larger in CCF patients and was 100% sensitive and 93% specific in distinguishing CCF patients from normals or from those with other orbital diseases.28 Orbital color Doppler ultrasound of the retrobulbar vessels can detect reversal of flow in the SOV, helping distinguish CCF from other causes of SOV enlargement such as thyroid-associated orbitopathy and allowing noninvasive monitoring of the progression or resolution of the fistula over time.6,29,30 Blood flow velocity and pulsatility index parameters specific to CCF can reportedly be detected with transcranial Doppler ultrasound, even in lesions with normal CT or MRI findings.31 Appropriate diagnosis and management can result in partial or full resolution of signs and symptoms. In this series, 90% of patients for whom follow-up data were available showed full anatomical resolution of CCF after endovascular intervention comparable with the rates reported elsewhere ranging from 84% to 100%.8,9,32,33 The fistula can be accessed via orbital vessels or through a more distal vascular access point.34,35 In one review, embolization produced a higher rate of resolution (91%) than gamma knife radiation (72%), and the transvenous approach for embolization had a superior success rate (98%) compared with transarterial (88%).33 Resolution after stereotactic radiosurgery and spontaneous involution with observation alone have also been reported.36,37 Of interest, one patient in this series continued to experience congestive orbital symptoms after endovascular treatment, despite complete anatomical occlusion of the CCF demonstrated by angiography. Similarly, in a series of 60 patients with CCF who underwent endovascular occlusion, anatomical cure was seen more frequently (63%) than clinical cure (51%).3 This study is limited by its retrospective nature, including incomplete prereferral records and incomplete follow-up data. As a tertiary medical center, a high percentage of patients return to the referring provider after diagnosis or intervention. Effort was made to obtain reports and images from outside providers and to standardize the interpretation of the imaging studies through review of each image series by the authors. In conclusion, CCF can give rise to neurologic and ophthalmologic signs and symptoms. It is frequently misdiagnosed, especially when presenting with atypical clinical features or subtle imaging characteristics. SOV enlargement and other findings of fistulas may not be detectable with noninvasive imaging. Even with catheter angiography, 8% of patients with CCF and clinical signs of orbital congestion lacked SOV enlargement in this series. To avoid inappropriate interventions or delays in
Carotid Cavernous Fistula
diagnosis and care, patients with clinical features suspicious for CCF should undergo catheter angiography to identify fistula flow. Endovascular treatment can produce clinical improvement or resolution.
REFERENCES 1. Keltner JL, Satterfield D, Dublin AB, et al. Dural and carotid cavernous sinus fistulas. Diagnosis, management, and complications. Ophthalmology 1987;94:1585–600. 2. Miller NR. Dural carotid-cavernous fistulas: epidemiology, clinical presentation, and management. Neurosurg Clin N Am 2012;23:179–92. 3. Preechawat P, Narmkerd P, Jiarakongmun P, et al. Dural carotid cavernous sinus fistula: ocular characteristics, endovascular management and clinical outcome. J Med Assoc Thai 2008;91:852–8. 4. Théaudin M, Saint-Maurice JP, Chapot R, et al. Diagnosis and treatment of dural carotid-cavernous fistulas: a consecutive series of 27 patients. J Neurol Neurosurg Psychiatry 2007;78:174–9. 5. Yoshihara M, Saito N, Kashima Y, Ishikawa H. The Ishikawa classification of cavernous sinus lesions by clinic-anatomical findings. Jpn J Ophthalmol 2001;45:420–4. 6. Wei R, Cai J, Ma X, et al. Imaging diagnosis of enlarged superior ophthalmic vein. Zhonghua Yan Ke Za Zhi 2002;38:402–4. 7. Hu Y, Wang Z, Quan W, et al. Digital subtraction angiography of cerebrum for diagnosis of cavernous sinus fistula and its intravascular treatment. Zhonghua Yan Ke Za Zhi 1999;35:197–9. 8. Barry RC, Wilkinson M, Ahmed RM, et al. Interventional treatment of carotid cavernous fistula. J Clin Neurosci 2011;18:1072–9. 9. Gemmete JJ, Chaudhary N, Pandy A, Ansari S. Treatment of carotid cavernous fistula. Curr Treat Options Neurol 2010;12:43–53. 10. Taniguchi RM, Goree JA, Odom GL. Spontaneous carotid cavernous shunts presenting diagnostic problems. J Neurosurg 1971;35:384–91. 11. Gossman MD, Berlin AJ, Weinstein MA, et al. Spontaneous direct carotid-cavernous fistula in childhood. Ophthal Plast Reconstr Surg 1993;9:62–5. 12. Lau FH, Yuen HK, Rao SK, et al. Spontaneous carotid cavernous fistula in a pediatric patient: case report and review of literature. J AAPOS 2005;9:292–4. 