Focus on the target: Angiographic features of the fistulous point and prognosis of transvenous embolization of cavernous sinus dural arteriovenous fistula

Abstract

Background and purpose

Transvenous embolization (TVE) is widely utilized as an effective and safe treatment option for cavernous sinus dural arteriovenous fistula (CS-dAVF); however, detecting the exact location of the fistula is challenging. The present study identified the angiographic features of the fistulous point and evaluated the match with the microcatheter tip and fistulous point.

Materials and methods

An analysis cohort of 45 consecutive patients with CS-dAVF treated by TVE was analyzed retrospectively. The patients were divided into two groups, 22 matches and 23 mismatches, according to whether the fistulous point and the microcatheter tip were in the same compartment of the cavernous sinus (CS). The angiographic findings, the location of the fistulas, the position of the microcatheter tips, the volume of embolic materials, complications, and outcomes were assessed.

Results

Several angiographic features defined the fistulous points, such as the early opacified area, jellyfish-like sign, changes in the density of the contrast medium, the juncture of different arterial supply, enlarged feeders, and hand-injection angiograms. The fistulas were primarily in the posterosuperior portion of the CS (80%) and medial side (73.3%) according to the internal carotid artery. Both groups achieved effective TVE; the matched group required less embolic material than the mismatched group (p = 0.024). The patients with cranial nerve dysfunction (CND) required more embolic materials than others (p = 0.032).

Conclusion

The fistulous point in most of the CS-dAVFs could be isolated by careful analysis of the angiography images. The matching of the microcatheter tip and fistulous point in the same compartment of CS can reduce the dosage of embolic materials, and a low volume of embolic materials might cause fewer CND complications.

Introduction

Cavernous sinus dural arteriovenous fistulas (CS-dAVFs) are abnormal communications between the dural branches of the internal (ICA) and external carotid artery (ECA) and the cavernous sinus (CS).^1–3 CS-dAVFs have a benign natural course and a high incidence of spontaneous regression. ⁴ Nevertheless, endovascular treatment is indicated for high-risk lesions associated with cortical venous drainage, progressive visual loss, neurological deficits, or hemorrhage, as well as, in cases presenting with intolerable diplopia, severe headache, or severe cosmetic disfigurement.^5–8 Currently, there are two different endovascular treatment modalities for CS-dAVFs: transarterial embolization (TAE) of feeder vessels and transvenous embolization (TVE). Although the transvenous approach may be technically challenging, it offers a high success rate with a relatively low likelihood of serious complications as compared to TAE.^9,10 A dense packing of the entire CS with coils and Onyx (eV3, Irvine, CA, USA) may aggravate the symptoms, and embolic material volume might be correlated to the occurrence of cranial nerve dysfunction (CND). ¹¹

Targeted compartmental embolization of CS needs less embolic material to obliterate fistulas. However, in a majority of the cases, detecting the exact location of the fistula in the wall of the CS is challenging. ¹⁰ Thus, the first step is to find the target, the exact location of the fistula, and then to navigate the tip of the microcatheter into it to perform the embolization. Nevertheless, there are only a few studies addressing the angiographic features and locations of fistulous points in CS-dAVFs.^12,13

In this study, we sought to identify the angiographic features of the fistulous point and whether the matching of the microcatheter tip and fistula site are associated with embolic material volume and complications in CS-dAVF patients who underwent TVE.

Materials and methods

Study population and design

This retrospective study was approved by the Ethics Committee of Beijing Tiantan Hospital. All patients included in this study provided written informed consent to publish the related research results.

Patients diagnosed with CS-dAVF by conventional angiography at our hospital between December 2011 and May 2015 were eligible for enrollment in the present study. We conducted a retrospective analysis of 45 consecutive patients who underwent TVEs for CS-dAVF treatment, excluding those patients with any of the following: other treatment for CS-dAVF (TAE, gamma knife radiation, only manual compression), transorbital puncture of the CS, technically failed TVE, incomplete data for clinical evaluation, insufficient image quality for a radiological evaluation, or follow-up loss within six months after embolization.

The radiological findings, medical records, and clinical follow-up outcomes of each patient were retrospectively reviewed and evaluated, with an emphasis placed on evaluating the exact location of the fistula, the position of the tip of the microcatheter, the volume of embolic material, CND complications, and outcomes.

Treatment

Complete cerebral angiography of the bilateral ICA, ECA, and vertebral artery were performed in all patients to confirm the diagnosis and angiographic characteristics. All 45 patients were operated on using Onyx or Onyx coupled with detachable coils under general anesthesia by the cerebrovascular team at our hospital.

