Article Text
PDF
Abstract
Objective The combination of coils and Onyx for the treatment of carotid-cavernous fistulas (CCFs) is an interesting new development. The purpose of the current study is to evaluate our preliminary experience with the combined use of coils and Onyx for the treatment of traumatic CCFs.
Methods Between April 2009 and July 2014, 16 patients with 17 traumatic CCFs were embolized with the so-called ‘armored concrete’ treatment modality using coils, Onyx-18, and a non-detachable balloon via the transarterial approach. The outcomes were assessed both clinically and radiologically. Digital subtraction angiography (DSA) follow-up was performed 3 or 6 months after endovascular treatment while clinical follow-up was continued until December 2014.
Results Obliteration of the CCFs was obtained with patency of the parent artery in all 16 cases. Follow-up DSA demonstrated stable occlusion of all the fistulas. Symptoms related to the CCFs were either resolved immediately or gradually over 2 months. No worsening of the cranial neuropathies was observed during the follow-up period which averaged 32.6 months.
Conclusions The ‘armored concrete’ treatment modality using coils, Onyx, and a non-detachable balloon promises to be a safe, economical, and effective alternative in the management of traumatic CCFs.
- Balloon
- Coil
- Fistula
- Embolic
- Intervention
Statistics from Altmetric.com
Introduction
Carotid-cavernous fistulas (CCFs) are direct arteriovenous connections between the internal carotid artery (ICA) and cavernous sinus.1 Although endovascular occlusion of the fistulas is currently the regular treatment for CCFs, various treatment modalities have had variable outcomes. Recently, coils and liquid embolization agents such as ethylene vinyl alcohol (Onyx; ev3 Endovascular) have increasingly been used in the treatment of CCFs.2–4 However, the combined use of coils and Onyx (called the ‘armored concrete’ treatment modality) has rarely been reported and case numbers are very limited.5 ,6 The purpose of the current study was to evaluate our preliminary experience with the use of the ‘armored concrete’ treatment modality to treat 16 patients with 17 traumatic CCFs.
Methods
Patient population
The records of 42 patients diagnosed with traumatic CCFs at our institute between April 2009 and July 2014 were reviewed. Of these patients, 16 (12 men and 4 women) of mean age 35.7 years (range 19–59 years) were managed with the ‘armored concrete’ treatment modality. Of the 16 cases, two had undergone detachable balloon embolization before being transferred to our center, one failed due to the small size of the tear, and another recurred because of early balloon deflation.
Clinical presentations
The onset of the first symptoms occurred between 3 weeks and 5 months after the trauma. All of the patients presented with pulsating tinnitus; 15 had proptosis and significant conjunctival chemosis. The other symptoms included disorder of ocular movement in 10 patients and acute vision deterioration in four patients. CT showed skull fractures in 11 patients and MRI demonstrated a dilated superior ophthalmic vein in 13 patients. The definitive diagnosis was established by digital subtraction angiography (DSA), which revealed 17 fistulas (one had bilateral CCFs). In addition to the blood supply from the ICA, blood supply from the external carotid artery was revealed in two cases. The venous drainage involved superior (15 patients) and inferior (12 patients) ophthalmic veins, superior (6 patients) and inferior (11 patients) petrosal sinuses, cortical veins (6 patients), and contralateral sinuses (5 patients). All cases were treated electively unless there was an emergency situation such as progressive loss of vision or intracerebral hemorrhage due to recruitment of cortical vein drainage.
Endovascular procedure
Under general endotracheal anesthesia and systematic heparinization, a 6 Fr guide catheter was introduced into the ICA harboring the orificium fistula. A non-detachable compliant balloon (Hyperglide balloon 4×20 mm (ev3 Neurovascular) in 8 cases; Scepter C balloon 4×20 mm (Microvention) in 8 cases) was placed in the ICA and left in position spanning the fistulous site. A microcatheter (Echelon14; ev3 Neurovascular) was then navigated through the fistulous point into the cavernous sinus over a 0.014 inch microwire. With the balloon inflated, angiography via a microcatheter was performed to confirm that the microcather tip was in the cavernous sinus. After deployment of 2–3 coils in the cavernous sinus to make a coil framework clinging to the tear, Onyx-18 was carefully injected through the microcatheter. The balloon was fully inflated during coil deployment or Onyx injection. By controlling the injection speed, Onyx-18 was allowed to diffuse within and around the coil scaffold. If an unwanted cast emerged, the injection was halted for 2 min to solidify the Onyx polymerization. To minimize the risk of embolic events we followed a 3-2-2 protocol: (1) Onyx injection for no more than 3 min; (2) Onyx solidification for 2 min; and (3) balloon deflation for 2 min. The protocol was repeated until angiography demonstrated complete occlusion of the fistula (figure 1). In the patient with bilateral fistulas we adopted a staged strategy to lower the risk of hyperperfusion syndrome.
Procedure of the ‘armored concrete’ treatment modality via the transarterial approach. A non-detachable balloon is left in position spanning the fistula site and a microcatheter is navigated into the cavernous sinus (A). Injection via the microcatheter confirms the microcatheter tip is in the cavernous sinus (B). The balloon is fully inflated and the coils are deployed to make a coil framework clinging to the fistula site (C). Onyx-18 is injected slowly and allowed to diffuse mainly within the coil framework (D). The non-detachable balloon is inflated during the procedure to demarcate the boundaries of the internal carotid artery (E) and to localize the fistula site (red arrow).