13. Kurata A, Suzuki S, Iwamoto K, et al. Dural arteriovenous fistulas in the cavernous sinus: clinical research and treatment. ISRN Neurol 2011;2011:453834. 14. Miyamoto N, Naito I, Takatama S, et al. Clinical and angiographic characteristics of cavernous sinus dural arteriovenous fistulas manifesting as venous infarction and/or intracranial hemorrhage. Neuroradiology 2009;51:53–60. 15. Hashimoto M, Ohtsuka K, Suzuki Y, et al. A case of posterior ischemic optic neuropathy in a posterior-draining dural cavernous sinus fistula. J Neuroophthalmol 2005;25:176–9. 16. Kuwayama N. Classification and diagnosis of intracranial dural arteriovenous fistulas. Brain Nerve 2008;60:887–95. 17. Miyagishima T, Hara T, Inoue M, et al. Pontine venous congestion due to dural arteriovenous fistula of the cavernous sinus: Case report and review of the literature. Surg Neurol Int 2012;3:53. 18. Tsai RK, Chen HY, Wang HZ. Painful fourth cranial nerve palsy caused by posteriorly-draining dural carotid-cavernous sinus fistula. J Formos Med Assoc 2000;99:730–2. 19. Masaya-Anon P. Isolated oculomotor nerve palsy in a white-eyed patient with dural carotid-cavernous sinus fistulas: a case report. J Med Assoc Thai 2012;95(suppl 4):S143–6. 20. Oestreicher JH, Frueh BR. Carotid-cavernous fistula mimicking Graves’ eye disease. Ophthal Plast Reconstr Surg 1995;11:238–44. 21. Slusher MM, Lennington BR, Weaver RG, et al. Ophthalmic findings in dural arteriovenous shunts. Ophthalmology 1979;86:720–31. 22. Feiner L, Bennett J, Volpe NJ. Cavernous sinus fistulas: carotid cavernous fistulas and dural arteriovenous malformations. Curr Neurol Neurosci Rep 2003;3:415–20. 23. Miller NR. Diagnosis and management of dural carotid-cavernous fistulas. Neurosurg Focus 2007;23:E13. 24. Ros de San Pedro J, Pérez CJ, Parra JZ, et al. Bilateral ethmoidal dural arteriovenous fistula: unexpected surgical diagnosis. Clin Neurol Neurosurg 2010;112:903–8.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
195
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
S. M. Jacobs et al.
25. Zhang J, Stringer MD. Ophthalmic and facial veins are not valveless. Clin Experiment Ophthalmol 2010;38:502–10. 26. Théaudin M, Chapot R, Vahedi K, et al. Dural carotid-cavernous fistula: relationship between evolution of clinical symptoms and venous drainage changes. Cerebrovasc Dis 2008;25:382–4. 27. Ikeda K, Deguchi K, Tsukaguchi M, et al. Absence of orbito-ocular signs in dural carotid-cavernous sinus fistula with a prominent anterior venous drainage. J Neurol Sci 2005;236:81–4. 28. Golnik KC, Miller NR. Diagnosis of cavernous sinus arteriovenous fistula by measurement of ocular pulse amplitude. Ophthalmology 1992;99:1146–52. 29. Hatton MP, Remulla HD, Tolentino MJ, et al. Clinical applications of color Doppler imaging in the management of orbital lesions. Ophthal Plast Reconstr Surg 2002;18:462–5. 30. Jianu DC, Jianu SN. Carotid-cavernous fistulas of low-flow type: color Doppler imaging of retrobulbar vessels findings. Oftalmologia 2010;54:83–90. 31. Kiliç T, Elmaci I, Bayri Y, et al. Value of transcranial Doppler ultrasonography in the diagnosis and follow-up of carotid-cavernous fistulae. Acta Neurochir (Wien) 2001;143:1257–64.
196
32. Jung KH, Kwon BJ, Chu K, et al. Clinical and angiographic factors related to the prognosis of cavernous sinus dural arteriovenous fistula. Neuroradiology 2011;53:983–92. 33. Choi BS, Park JW, Kim JL, et al. Treatment strategy based on multimodal management outcome of cavernous sinus dural arteriovenous fistula (CSDAVF). Neurointervention 2011;6: 6–12. 34. Badilla J, Haw C, Rootman J. Superior ophthalmic vein cannulation through a lateral orbitotomy for embolization of a cavernous dural fistula. Arch Ophthalmol 2007;125:1700–2. 35. Michels KS, Ng JD, Falardeau J, et al. Transvenous embolization of a dural carotid-cavernous sinus fistula via the inferior ophthalmic vein. Ophthal Plast Reconstr Surg 2007;23:480–2. 36. Chong GT, Mukundan S, Kirkpatrick JP, et al. Stereotactic radiosurgery in the treatment of a dural carotid-cavernous fistula. J Neuroophthalmol 2010;30:138–44. 37. Bujak M, Margolin E, Thompson A, et al. Spontaneous resolution of two dural carotid-cavernous fistulas presenting with optic neuropathy and marked congestive ophthalmopathy. J Neuroophthalmol 2010;30:222–7.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Carotid Cavernous Sinus Fistulas Without Superior Ophthalmic Vein Enlargement Sarah M. Jacobs, M.D.*, Eric J. Arias, M.D.†, Colin P. Derdeyn, M.D.†‡, Steven M. Couch, M.D.*, and Philip L. Custer, M.D.* *Department of Ophthalmology and Visual Sciences, †Department of Neurosurgery, and ‡Departments of Neurology and Radiology, Washington University, St. Louis, Missouri, U.S.A.