The TVE was initially attempted through the ipsilateral inferior petrosal sinus (IPS), as well as in cases with occluded IPS. A 6-French sheath was placed in the femoral vein while a 5-French sheath was placed in the contralateral femoral artery. A diagnostic catheter with a continuous heparinized flushing system was positioned in the common carotid artery for observation of the shunt, acquisition of roadmaps, and angiographic monitoring during the procedure. A guide catheter was placed in the ipsilateral internal jugular vein. Utilizing the road-mapping techniques, a microcatheter with a microwire was navigated into the CS. Then, Onyx injections were administered using real-time roadmaps or detachable platinum coils were first inserted to reduce the venous outflow and act as an attachment point for the subsequent Onyx injection. The embolization progress was monitored by intermittent angiograms that evaluated the fistula occlusion, status of the venous drainage, and patency of the ICA. The procedure was completed as soon as a control angiogram confirmed total obliteration of the fistula. If the ipsilateral IPS approach failed, the superior ophthalmic vein (SOV) through a facial vein or the contralateral IPS and inter-CS were also used. All the TVE techniques mentioned above have been described previously.^14–16 The angiographic definitions were as follows: a complete occlusion was a complete obliteration of the fistula, a subtotal occlusion was a minor residual shunt that was considered likely to thrombose, whereas a partial occlusion indicated the presence of a significant residual shunt. We also defined subtotal and partial occlusion as an incomplete occlusion.

Clinical evaluation and follow-up

We retrospectively collected data on patient demographics and clinical characteristics (age, sex, symptoms, symptom duration, treatment, and clinical outcome). All patients underwent a repeat of the post-embolization neurological and ophthalmological examinations. CND complications were defined as the aggravation of preexisting CND and TVE-induced CND. The patients commonly underwent a follow-up clinical evaluation at an interval of three to six months, as well as in the interim during any change or deterioration of the symptoms. In the case of any delay in patient recovery or a clinical suspicion of recurrent disease, an imaging study was recommended; 15 cases underwent radiological follow-up (14 received digital subtraction angiography (DSA) and one magnetic resonance imaging (MRI)). Recurrence was defined as the reappearance of the embolized fistula or development of a de novo fistula during the follow-up period after a successful embolization. The follow-up period ranged from 24 to 48 (average, 34.5) months. Herein, we evaluated the final clinical outcome of each patient at a follow-up visit or on the telephone, and stratified them into normal (asymptomatic patients) and abnormal (patients with residual primary symptoms or new symptoms without complete recovery).

Radiological evaluation

The radiological data of all patients were collected retrospectively, including pre-embolization, during embolization, post-embolization, and follow-up.

In order to analyze the location of the fistulous point and the tip of the microcatheter on the angiogram, we divided both sides of the CS into four compartments according to the relative positions based on the intracavernous ICA, anteroinferior/posterosuperior in the lateral projection, and medial/lateral (inside/outside) in the frontal projection (Figure 1). The patients were divided into two groups, matched or mismatched, according to localization of the fistulous point and the tip of the microcatheter—whether they were in the same compartment of CS.

Figure 1.

The left cavernous sinus was divided into four compartments approximately according to the (a) anteroinferior/posterosuperior in lateral projection, (b) medial/lateral in frontal projection, and (c) the relative positions to the intracavernous internal carotid artery.

We identified fistulous points synthetically based on several angiographic findings as follows (Figures 2 and 3).

1. The early opacified area of the CS in the early arterial phase of DSA.

2. Jellyfish-like sign: Numerous tenuous branches of arterial supply converged at the fistulas, similar to the tentacles connected to the body of jellyfish. The “jellyfish body” commonly represents the fistulous point.

3. The enlarged and tortuous branches of arterial supply can directly point to the fistulas.

4. The density of contrast medium changes from dark in the artery to gray in the veins on the fistulous point.

5. Different arterial supplies converge into the same venous drainage; the junction of different streams forms the fistulous point.

6. Subsequent to the advancement of the microcatheter tip into the CS, a hand-injection angiogram was performed to confirm the position before the injection of Onyx. The more venous drainage observed, the greater the proximity to the fistulous point.

Figure 2.