Results
Early outcomes
The endovascular procedure using coils, Onyx, and a non-detachable balloon was technically successful and uneventful. Obliteration of the CCFs was obtained with patency of the parent artery in all 16 cases. An average of 2.53 coils and 2.78 mL Onyx-18 was needed for complete occlusion of one fistula (see table 1). For the patient with bilateral fistulas, a staged procedure with an interval of 1 month was adopted, finally achieving total occlusion. Pulsating tinnitus disappeared immediately after treatment in all cases, while eye symptoms and neurologic deficits were resolved gradually within 2 months. Routine postoperative CT and MRI scanning were performed in 13 cases, revealing no ischemic events. In the patient with bilateral fistulas, minor hemorrhage in the basal ganglion region was revealed after embolization of the left CCF; however, there were no symptoms associated with the hemorrhage. After 1 month the right CCF was uneventfully occluded using the same method (figure 2).
Management of 16 patients with 17 CCFs using the so-called ‘armored concrete’ treatment modality
Staged treatment of a patient with bilateral carotid-cavernous fistulas (CCFs). Left internal carotid artery (ICA) injection shows the CCF draining into the superior ophthalmic vein, basal vein of Rosenthal, inferior petrosal sinuses, and contralateral sinuses (A1). The left-sided CCF is completed occluded with the ‘armored concrete’ treatment modality (A2). Anteroposterior view of the right ICA angiogram showing the CCF in the right side (B1). Branches from the external carotid artery join the supply of the right CCF (B2). One month later the right-sided CCF is uneventfully occluded using the same method (B3). Plain skull X-ray showing bilateral ‘armored concrete’ blockages (C). CT scan shows minor hemorrhage in the basal ganglion region after embolization of the left CCF (D), but no symptoms associated with the hemorrhage. Proptosis and conjunctival chemosis before treatment (E1) were successfully resolved (E2).
Late outcomes
Three-month follow-up in six patients and 6-month follow-up in 10 patients with DSA demonstrated stable occlusion of the fistulas and patency of the ICA. No worsening of the cranial neuropathies was observed during the follow-up period which ranged from 4 months to 5 years (mean 32.6 months).
Discussion
CCF was an intractable vascular disease before the endovascular era. Over the past few decades, fistula occlusion with a detachable balloon, which is a relatively simple and economical method, has been the standard treatment of choice in many countries and areas. However, complications associated with embolization with a detachable balloon are not uncommon.7–10 With the availability of new embolization materials such as coils, liquid embolic agents and stents, the chance of occluding the fistula while keeping the patency of the ICA has improved. However, treatments remain to be optimized and comparisons of the various treatment modalities are required. Coil embolization is frequently used in the treatment of CCFs.11 However, transarterial embolization of traumatic CCFs with coils may lead to problems of the coil protruding into the lumen of the ICA, particularly when the defect in the wall of the ICA is large. The recurrence of the fistula due to compression and/or migration of coils within the cavernous sinus over time has also been reported.12 ,13 Tight packing with coils is critical to avoid recurrence, but it inevitably causes a mass effect in the cavernous sinus.14 Stent-assisted coiling or cover stent placement is suggested and has been reported to be valuable in certain cases.15–17 Nevertheless, considering the lifetime of antiplatelet treatment, stenting is not a preferred option, especially in young patients. The liquid embolic agent Onyx has been shown to be useful in the embolization of arteriovenous malformations, dural arteriovenous fistulas, and other vascular diseases. Treatment modalities based on the exclusive use of Onyx have also recently been reported for the embolization of CCFs with promising results. All 23 cases with traumatic CCFs treated with Onyx by Yu and Huang were cured with no significant complications.3 However, the risk of Onyx reflux into the ICA as well as the overfilling of Onyx in the cavernous sinus makes this treatment method controversial. To make the best of coils, Onyx, and balloons, we designed an ‘armored concrete’ treatment modality. Indeed, the procedure was technically successful and all cases were cured with no neurological deficit. The advantages of this protocol are significant and include: (1) 2–3 coils (mean 2.53 coils per fistula) were sufficient to produce a coil framework clinging to the fistula, indicating less cost and mild mass effect; (2) the diffusion of Onyx was greatly limited to the coil framework, thus reducing the Onyx injection volume; and (3) more significantly, the combination of Onyx and coils produced a strong ‘armored concrete’ blockage which embolized the fistula reliably. Taken together, the combination of coils and Onyx makes embolization of CCFs less expensive and more effective.
In the transarterial approach, having a balloon inflated in the ICA with a trapped microcatheter has the following advantages: (1) it allows the demarcation of the ICA boundaries; (2) it helps to localize the fistulous site; (3) it protects the ICA by preventing accidental Onyx reflux; (4) it facilitates the adhesion and solidification of Onyx within the coil framework by blocking the blood flow; and (5) it stabilizes the microcatheter tip in the cavernous sinus during Onyx injection. To fully occlude the fistula and prevent Onyx from refluxing into the intracranial circulation, we prefer an elongated conformal balloon. It cannot be ignored that the application of the balloon increases the risk of plaque detachment, thrombosis formation and vessel dissection, all of which are associated with ischemic infarction.18 ,19 To reduce the adverse effects resulting from the balloon, systematic heparinization and intermittent deflation of the balloon during the procedure are essential.
Conclusion
CCF represents a direct communication between the ICA and cavernous sinus; however, the treatment method varies due to the different concepts and experiences in using embolization materials. The ‘armored concrete’ modality using a combination of coils, Onyx, and a non-detachable balloon has multiple advantages and promises to be an effective alternative in the treatment of CCFs, particularly in refractory and recurrent cases.
References
Footnotes
XZ and WG contributed equally.
Contributors XZ and WG wrote the article. RS and JPS collected the data; JY and XZC modified the manuscript. LG provided guidance. ZQC designed the proposal and interventional operation. QBZ contributed to the protocol and provided assistance with surgery.
Competing interests None declared.
Ethics approval The study was reviewed and approved by the institutional review board at Shanghai Tenth People's Hospital, Tongji University School of Medicine.
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.