Purpose: Diagnosis of carotid cavernous fistula (CCF) relies on clinical findings, such as proptosis, chemosis, and pulsatile tinnitus, plus imaging features including enlargement of the superior ophthalmic vein (SOV). This study reviewed patients with CCF, with a focus on those who were clinically symptomatic but had a normal-appearing SOV on routine scans. Methods: Retrospective review was conducted on the clinical records of patients with CCF seen by ophthalmology or interventional neuroradiology, with attention to clinical and imaging features, angiography findings, management, and outcomes. Results: Forty patients presented with CCF. History of head trauma was present in 13 (average age 43.8 years; all direct or complex), while the remainder occurred spontaneously (average 66 years; 85% indirect). The most common presenting ophthalmologic signs or symptoms were proptosis (65%), binocular diplopia (60%), redness (57.5%), and chemosis (47.5%). After diagnosis, 36 underwent endovascular treatment, with successful occlusion achieved in 90% of cases for whom follow-up data was available (n = 21). Notably, 3 patients with CCF did not have SOV enlargement on any imaging modality including catheter angiography. Conclusions: In this series of patients with clinical signs of CCF, there was no radiologic evidence of enlarged SOV in 26% of patients on noninvasive imaging and in 8% on catheter angiography. To avoid inappropriate interventions or delays in diagnosis and care, it is important to recognize that CCF can exist without SOV enlargement. Patients with clinical features suspicious for CCF should undergo catheter angiography if treatment is being considered. Endovascular treatment can produce clinical improvement or resolution. (Ophthal Plast Reconstr Surg 2015;31:191–196)
low-pressure venous space, resulting in elevated venous pressure. The resulting congestion within the orbital venous system gives rise to clinical signs and symptoms including diplopia and the classic triad of conjunctival chemosis, proptosis, and pulsatile tinnitus.1–5 In addition to clinical findings, practitioners often rely on noninvasive imaging methods such as CT or MRI to identify superior ophthalmic vein (SOV) enlargement which supports the diagnosis of CCF.2,6 Invasive catheter angiography is currently the most sensitive imaging modality for diagnosis and characterization of CCF.3,7 Timely diagnosis and endovascular treatment of CCF can produce clinical improvement or resolution.3,8,9 Unusual clinical or imaging characteristics of CCF can cause delays in diagnosis and management. This retrospective case series reviewed the clinical presentation, noninvasive imaging results, catheter angiography findings, management, and outcomes of 40 patients with CCF, with attention to features leading to initial misdiagnosis or delayed diagnosis.
MATERIALS AND METHODS The Washington University Institutional Review Board approved this study. A retrospective chart review was performed to identify patients with CCF who had presented to ophthalmology and/or interventional neuroradiology between January 2002 and April 2013. Age, gender, history (including trauma), time of onset and evolution of signs and symptoms, clinical course leading to diagnosis, management, and follow-up outcomes were collected. Imaging studies (CT, CT angiogram, MRI, MR angiogram, and catheter angiogram) were reviewed by an interventional neuroradiologist (author C.P.D.) to establish the presence or absence of SOV enlargement, to determine the CCF subtype (direct, indirect, or complex), and to characterize the venous drainage pathway of each lesion.
RESULTS
C
arotid cavernous fistula (CCF) is a connection between the cavernous sinus and the internal carotid artery (ICA), its branches, or branches of the external carotid artery (ECA) without an intervening capillary bed. This connection allows blood flow from the high-pressure arterial system to enter the
Accepted for publication June 2, 2014. Presented at 2013 ASOPRS Fall Scientific Symposium, New Orleans, LA, U.S.A. The authors have no financial or conflicts of interest to disclose. Address correspondence and reprint requests to Sarah M. Jacobs, M.D., Department of Ophthalmology and Visual Sciences, Washington University, 660 South Euclid Avenue, Campus Box 8096, St. Louis, MO 63110. E-mail: [email protected] DOI: 10.1097/IOP.0000000000000241
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
Forty patients were identified with CCF. The average age at diagnosis was 58.4 years (range 15–99 years), and 42.5% were male (n = 17). Direct fistulas between the ICA and the cavernous sinus were present in 37.5% (n = 15), while indirect fistulas involving branches of the ICA and/or ECA were present in 57.5% (n = 23), and complex or combined direct/indirect fistulas were found in the remaining 5% (n = 2). Subgroup analysis found that 32.5% of the CCF patients had a history of preceding head trauma (n = 13, 46% male, 11 direct and 2 complex, average age 43.8 years, age range 15–99 years). The remaining 67.5% appeared to have arisen spontaneously (n = 27, 37% male, 4 direct and 23 indirect, average age 66 years, age range 39–96 years). The average time from symptom onset to correct diagnosis was 2.9 months (data available for n = 37; range 1 day to 11 months), with the trauma subgroup diagnosed more quickly (average 1.5 months, range 1 day to 9 months) compared with the subgroup with spontaneous onset (average 3.7 months, range 3 weeks to 11 months).