Angiographic features of fistulous points in the current cases (black arrows show the fistulous points). A1, A2 (case 1); B1, B2 (case 13): The fistulous points were difficult to define in the arterial phase of the lateral view of the common carotid artery (CCA) injection (A1, case 1) or the external carotid artery (ECA) injection (B1, case 13); the first early opacified area of the cavernous sinus (CS) in the early arterial phase shows the fistulous points (A2, B2). C1 (case 7); C2 (case 22); C3 (case 36): The “jellyfish like sign”: Lateral view of the ECA injection in the three cases shows numerous tenuous branches of arterial supply converging at the fistulas, similar to the tentacles connected to the body of a jellyfish. The “jellyfish body” represents the fistulous point. D1 (case 44); D2 (case 8): Lateral view of the ECA injection (D1, case 44) or the CCA injection (D2, case 8) shows the enlarged and tortuous branches of arterial supply pointing to the fistulous points. E1 (case 38); E2 (case 15): Lateral view of the ascending pharyngeal artery selective injection in the two cases shows that the density of contrast medium changes on the fistulous points.

Figure 3.

Angiographic features of the fistulous points in the current cases (black arrows show the fistulous points). A1, A2 (case 25): Anteroposterior view of bilateral external carotid artery (ECA) injection shows that different streams converge into the fistulous point. B1, B2 (case 22): Anteroposterior view of bilateral ICA injection shows different streams converging into the fistulous point. C1, C2 (case 15); D1, D2 (case 43): In two of the current cases, the hand-injection angiogram through the microcatheter show the venous drainage, after the microcatheter tip was inserted near the fistulous point through the superior ophthalmic vein.

The dense packing of the fistulous point can eventually cure the CS-dAVF, which can verify the preoperational judgment of the fistulous point.

According to the analysis of the angiographic features and the location of the fistulous point, the angiograms of each patient were reviewed by two neuroradiologists, independently, blinded to the clinical information.

The volume of each coil was calculated as the cross-sectional area × the coil length, and the cumulative coil volume added to the volume of Onyx represented the total volume of all embolic materials.

Statistical analysis

The results are expressed as mean ± standard deviation or numbers (percentages). The measurement data were analyzed via the independent sample t-test. The frequency distributions of categorical data were compared using Pearson’s chi-square tests or Fisher’s exact test. All statistical tests were two sided. A p < 0.05 was considered to indicate a statistically significant difference. Analyses were performed with SPSS 19.0.0 (IBM SPSS, Chicago, IL, USA).

Results

Overall characteristics

The present cohort consisted of 17 males and 28 females, with a mean age of 53.4 (range 24–81) years. The preoperative signs and symptoms included conjunctival congestion or chemosis in 41 cases, proptosis in 37, pulsatile bruit or tinnitus in 36, diplopia in 26, visual loss in 19, retroocular pain or headache in 16, facial numbness in two, hearing loss in one, and intracerebral hemorrhage-mediated epileptic seizure in one.

A total of 45 successful transvenous approaches were performed in this cohort, including the ipsilateral IPS approach in 23 cases, transcontralateral IPS and interCS in six, transfacial vein SOV in 14, trans-superior petrosal sinus (SPS) in one, and transpterygoid plexus in one.

Of the 45 CS-dAVFs, postoperative control angiography revealed complete occlusion in 31 lesions, subtotal occlusion in 11 cases, and partial occlusion in three cases.

The characteristics of the patients are summarized in Table 1.

Table 1.

Characteristics of the patients and results of transvenous embolizations.

Characteristics	Incidence (%)
Age (years)	24–81; mean (53.4)
Sex
Male	17 (37.8)
Female	28 (62.2)
Initial presenting symptoms
Conjunctival congestion or chemosis	41 (91.1)
Proptosis	37 (82.2)
Pulsatile bruit or tinnitus	36 (80)
Diplopia	26 (57.8)
Visual loss	19 (42.2)
Retroocular pain or headache	16 (35.6)
Facial numbness	2 (4.4)
Hearing loss	1 (2.2)
ICH leading to epileptic seizure	1 (2.2)
Venous approach routes
IPS	23 (51.1)
Inter-cavernous sinus via contralateral IPS	6 (13.3)
Facial vein-SOV	14 (31.1)
SPS	1 (2.2)
Pterygoid plexus	1 (2.2)
Angiographic results
Complete occlusion	31 (68.9)
Subtotal occlusion	11 (24.4)
Partial occlusion	3 (6.7)
Clinical outcome
Normal	40 (88.9)
Abnormal	5 (11.1)

ICH: intracerebral hemorrhage; IPS: inferior petrosal sinus; SOV: superior ophthalmic vein; SPS: superior petrosal sinus.

The incidence of angiographic features of the fistulous points in the patients in the present study are as follows.

1. The early opacified area of the CS: 30 (66.7%).

2. Jellyfish-like sign: 25 (55.6%).

3. The enlarged and tortuous branches of the arterial supply: 13 (28.9%).

4. The density of contrast medium changes: 35 (77.8%).

5. Different arterial supply converges into the fistulous point: 34 (75.6%).

6. Hand-injection angiogram via the microcatheter showed all the venous drainage:14 (31.1%).

The location of the fistulous point was on the left side only in 27 cases (60%), on the right side in 17 cases (37.8%), and bilateral in one case (2.2%). The case numbers of each compartment of CS with the fistulous point are listed in Table 2.

