Background Hypoglossal canal dural arteriovenous fistulae (HC-dAVF) are a rare subtype of skull base fistulae involving the anterior condylar confluence or anterior condular vein within the hypoglossal canal. Transvenous coil embolization is a preferred treatment strategy, however delineation of fistula angio-architecture during workup and localization of microcatheter tip during embolization remain challenging on planar DSA. For this reason, our group have utilized intra-operative cone beam CT (CBCT) and selective cone beam CT angiography (sCBCTA) as adjuncts to planar DSA during workup and treatment. The purpose of this article is to present our experience in the treatment of HC-dAVF using transvenous coil embolization (TVCE) with cone beam CT assistance, describing our technique as well as presenting our angiographic and clinical outcomes.
Methods Ten patients with symptomatic HC-dAVF were treated using TVCE with intra-operative cone beam CT assistance. Prospectively collected data regarding clinical and angiographic results and complication rates was recorded and reviewed.
Results Complication-free fistula occlusion was achieved in our entire patient cohort. The dominant symptom of pulsatile tinnitus resolved in all 10 patients.
Conclusions This study demonstrates that TVCE with CBCT assistance is a highly effective treatment option for HC-dAVF, achieving complication-free fistula occlusion in our entire patient cohort. We have found low-dose sCBCTA and CBCT to be an extremely useful adjunct to planar DSA imaging during both workup and treatment of these rare fistulae.
- vascular malformation
- ct angiography
Statistics from Altmetric.com
Background and purpose
Hypoglossal canal dural arteriovenous fistulae (HC-dAVF) are a rare subtype of skull base fistula involving either the anterior condylar vein (ACV) as it traverses the hypoglossal canal or the anterior condylar confluence (ACC) lying at its extra-cranial aperture. They account for less than 5% of dAVF cases, with a recent systematic review identifying only 120 reported HC-dAVF that have undergone intervention.1
Surgery has a limited role to play in the treatment of HC-dAVF while trans-arterial therapy often requires embolization utilizing liquid embolic agents via the ascending pharyngeal artery. Although the ascending pharyngeal artery approach has been successfully utilized for dAVF embolization,2 the procedure does carry a significant risk due to the potential for non-target embolization of arteries supplying the vasa nervosum of the 9th to 12th cranial nerves.1 3 Trans-venous embolization using coils has therefore become a favored treatment technique achieving high curative rates with relatively low morbidity.1 4 5 However, mapping the complex arterial and venous anatomy of these fistulae, and the subsequent trans-venous navigation of catheters into the hypoglossal canal, remain challenging using planar digital subtraction angiography (DSA).
Our group have found intra-operative selective cone beam CT angiography (sCBCTA) and intra-operative cone beam CT (CBCT) to be extremely useful adjuncts to planar DSA both during workup and treatment of these fistulae. We present our experience in the treatment of HC-dAVF using transvenous coil embolization (TVCE) with cone beam CT assistance, describing our technique as well as our angiographic and clinical outcomes.
Institutional ethics board approval was obtained for this retrospective review of prospectively collected data in our institution’s dAVF database.
All patients referred to our institution with suspected HC-dAVF on clinical assessment and MR imaging subsequently underwent a cerebral DSA. In addition to standard bilateral internal carotid, external carotid, and vertebral artery angiograms, selective angiography of the dominant arterial branches supplying the fistula was undertaken. Selective planar DSA and low-dose selective cone beam CT angiography (sCBCTA) were performed via the arterial branch which produced optimum fistula opacification in order to confirm the diagnosis of a HC-dAVF and to aid in treatment planning (figure 1). Our sCBCTA protocol involved a 5 s, 70kV CBCT rotation, tightly coned around the hypoglossal canal with a frame acquisition rate of one frame per 1.5-degree rotation. The volume of contrast utilized was between 2–5 mL injected at a rate of between 1–3 mL/sec depending on the size of the catheterized artery .
The arterial supply and venous drainage of fistula was prospectively documented at the time of the planning DSA and the venous drainage pattern of the fistula was classified according to Cognard’s and Spittau’s systems.1 6
Endovascular treatment technique
All TVCE procedures were performed by one of four experienced interventional neuroradiologists, under general anesthetic, and using a bi-planar angiography unit with cone beam CT capability (Artis zee, Seimens, Erlangen, Germany).