191
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
S. M. Jacobs et al.
Presenting signs and symptoms included proptosis (65%, n = 26), binocular diplopia (60%, n = 24), red eye due to episcleral venous engorgement (57.5%, n = 23), conjunctival chemosis (47.5%, n = 19), retrobulbar or periocular pain (37.5%, n = 15), pulsatile tinnitus (17.5%, n = 7), eyelid ptosis (10%, n = 4), and seizures (5%, n = 2). Unilaterally elevated intraocular pressure was present in 53.3% (n = 16) of the 30 patients in whom measurements were taken. The complete triad of conjunctival chemosis, proptosis, and pulsatile tinnitus was observed in only 2 cases (Table 1). Most patients in the series were seen by an ophthalmologist at some point during their management (n = 32). Sixteen patients sought out an ophthalmologist as the first healthcare provider after the onset of symptoms. Of the 27 patients who were seen by an ophthalmologist prior to CCF diagnosis, 48% were correctly diagnosed at the first evaluation and 41% on a subsequent visit or after referral to a different ophthalmologist, and in 11% the diagnosis was made by another specialist after being missed by ophthalmology. Overall, 18 patients (45%) were initially misdiagnosed by their first provider (primary care, neurology, ophthalmology), with causes ranging from conjunctivitis to Bell’s palsy (Table 2). Early misdiagnosis
TABLE 1. Demographics and clinical and radiologic features in 40 patients with carotid cavernous fistula Number of cases (%)* Demographics Male Female Traumatic Spontaneous Average age: 58.5 (range 15–99) years Indirect CCF: 70.3 (range 47–96) years Direct CCF: 43.6 (range 15–99) years Complex CCF: 53.5 (range 30–77) years Average time from symptoms to diagnosis: 2.9 months (range 1 day–11 months) Traumatic: 1.5 months (range 1 day–9 months) Spontaneous 3.7 months (range 3 weeks–11 months) Clinical signs and symptoms Proptosis Binocular diplopia Red eye/episcleral venous engorgement Elevated intraocular pressure Conjunctival chemosis Retrobulbar or periocular pain Pulsatile tinnitus Eyelid ptosis New-onset seizures Imaging Noninvasive imaging done prior to catheter angiography Positive for SOVE Negative SOVE on noninvasive imaging, but positive SOVE on CA Negative SOVE on noninvasive imaging and on CA Catheter angiography Positive for CCF Positive for SOVE Negative for SOVE
17 (42.5) 23 (57.5) 13 (32.5) 27 (67.5) 22 (55) 16 (40) 2 (5)
26 (65) 24 (60) 23 (57.5) 16 of 30 (53.3) 19 (47.5) 15 (37.5) 7 (17.5) 4 (10) 2 (5) 30 (75) 22 of 30 (73.3) 6 of 30 (20) 2 of 30 (6.7) 39 of 39 (100) 36 of 39 (92) 3 of 39 (8)
*Percentage calculated as number of cases out of 40, unless otherwise noted. CA, catheter angiography; CCF, carotid cavernous fistula; CTA, CT angiogram; MRA, MR angiogram; PVA, polyvinyl alcohol; SOVE, superior ophthalmic vein enlargement.
192
was more common with indirect (13 of 23, 56%) than direct fistulas (5 of 15, 33%). Three patients underwent surgery based on an incorrect diagnosis: one for “traumatic strabismus,” the second for “aponeurotic ptosis,” and the third for “unilateral glaucoma.” Prior to proceeding with diagnostic catheter angiography, 30 patients had one (n = 22) or more (n = 8) noninvasive imaging studies. Noninvasive imaging modalities included CT (n = 12), CT angiogram (n = 7), MRI (n = 13), and MR angiogram (n = 6). Among patients with noninvasive imaging, SOV enlargement was present in 73.3% (n = 22). The remaining 26.7% (n = 8) had no appreciable SOV enlargement on noninvasive studies. Clinical findings were similar in patients with and without radiographic evidence of SOV enlargement (Fig. 1). Thirty-nine of 40 patients underwent catheter angiography, which confirmed the presence of CCF. One patient’s health and age precluded catheter angiogram, but her indirect CCF was apparent on CT angiogram, despite a normal-appearing SOV on all her imaging studies. Of the 8 patients who did not show SOV enlargement on noninvasive imaging, 6 did show SOV enlargement on subsequent catheter angiogram (false negatives: 3 by CTA, 2 by postcontrast MRI, and 1 by MRA), while 2 did not (true negatives: 2 by MRA). Although clinical signs and symptoms of orbital venous congestion were present, 3 of 39 patients (8%) had no SOV enlargement on catheter angiography (Table 3). Among this SOV-negative subset:
• Patient
1 was an 85-year-old female with 6 to 8 months of gradually progressive symptoms including diplopia in right- and
TABLE 2. Initial misdiagnosis in patients presenting with clinically symptomatic CCF Number of cases (%)* Overview Initially misdiagnosed Incidence of initial misdiagnosis in patients with indirect vs. direct fistula Indirect Direct Incidence of initial misdiagnosis in patients with spontaneous vs. traumatic etiology Spontaneous Traumatic Evaluation by ophthalmology Saw ophthalmologist during CCF diagnostic process Saw ophthalmologist as first provider after symptom onset Result of ophthalmology exam Diagnosed CCF at first evaluation Diagnosed CCF at subsequent evaluation CCF undiagnosed by ophthalmologist Initial misdiagnoses Infectious conjunctivitis Allergic conjunctivitis Nonspecific ocular irritation Bell’s palsy Scleritis Sinusitis Orbital cellulitis Headache syndrome Tolosa hunt Unilateral glaucoma Traumatic strabismus Aponeurotic ptosis
18 (45) 13 of 23 (56) 5 of 15 (33) 15 of 27 (55.6) 3 of 13 (23.1) 27 (67.5) 16 (40) 13 of 27 (48) 11 of 27 (41) 3 of 27 (11) 6 3 2 2 1 1 1 1 1 1† 1† 1†
*Percentage calculated as number of cases out of 40, unless otherwise noted. †Underwent surgery for the misdiagnosed entity prior to diagnosis of CCF.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
Carotid Cavernous Fistula
FIG. 1. Clinical appearance in CCF patients with or without radiographic evidence of SOV enlargement. A, 61-year-old male with marked chemosis, episcleral venous injection, and proptosis clinically, and asymmetric enlargement of the right SOV on postcontrast fat-saturated T1 coronal MR. B, 59-year-old female with mild chemosis, episcleral venous injection, and diplopia clinically, but no apparent SOV asymmetry on postcontrast fat-saturated T1 coronal MR. upgaze, episcleral injection, conjunctival chemosis, proptosis, pain, and decreasing visual acuity. Unilaterally elevated intraocular pressure (12 mm Hg higher on the affected side) led to glaucoma surgery. When symptoms persisted postoperatively, CCF was suspected. Catheter angiography showed an indirect left CCF draining through the circular sinus to the right cavernous and transverse sinuses via a network of cortical veins, without enlargement or retrograde filling of either SOV (Fig. 2). The inferior petrosal sinuses were thrombosed. The patient opted against treatment of her fistula. A follow-up angiogram 14 months later showed persistent CCF, still without SOV enlargement.