Table 2.

Location of fistulous points in each compartment of the cavernous sinus.

Location	Medial	Lateral	Total
Posteosuperior	30 (66.7%)	6 (13.3%)	36 (80%)
Anteroinferior	3 (6.7%)	6 (13.3%)	9 (20%)
Total	33 (73.3%)	12 (26.7%)	45 (100%)

The comparison between the matched group (22 cases) and mismatched group (23 cases) with respect to the age, duration, incidence of CND, angiographic result, and the clinical outcome did not show any significant differences. The microcatheter tip mismatched the fistula in 28.6% (4/14) of trans-SOV cases and 65.5% (19/29) of trans-IPS cases (p = 0.023). The Onyx dosage in the matched group was 2.81 ± 1.45 ml, while that in the mismatched group was 4.11 ± 2.19 ml; the difference was statistically significant (p = 0.024). The mean of coil numbers used in the matched group was 1.23 ± 1.38, while that in the mismatched group was 2.09 ± 1.47, without any statistically significant difference (p = 0.050; “borderline significant”). The mean of coil length used in the matched group was 18.27 ± 23.60 cm, while that in the mismatched group was 26.70 ± 20.01 cm; however, no statistically significant difference was observed (p = 0.203). The total volume of the embolic materials was 2.82 ± 1.46 ml in the matched group and 4.13 ± 2.20 ml in the mismatched group; the difference between the two groups was statistically significant (p = 0.024). The final clinical outcome was abnormal in two matched (visual loss in one and tinnitus in one) and three mismatched group patients (diplopia in one and tinnitus in two).

The comparison results of the two groups are listed in Table 3.

Table 3.

Comparison of matched and mismatched groups.

	Age (years)	Duration (months)	Approach (SOV/IPS)	Material				CND (yes/no)	Result (complete/ incomplete)
				Onyx (ml)	Coil length (cm)	Coil number	Onyx + coil (ml)
Match	52.1	5.2	10/10b	2.81	18.27	1.23	2.82	3/19	17/5
Mismatch	54.6	4.1	4/19	4.11	26.70	2.09	4.13	9/14	14/9
p	0.541	0.549	0.023 a	0.024 a	0.203	0.050	0.024 a	0.053	0.235

^ap < 0.05. ^bExcluded patients who underwent trans-superior petrosal sinus and transpterygoid plexus approach.

SOV: superior ophthalmic vein; IPS: inferior petrosal sinus; CND: cranial nerve dysfunction.

CND complications occurred in three matched and nine mismatched group patients, albeit without statistical significance (p = 0.053). The Onyx dosage in patients with CND was 4.50 ± 1.93 ml, while in the non-CND patients it was 3.10 ± 1.86 ml, which was significantly different (p = 0.032). However, the length and number of coils had no statistical difference.

Herein only one patient (case 6) was confirmed with the recurrence of fistula. The 24-year-old male underwent a complete occlusion of the fistula in a lateral anteroinferior compartment of the right-side CS. The microcatheter tip matched the fistula, and 2.5 ml Onyx (0.9 ml Onyx34, 1.6 ml Onyx18) was used. The patient complained of a recurrence of mild tinnitus seven months after embolization, and the symptoms were stable. The angiograms after 34 months post-embolization showed a mild recanalization of the previous fistula, and a new supply originated from the right middle meningeal artery. After manual compression of the right-side carotid artery for one month, the tinnitus totally vanished until the most recent follow-up of 42 months.

Discussion

Compared to the transverse-sigmoid sinus as the most frequent site of occurrence in Western countries, the CS is the most common location of dAVFs in Asians.^17–19

TVE is preferred over TAE because of its simplicity, low ischemic risk, high rate of success and ability to cure the fistula in a single session.^20,21 The TVE aimed to specifically catheterize the abnormal CS and occlude the fistula without rerouting venous drainage to the cortical structures. In TAE, the microcatheter tip can hardly reach the fistulous point through the small-sized arterial feeders. On the other hand, in TVE, an appropriate area in the enlarged CS connecting to several enlarged draining veins should be selected to perform the embolization.^13,15,22