Venous access was via ultrasound guided common femoral vein puncture. Equipment chosen to perform the procedure differed slightly between operators, however all negotiated a bi-axial or tri-axial venous system into the IJV ipsilateral to the HC-dAVF. The most commonly utilized equipment was a tri-axial system composed of a 6Fr Shuttle sheath (Cook Medical, Bloomington In. USA), 100 cm Envoy 070 (Codman Neuro, Raynham, MA, USA), an Echelon-10 017 microcatheter (Medtronic, Dublin, Ireland), and a Traxcess 014 micro-wire (Microvention, Aliso Viejo, CA, USA).
Following ultrasound-guided common femoral artery access, the arterial branch identified on planning DSA as producing optimum opacification of the HC-dAVF was directly catheterized using a 5Fr diagnostic catheter. Angiographic runs were performed through this catheter and roadmaps obtained which allowed negotiation of the microcatheter from the IJV to the point of fistulation within the ACV or ACC. CBCT or sCBCTA was then performed, either directly after positioning of the microcatheter or prior to deployment of the first coil, in order to confirm position within the hypoglossal canal at the point of fistulation. Axium Prime Extra Soft coils (Medtronic, Dublin, Ireland) were utilized to pack the ACC and ACV until cessation of flow within the fistula was achieved (figure 2).
Angiographic and clinical outcomes
Angiographic outcomes were assessed at the end of the TVCE procedure and on MRI/MRA performed 6 months following embolization. In order to define the angiographic outcome, two experienced neuroradiologists, one of whom was not involved in treatment, assessed all imaging. Angiographic outcomes were classified as complete fistula occlusion (no residual fistula visible), subtotal fistula occlusion (75%–99% of fistula obliterated), and residual fistula (<75% of fistula obliterated).
Clinical outcomes were recorded on day 1 post-TVCE and at outpatient clinic assessment 6 months post-treatment. Subjective symptom improvement or deterioration was categorized as complete symptom resolution, symptom improvement, or symptom deterioration. The Modified Rankin Score was also recorded at 6-month follow-up.
All 10 patients experienced pulsatile tinnitus as their dominant symptom. Four patients also suffered from headaches and two experienced hypoglossal nerve palsies. Mean duration of these symptoms prior to endovascular treatment was 4.7 months (range 2–8 months).
The dominant arterial supply to the HC-dAVF was via neuromeningeal branches of the ipsilateral ascending pharyngeal artery in nine cases and in six of these fistulae contralateral ascending pharyngeal artery supply was also seen. Numerous other external carotid artery branches contributed supply, with the occipital and middle meningeal artery most commonly involved. Internal carotid artery supplying the fistula was noted in six cases. This was always minimal and most commonly involving the ipsilateral meningiohypophyseal trunk. Arterial supply from the vertebrobasilar system was seen in five cases. Nine fistula demonstrated antegrade type 1 venous drainage into the IJV via condylar veins with one fistula demonstrating a type 2 retrograde venous drainage pattern into the IPS, cavernous sinus, and ophthalmic veins. In eight cases the Cognard grade was 2A and in one case each grade 1 and grade 2A+2B. A summary of the angiographic characteristics of the HC-dAVF in our patient cohort is provided in table 1.
CBCT or sCBCTA was performed in all cases during the TVCE procedure. In two cases positioning of the microcatheter tip was thought to be incorrect on planar DSA imaging, however CBCT confirmed satisfactory position meaning that multiple further planar DSA runs were not required. In one case the first coil was felt to be malpositioned outside the hypoglossal canal, however CBCT performed prior to coil detachment confirmed that the coil was confined to the ACV with the hypoglossal canal and the coil was then deployed. In a further case CBCT performed prior to detachment of the first coil detachment demonstrated that the coil lay proximal to the point of fistulation within the ACC: the coil was then re-sheathed and the microcatheter advanced into the ACV with subsequent successful emboilization.
Immediate angiographic fistula occlusion was achieved in all 10 cases following TVCE with no residual or recurrent fistulae seen on 6-month MRA follow-up. Complete symptom resolution was achieved in nine patients at 6 months. In one patient the dominant symptom of pulsatile tinnitus resolved, however, although much improved, a mild hypoglossal nerve palsy persisted. No complications related to TVCE were reported at 6-month clinic follow-up. The angiographic and clinical outcomes of our patient cohort is summarized in table 1.