• Patient 2 was a 30-year-old male with a history of multiple skull fractures 6 years prior to symptom onset. After 6 weeks of gradually worsening bilateral eyelid edema and conjunctival chemosis, he experienced new-onset seizures that prompted diagnostic work-up. Catheter angiogram showed a complex direct CCF with brisk flow from the left internal carotid artery into the left cavernous sinus (Fig. 2). The fistula drained through the left superior petrosal vein into the transverse sinus, as well as to the left jugular bulb via the inferior petrosal sinus. The patient underwent transarterial embolization with coils and embolic glue, followed by additional coils and placement of 2 Amplatzer vascular closure
TABLE 3. Carotid cavernous fistula patients with symptomatic orbital congestion but no superior ophthalmic vein enlargement on catheter angiography
Patient
Sex
Fistula type
Trauma
1
85
F
Indirect
No
2
30
M
Complex
Yes
3
77
F
Complex
Yes
Signs, symptoms, and clinical course
Imaging performed (all negative for SOVE)
Fistula flow
6–8 months of binocular diplopia in rightand upgaze, with increasing episcleral injection, conjunctival chemosis, proptosis, retrobulbar pain, headache, and decreasing visual acuity. >10 mm Hg of IOP asymmetry prompted a unilateral glaucoma surgery. History of significant blunt head trauma 6 years prior to symptom onset, with multiple skull fractures. 6 weeks of mild bilateral eyelid edema with conjunctival chemosis. New-onset seizures prompted diagnostic work-up.
1. CA
Indirect flow into circular sinus which then drains through cortical veins into the right cavernous and transverse sinuses
1. MRI and MRA 2. CA
Fall down stairs 4 months prior to symptom onset, with skull base fracture through petrous apex and carotid canal. 3 weeks of unilateral right eyelid edema, conjunctival chemosis, and retrobulbar pain, with unilaterally elevated IOP. Gradually worsening binocular diplopia with bilateral extraocular motility deficits (bilateral CN6 palsy, left CN3 palsy). New-onset seizures prompted repeat angiography.
The day of her fall: 1. CT 2. MRA 3. CA After seizure onset: 4. CT 5. CA
Aneurysmally dilated ICA directly flows into enlarged cavernous sinus, which drains through the superior petrosal vein to the transverse sinus, and through inferior petrosal sinus to the jugular bulb Initial CA showed small CCF with drainage to pterygoid plexus. Subsequent CA showed direct flow from the right ICA with complex drainage to MCV, SSS, ipsilateral spinal and mesencephalic veins, as well as indirect routes from the ethmoidal arteries into SSS
CA, Catheter angiography; CCF, carotid cavernous fistula; CN3, cranial nerve III; CN6, cranial nerve VI; CTA, CT angiogram; ICA, internal carotid artery; IOP, intraocular pressure; MCV, middle cerebral veins; MRA, MR angiogram; SOVE, superior ophthalmic vein enlargement; SSS, superior sagittal sinus.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
193
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
S. M. Jacobs et al.
FIG. 2. Catheter angiography in CCF patients without SOV enlargement. A, 85-year-old female with indirect right CCF draining from the circular sinus through the cavernous and transverse sinuses, without SOV filling or enlargement. B, 30-year-old male demonstrating absence of SOV filling or enlargement in the setting of a direct left CCF with complex drainage through the petrosal vein, transverse sinus, and internal jugular veins. (C) 77-year-old female with a slow-flowing direct right CCF draining to the superior sagittal sinus and bilateral pterygoid plexuses.
