Dural arteriovenous fistulae (dAVF) are anomalous connections between arterial feeders and dural venous sinuses that often present with symptoms including hemorrhage, focal neurologic deficits, seizures, and flow-related symptoms [12,13,18]. Surgical approaches can be employed for dAVFs that carry certain anatomical characteristics, however, dAVFs are most commonly treated endovascularly using either the transarterial or transvenous approach [3,8,14].

While large, multi-center efforts are underway to assess the natural history, treatment strategies, and clinical outcomes of dAVFs, large studies of this magnitude can preclude detailed description of treatment strategies and outcomes for rare subtypes of dAVFs [3]. One interesting subtype of dAVF is those affecting the anterior condylar confluence adjacent to the canal of the hypoglossal nerve [9]. Though little is known about these rare dAVFs, a small number of studies have been published describing the treatment of these lesions [9]. Building on these efforts, we share our experience with endovascular management of 3 unique cases of dAVFs affecting the anterior condylar confluence, providing details surrounding presentation, treatment strategy, imaging, and patient outcomes.

For procedural details of the cases reported in this study, see Table 1.

In this work, we report outcomes of endovascularly-treated dAVFs located at the anterior condylar confluence. Among 3 treated dAVFs, there were no technical complications or treatment-related clinical complications. At final follow-up, presenting symptoms had resolved in all patients and there was no angiographic recurrence of any dAVF.

While none have proposed pathophysiology specific to dAVFs in the anterior condylar confluence, it is likely that the etiology of these lesions is shared with dAVFs in other regions. Known risk factors for dAVFs include local infection, cranial trauma, prior intracranial surgery, and venous sinus thrombosis [13,17]. The majority of dAVFs are thought to be acquired lesions, though the presence of dAVFs in infants suggests that there may be a small number that are either congenital or secondary to antecedent insults occurring very early in life [17]. Inceptive studies in dAVF pathophysiology suggested that inflammation secondary to organization of sinus thrombosis was a trigger for the formation of aberrant arterial-venous communications, however subsequent evidence implicated angiogenesis induced by ischemia from venous hypertension as the main driver of dAVF development [4,7]. The etiology of dAVFs continues to be an area of active scientific interest, with recent studies implicating neutrophilic, monocytic, and cell-mediated inflammation as important players in the organization of venous thrombosis that may precede dAVF formation [2,5]. Despite these risk factors and potential etiologies of dAVFs, no patients in our group had a history of any previously described risk factor.

The anterior condylar confluence is an extracranial venous system at the skull base that comprises communications between regional veins and dural sinuses [15]. It is located medial to the jugular vein, anterior to the aperture of the hypoglossal canal, and inferior to the jugular bulb and inferior petrosal sinus [15]. This confluence provides connections between the dural venous sinuses of the posterior fossa, branches of the internal jugular vein, and branches of the vertebral venous system [11]. Prior work has suggested that the ascending pharyngeal artery is the main feeder for dAVFs in this location, and that contralateral arterial contribution is not uncommon [11]. Most cases have feeders from posterior meningeal branch of the VA. In our study, arterial supply was from the ascending pharyngeal artery in 3 dAVFs, from the occipital artery in 2 dAVFs, from the middle meningeal artery in 1 dAVF, and from branches of the vertebral artery in 1 dAVF. All arterial supply for dAVFs included in our study was ipsilateral to the fistula. Venous drainage was to the internal jugular vein in 3 cases, to the paraspinal venous plexus in 2 cases, to the inferior petrosal sinus in 2 cases, to the cavernous sinus in 1 case, and to the ophthalmic veins in 2 cases. Venous ectasia was present in 1 dAVF, while cortical venous reflux was not observed in any dAVFs.

The proximity of the anterior condylar confluence to important structures is easily correlated with presenting symptoms of patients in our study and prior studies. Two patients in our study had auditory symptoms at presentation (1 case each of pulsatile tinnitus and audible bruit), and two had ocular symptoms (1 patient with 6th nerve palsy and ocular chemosis, and 1 patient with proptosis and ocular hyperemia). Prior studies have reported similar ocular symptoms and auditory symptoms, in addition to hoarseness from 12th nerve compression and facial droop from 7th nerve compression [1,9,11]. The shunting that occurs near the petrous bone can be the cause of the tinnitus and pulsatile bruit, while ocular symptoms are likely due to the anterior draining pattern of these dAVFs [6]. It has been noted that the drainage and anatomy of these dAVFs make them similar to carotid-cavernous fistulae (CCF) [11]. In our series, 2 of 3 were initially diagnosed as CCFs before being re-classified due to catheter angiography studies. The proximity of the anterior condylar confluence to crucial structures also influences treatment approaches. Besides favoring endovascular approaches to avoid approach-related complications, exquisite caution must be taken when performing these interventions. When embolizing these lesions, dense packing of the anterior condylar vein with embolics should be avoided in order to minimize the risk of postoperative cranial nerve XII palsy, which has been observed in previous studies [11]. Additionally, transvenous approaches are commonly favored over transarterial approaches given the risk of embolic stroke in the latter [10,16]. Because of the inherent risk of treating dAVFs in this region, even when using minimally invasive endovascular approaches, further work is necessary to identify additional therapeutic avenues.

There are several limitations to this study that are common to studies of this type. Our study is an observational case series including a small number of cases due to the overall rarity of these lesions. Therefore, it is unclear if our findings are generalizable to all dAVFs of the condylar canal. In addition, our single-center perspective leads to treatment approaches described to reflect preferences of a small group of operators, further limiting the generalizability of our findings. Finally, this small series reflects careful patient selection and therefore our experience may not reflect all the technical nuances of treatment of dAVFs found in this location.

Treatment of dAVFs located in the anterior condylar confluence can be achieved endovascularly using both transarterial and transvenous approaches. No dAVF recanalization or treatment-related complications were observed.