Dural arteriovenous fistula (dAVF) are pathological communications between dural venous structures and the dural arteries which account for 10% to 15% of intracranial arteriovenous malformations.7–9
HC-dAVF are uncommon, accounting for less than 5% of dAVF with a recent systematic review identifying just 120 cases that have undergone endovascular or surgical intervention.1 The largest case series identified in this systematic review consisted of just 14 patients with other documented HC-dAVF being recorded as case reports or small series.1 10
HC-dAVF involve the anterior condylar vein (ACV), which traverses the hypoglossal canal, or the anterior condylar confluence (ACC), a small venous pouch measuring approximately 5 mm in diameter which lies just outside the hypoglossal canal at its extra-cranial aperture.1 5 11 12 The ACC is a major venous crossroads at the skull base with caudal connections to the ACV, lateral condylar vein, internal jugular vein, and vertebral venous plexus as well as cranial connections to the inferior petrosal sinus, cavernous sinus, and the superior ophthalmic vein.11 Consequently a dural arteriovenous fistula in this location can lead to arterialization of not only the ACC and ACV but also some or all of these communicating venous structures.
Clinical symptoms of patients presenting with HC-dAVF vary according to the dominant venous drainage pathways, with the most common presentation being pulsatile tinnitus and headache. However, patients may also present with orbital proptosis and chemosis due to arterialization of the superior ophthalmic vein or hypoglossal nerve palsy due to compression of the hypoglossal nerve within the canal.1
Our initial step with all referred patients was to perform a workup cerebral DSA in order to confirm the suspected diagnosis and to define the fistula angio-architecture. In spite of the high spatial and temporal resolution of planar DSA imaging, the complex arterial and venous anatomy associated with these fistulae and their location within the dense bone of the skull base makes accurately defining the angio-architecture and precise point of fistulation challenging. Our group have found that utilization of sCBCTA, performed via injection of the arterial feeder producing optimum fistula opacification on planar DSA, to be extremely useful in defining both the angio-architecture of these fistulae and their relationship to the hypoglossal canal. When reconstructed on a sharp, bone algorithm, our sCBCTA protocol allowed precise delineation of the point of fistulation within the hypoglossal canal (figure 1, Image F) and also demonstrated venous drainage pathways that would offer routes of trans-venous access to the fistulae. This low-dose protocol exposed patients to a mean DAP of just 638 microGy/cm2 (range 483–859 microGy/cm2), equivalent to a AP and lateral whole head external carotid artery DSA on our angiography unit.
Using the low-dose protocol described in the Methods section, CBCT without or without selective angiography (sCBCTA) was also utilized during the embolization phase in order to more accurately define the position of the trans-venously placed microcatheter in relation to the hypoglossal canal and adjacent bony structures. CBCT or sCBCTA was performed, either after positioning of the microcatheter or immediately prior to the deployment of the first coil. These cone beam images could then be compared with those of the workup sCBCTA to ensure correct positioning allowing deployment of the coil precisely at the point of fistulation within the hypoglossal canal. In four out of 10 cases, CBCT or sCBCTA performed during TCVE led to changes in the course of the procedure, either demonstrating correct positioning of the microcatheter or primary coil when positioning was in question on planar DSA imaging, or by demonstrating incorrect placement of initial coil leading to coil withdrawal and microcatheter repositioning.
Although absolute numbers of patients in our cohort are small, the rarity of symptomatic HC-dAVF requiring intervention means that this is the second largest cohort of patients to be published and the largest treated using a standardized trans-venous approach. The technique of TVCE with CBCT assistance has produced excellent angiographic and clinical results with complication-free fistula occlusion achieved in the entire cohort. We have found on table sCBCTA and CBCT to be an extremely useful adjunct to planar DSA imaging, providing greater accuracy in diagnosis and greater precision during embolization. Utilization of the low-dose CBCT protocol described means that radiation exposure to the patient from CBCT is minimal and equivalent to a single external carotid artery DSA run: in fact by utilizing CBCT the overall radiation exposure to the patient may in fact be reduced by negating the need for multiple planar DSA runs to accurately define the fistula angio-architecture and microcatheter or coil position.
This study demonstrates that TVCE with CBCT assistance is a highly effective treatment option for HC-dAVF achieving complication-free fistula occlusion in our entire patient cohort. We have found low-dose sCBCTA and CBCT to be an extremely useful adjunct to planar DSA imaging during both workup and treatment of these uncommon fistulae.
Mr Roy McFaul, Medical Imaging Technician, assistance with image reproduction.
Contributors MTC; conception and design of study; acquisition, analysis, and interpretation of data; writing of manuscript. AHYC, TPC, WM; acquisition, analysis, and interpretation of data. TJP; acquisition, analysis, and interpretation of data, revision of manuscript and approval of final version to be published.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial, or not-for-profit sectors.
Competing interests None declared.
Patient consent Not required.
Ethics approval Western Australia Health GEKO committee.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.