•
devices (St. Jude Medical). A repeat angiogram 4 months after treatment showed complete occlusion of the CCF. The patient experienced complete resolution of his signs and symptoms. Patient 3 was a 77-year-old female who experienced skull base fractures through the petrous apex and carotid canal. Angiogram at the time of her injury showed a small direct CCF without SOV enlargement, draining to the right pterygoid plexus. No intervention was performed at that time. Four months later, she developed unilateral eyelid edema, conjunctival chemosis, pain, left third nerve palsy and bilateral sixth nerve palsies, unilateral intraocular pressure elevation, and new-onset seizures. Repeat angiogram demonstrated a complex direct CCF arising from the right internal carotid artery just proximal to the ophthalmic artery, with a relatively slow arteriovenous shunt draining bilaterally through the pterygoid plexuses and superior sagittal sinus (Fig. 2). Indirect fistula routes were also present from the ethmoidal arteries into the superior sagittal sinus. The patient underwent transvenous coil embolization. Despite complete occlusion of the CCF, she still demonstrated residual cranial nerve palsies and symptomatic diplopia at 12-month follow up.
In the entire series, after CCF diagnosis, 36 of the 40 patients underwent endovascular treatment once (n = 31) or twice (n = 5), with coils (n = 15), embolization with Onyx copolymer (Micro Therapeutics Inc., Irvine, CA) (n = 3), or a combination thereof (n = 17). One patient required subsequent placement of Amplatzer vascular plugs (AGA Medical Corp., Golden Valley, MN). Among the 20 patients for whom follow-up data are available, successful complete occlusion was achieved in 18, while partial occlusion was achieved in 2. Follow-up data were unavailable for 16 treated patients due to death unrelated to CCF or its treatment (n = 2) or because subsequent care was provided elsewhere (n = 14). Seven of the patients who demonstrated complete post-treatment occlusion of the CCF still reported ophthalmologic symptoms on follow up, including diplopia related to persistent cranial nerve palsy (n = 6) and episodic chemosis plus retrobulbar pain (n = 1). Two patients experienced complications of endovascular treatment: one with new cranial nerve VI palsy, and the other with transient diplopia for 3 weeks postprocedure. Of the 4 patients managed by observation only, 1 spontaneously resolved, 2 had a persistent symptomatic CCF, and 1 was lost to follow up.
DISCUSSION CCF classically presents with various combinations of diplopia, proptosis, conjunctival chemosis, episcleral venous engorgement, and pulsatile tinnitus and with enlargement of the SOV on CT or MRI. However, CCF is often misdiagnosed, especially when typical clinical features or imaging findings are not present. One series reported that patients presenting
194
without orbital bruit were more likely to pose a diagnostic challenge.10 Barry et al.8 reported that initial misdiagnosis is more common for patients with indirect fistula (58% misdiagnosed) than direct (25% misdiagnosed), similar to the findings in this study. Patients with other traumatic deficits that confound the diagnostic picture, or with unusual demographics for CCF (such as pediatric age), are also more likely to be misdiagnosed initially.11,12 Patients with CCF are commonly evaluated by ophthalmology prior to diagnosis, and 40% of those in this series initially presented to their ophthalmologist. Nearly half of these patients were initially misdiagnosed. Delayed diagnosis and treatment may result in higher likelihood of persistent ophthalmologic deficits.8 In a 76-patient series, the average time from symptom onset to diagnosis was 16.1 weeks, with longer diagnostic delays occurring in patients who initially presented with cranial nerve palsy before the onset of other orbital congestive symptoms.13 Fistulas with a higher proportion of drainage via cortical veins were more likely to have an atypical presentation pattern resulting in diagnostic delays.13 This is consistent with other case reports in which CCF presented atypically with venous infarction, intracranial hemorrhage, dural venous sinus thrombosis, posterior ischemic optic neuropathy, isolated cranial nerve palsies, or elevated intracranial pressure mimicking pseudotumor cerebri, leading to initial misdiagnosis.14–20 It is standard teaching that a CCF with orbital clinical findings will demonstrate SOV enlargement on imaging.1,2,21–23 However, in this series, 26.7% (8 of 30) did not have SOV enlargement when studied with various noninvasive modalities (CT, CTA, MRI, MRA). There are several inherent sources of error in the interpretation of noninvasive images, including subtle venous asymmetry that can be undetectable, or bilateral venous drainage from the CCF leading to mild symmetric SOV enlargement that is difficult to appreciate.24 Three of the patients (8%) had no SOV enlargement on any imaging study, including catheter angiography. The question arises how the orbital venous system can be pressurized enough to cause clinical symptoms without distending the SOV. In all 3 cases, an alternate venous drainage pathway was present, apparently decompressing the SOV sufficiently to render it normal in caliber radiologically. Furthermore, vessel caliber is a function of not only pressure but also flow volume. As all 3 patients had drainage back through intracranial venous channels rather than forward through the orbit to the angular and facial veins, fistula flow through the orbit was limited. This lower flow state may have allowed the SOV to remain normal in caliber, despite relative pressurization of the orbital venous system. Another possible mechanism may relate to the presence of competent valves, decreasing the volume of retrograde flow from
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
the fistula into the SOV in these 3 patients. Although classically believed to be devoid of valves, recent anatomical studies have shown that valves are present in up to 75% of ophthalmic veins.25 Finally, SOV stenosis or occlusion could potentially prevent enlargement of the SOV despite elevated orbital venous pressure, but this was not seen in this series. There have been other reports of patients with ophthalmologic symptoms but no SOV enlargement on noninvasive imaging, including a case report of a 75-year-old woman with diplopia, episcleral venous injection, ophthalmoplegia, and ptosis, but a normal CT and MRI.26 Kurata et al.13 found that 3 of 51 patients with symptomatic orbital congestion had no appreciable drainage through the SOV on angiography. Conversely, a patient with radiologic evidence of CCF with SOV enlargement but without any ophthalmologic symptoms has also been reported.27 Several publications have suggested alternate means for assessing elevated venous pressure within the orbit to help diagnose CCF. One study found that the ocular pulse amplitude (difference between minimum and maximum intraocular pressure during the cardiac cycle, as measured by pneumotonometry) was significantly larger in CCF patients and was 100% sensitive and 93% specific in distinguishing CCF patients from normals or from those with other orbital diseases.28 Orbital color Doppler ultrasound of the retrobulbar vessels can detect reversal of flow in the SOV, helping distinguish CCF from other causes of SOV enlargement such as thyroid-associated orbitopathy and allowing noninvasive monitoring of the progression or resolution of the fistula over time.6,29,30 Blood flow velocity and pulsatility index parameters specific to CCF can reportedly be detected with transcranial Doppler ultrasound, even in lesions with normal CT or MRI findings.31 Appropriate diagnosis and management can result in partial or full resolution of signs and symptoms. In this series, 90% of patients for whom follow-up data were available showed full anatomical resolution of CCF after endovascular intervention comparable with the rates reported elsewhere ranging from 84% to 100%.8,9,32,33 The fistula can be accessed via orbital vessels or through a more distal vascular access point.34,35 In one review, embolization produced a higher rate of resolution (91%) than gamma knife radiation (72%), and the transvenous approach for embolization had a superior success rate (98%) compared with transarterial (88%).33 Resolution after stereotactic radiosurgery and spontaneous involution with observation alone have also been reported.36,37 Of interest, one patient in this series continued to experience congestive orbital symptoms after endovascular treatment, despite complete anatomical occlusion of the CCF demonstrated by angiography. Similarly, in a series of 60 patients with CCF who underwent endovascular occlusion, anatomical cure was seen more frequently (63%) than clinical cure (51%).3 This study is limited by its retrospective nature, including incomplete prereferral records and incomplete follow-up data. As a tertiary medical center, a high percentage of patients return to the referring provider after diagnosis or intervention. Effort was made to obtain reports and images from outside providers and to standardize the interpretation of the imaging studies through review of each image series by the authors. In conclusion, CCF can give rise to neurologic and ophthalmologic signs and symptoms. It is frequently misdiagnosed, especially when presenting with atypical clinical features or subtle imaging characteristics. SOV enlargement and other findings of fistulas may not be detectable with noninvasive imaging. Even with catheter angiography, 8% of patients with CCF and clinical signs of orbital congestion lacked SOV enlargement in this series. To avoid inappropriate interventions or delays in
Carotid Cavernous Fistula
diagnosis and care, patients with clinical features suspicious for CCF should undergo catheter angiography to identify fistula flow. Endovascular treatment can produce clinical improvement or resolution.
REFERENCES 1. Keltner JL, Satterfield D, Dublin AB, et al. Dural and carotid cavernous sinus fistulas. Diagnosis, management, and complications. Ophthalmology 1987;94:1585–600. 2. Miller NR. Dural carotid-cavernous fistulas: epidemiology, clinical presentation, and management. Neurosurg Clin N Am 2012;23:179–92. 3. Preechawat P, Narmkerd P, Jiarakongmun P, et al. Dural carotid cavernous sinus fistula: ocular characteristics, endovascular management and clinical outcome. J Med Assoc Thai 2008;91:852–8. 4. Théaudin M, Saint-Maurice JP, Chapot R, et al. Diagnosis and treatment of dural carotid-cavernous fistulas: a consecutive series of 27 patients. J Neurol Neurosurg Psychiatry 2007;78:174–9. 5. Yoshihara M, Saito N, Kashima Y, Ishikawa H. The Ishikawa classification of cavernous sinus lesions by clinic-anatomical findings. Jpn J Ophthalmol 2001;45:420–4. 6. Wei R, Cai J, Ma X, et al. Imaging diagnosis of enlarged superior ophthalmic vein. Zhonghua Yan Ke Za Zhi 2002;38:402–4. 7. Hu Y, Wang Z, Quan W, et al. Digital subtraction angiography of cerebrum for diagnosis of cavernous sinus fistula and its intravascular treatment. Zhonghua Yan Ke Za Zhi 1999;35:197–9. 8. Barry RC, Wilkinson M, Ahmed RM, et al. Interventional treatment of carotid cavernous fistula. J Clin Neurosci 2011;18:1072–9. 9. Gemmete JJ, Chaudhary N, Pandy A, Ansari S. Treatment of carotid cavernous fistula. Curr Treat Options Neurol 2010;12:43–53. 10. Taniguchi RM, Goree JA, Odom GL. Spontaneous carotid cavernous shunts presenting diagnostic problems. J Neurosurg 1971;35:384–91. 11. Gossman MD, Berlin AJ, Weinstein MA, et al. Spontaneous direct carotid-cavernous fistula in childhood. Ophthal Plast Reconstr Surg 1993;9:62–5. 12. Lau FH, Yuen HK, Rao SK, et al. Spontaneous carotid cavernous fistula in a pediatric patient: case report and review of literature. J AAPOS 2005;9:292–4. 13. Kurata A, Suzuki S, Iwamoto K, et al. Dural arteriovenous fistulas in the cavernous sinus: clinical research and treatment. ISRN Neurol 2011;2011:453834. 14. Miyamoto N, Naito I, Takatama S, et al. Clinical and angiographic characteristics of cavernous sinus dural arteriovenous fistulas manifesting as venous infarction and/or intracranial hemorrhage. Neuroradiology 2009;51:53–60. 15. Hashimoto M, Ohtsuka K, Suzuki Y, et al. A case of posterior ischemic optic neuropathy in a posterior-draining dural cavernous sinus fistula. J Neuroophthalmol 2005;25:176–9. 16. Kuwayama N. Classification and diagnosis of intracranial dural arteriovenous fistulas. Brain Nerve 2008;60:887–95. 17. Miyagishima T, Hara T, Inoue M, et al. Pontine venous congestion due to dural arteriovenous fistula of the cavernous sinus: Case report and review of the literature. Surg Neurol Int 2012;3:53. 18. Tsai RK, Chen HY, Wang HZ. Painful fourth cranial nerve palsy caused by posteriorly-draining dural carotid-cavernous sinus fistula. J Formos Med Assoc 2000;99:730–2. 19. Masaya-Anon P. Isolated oculomotor nerve palsy in a white-eyed patient with dural carotid-cavernous sinus fistulas: a case report. J Med Assoc Thai 2012;95(suppl 4):S143–6. 20. Oestreicher JH, Frueh BR. Carotid-cavernous fistula mimicking Graves’ eye disease. Ophthal Plast Reconstr Surg 1995;11:238–44. 21. Slusher MM, Lennington BR, Weaver RG, et al. Ophthalmic findings in dural arteriovenous shunts. Ophthalmology 1979;86:720–31. 22. Feiner L, Bennett J, Volpe NJ. Cavernous sinus fistulas: carotid cavernous fistulas and dural arteriovenous malformations. Curr Neurol Neurosci Rep 2003;3:415–20. 23. Miller NR. Diagnosis and management of dural carotid-cavernous fistulas. Neurosurg Focus 2007;23:E13. 24. Ros de San Pedro J, Pérez CJ, Parra JZ, et al. Bilateral ethmoidal dural arteriovenous fistula: unexpected surgical diagnosis. Clin Neurol Neurosurg 2010;112:903–8.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
195
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015
S. M. Jacobs et al.
25. Zhang J, Stringer MD. Ophthalmic and facial veins are not valveless. Clin Experiment Ophthalmol 2010;38:502–10. 26. Théaudin M, Chapot R, Vahedi K, et al. Dural carotid-cavernous fistula: relationship between evolution of clinical symptoms and venous drainage changes. Cerebrovasc Dis 2008;25:382–4. 27. Ikeda K, Deguchi K, Tsukaguchi M, et al. Absence of orbito-ocular signs in dural carotid-cavernous sinus fistula with a prominent anterior venous drainage. J Neurol Sci 2005;236:81–4. 28. Golnik KC, Miller NR. Diagnosis of cavernous sinus arteriovenous fistula by measurement of ocular pulse amplitude. Ophthalmology 1992;99:1146–52. 29. Hatton MP, Remulla HD, Tolentino MJ, et al. Clinical applications of color Doppler imaging in the management of orbital lesions. Ophthal Plast Reconstr Surg 2002;18:462–5. 30. Jianu DC, Jianu SN. Carotid-cavernous fistulas of low-flow type: color Doppler imaging of retrobulbar vessels findings. Oftalmologia 2010;54:83–90. 31. Kiliç T, Elmaci I, Bayri Y, et al. Value of transcranial Doppler ultrasonography in the diagnosis and follow-up of carotid-cavernous fistulae. Acta Neurochir (Wien) 2001;143:1257–64.
196
32. Jung KH, Kwon BJ, Chu K, et al. Clinical and angiographic factors related to the prognosis of cavernous sinus dural arteriovenous fistula. Neuroradiology 2011;53:983–92. 33. Choi BS, Park JW, Kim JL, et al. Treatment strategy based on multimodal management outcome of cavernous sinus dural arteriovenous fistula (CSDAVF). Neurointervention 2011;6: 6–12. 34. Badilla J, Haw C, Rootman J. Superior ophthalmic vein cannulation through a lateral orbitotomy for embolization of a cavernous dural fistula. Arch Ophthalmol 2007;125:1700–2. 35. Michels KS, Ng JD, Falardeau J, et al. Transvenous embolization of a dural carotid-cavernous sinus fistula via the inferior ophthalmic vein. Ophthal Plast Reconstr Surg 2007;23:480–2. 36. Chong GT, Mukundan S, Kirkpatrick JP, et al. Stereotactic radiosurgery in the treatment of a dural carotid-cavernous fistula. J Neuroophthalmol 2010;30:138–44. 37. Bujak M, Margolin E, Thompson A, et al. Spontaneous resolution of two dural carotid-cavernous fistulas presenting with optic neuropathy and marked congestive ophthalmopathy. J Neuroophthalmol 2010;30:222–7.
© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.
Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.