Angioarchitecture and angiographic features of CS-dAVF

The angioarchitecture of CS-dAVF encompassed arterial feeders, arteriovenous shunt (AVS), fistula site, and draining veins. The major arterial supply arises from the internal maxillary, middle meningeal, accessory meningeal, and ascending pharyngeal branches of the ECA, as well as the dural branches of the ICA, usually from the meningohypophyseal trunk, inferolateral trunk or ophthalmic artery recurrent branches. ²⁰ The feeding arteries sent out a large amount of AVS that converge on a relatively confined location in the CS to form the fistula site. ²³ Commonly, the feeders are small in size and large in number, so the fistula site could be opacified earlier than AVS owing to the concentration of the contrast in the early arterial phase of DSA, which was seen in 66.7% of our patients (30 cases). Subsequently, a large amount of AVS was visualized clearly because of the concentrated contrast on the fistula site, similar to the tentacles connected to the body of a jellyfish. This phenomenon was observed in 55.6% (25 cases) of our series, which we termed a “jellyfish-like sign.” Some of our cases (28.9%, 13 cases) did not exhibit abundant AVS, representing only a limited number of enlarged and tortuous feeders as a result of the accelerated shunt flow.

The multiple arterial feeders usually appeared slim and black on angiography, while the dilated CS and draining veins were broad and gray. The density of the contrast medium changes on the fistula site was noted in 77.8% (35) cases in our series.

The blood flow from different arterial supplies converged into the fistulous point, which is a relatively circumscribed location. Therefore, the juncture was easily identified when the angiographic images of different sides of the arterial supply were superimposed. The images of 34 cases (75.6%) in our series were treated using Adobe Photoshop CS6 (Adobe, San Jose, CA, USA) to define the location of the fistulas. A minimum of two layers of semitransparent images of different arterial supplies in the same position were superimposed over one another, coinciding with the shape of the skull and same drainage veins, due to which, the point of juncture will appear clearly. After the blood flow from different arterial supplies was mixed in the fistulous point, the same draining venous system was shared. Once the microcatheter tip was advanced into the fistula by transvenous approach, a hand-injection angiogram showed all venous drainage in theory. In 31.1% (14) cases of our series, the hand-injection angiogram through the microcatheter showed all the venous drainage.

A complete set of six arterial selective diagnostic angiographic images is essential to identify the exact location of fistulas. The position and angle of projection of the patient should be the same to make it easier to analyze by image overlaying using the software later. The oblique view or super-selective angiography is essential for some complex cases. The high rate of angiographic frame imaging (>5 frames/s) and rapid contrast injection rates (7–8 ml/s) or manual compression of the ipsilateral common carotid artery may assist in evaluating high-flow fistulas. ²⁰

Injecting the contralateral ICA and ECA allows better delineation of the fistula site in bilaterally supplied CS-dAVFs. In some cases, the fistula will emerge clearer after a few draining veins are partially embolized.

Although isolating the exact location of a fistula in complex cases is difficult, we can limit the scope in a relatively confined compartment in all the cases by careful analysis of the angiography.

Location of the fistulous point

The CS is divided into several chambers by trabecula; that which is involved in a fistula is termed the “culprit chamber.”^24,25 A few reports described the location of fistulas of CS-dAVF. The fistulas of CS-dAVF were usually posteriorly situated in the CS,^12,13,26 which was in agreement with our data. Fistulas were primarily situated at the posterosuperior region of the CS as assessed by the lateral projection angiography in 80% of our patients. At the same time, 73.3% of fistulas were located on the medial side of the ICA in frontal projection angiography. However, Yoshida et al. reported a complication of brainstem infarction after dense packing of the fistula located at the posterior wall of the CS. ²⁷

Furthermore, the dense packing in the posterolateral region of the CS adjacent to the lateral surface of the posterior bend of the intracavernous ICA can potentially cause new or aggravated abducens nerve palsy that might be attributed to the oculomotor, trochlear, ophthalmic, and maxillary nerves located in the lateral wall of the CS; the abducens nerve is located just lateral to the ICA.^8,11,27 This anatomical disposition may allow further vulnerability of the abducens nerve to the stretching and mass effects of the coil or thrombus within the sinus.

Thus, the medial side posterosuperior portion of the CS is a relatively safe compartment that harbors the fistulas in the majority of our cases (66.7%).

Approach and matching

The current microcatheter technology permits access to the CS via multiple routes. The inferior petrosal sinus route represents the easiest, shortest, safest, and the most commonly used venous approach even in patients with IPS thrombosis or occlusion. ¹⁹ In the current cohort, the inter-CS approach is also mediated via contralateral IPS. The anterior approach through SOV provides a convenient alternative pathway in the event of unsuccessful cannulation of the IPS, especially the anterior drained cases with dilated facial veins.^20,28 For the slow injection of Onyx, a controllable nonadhesive agent, the position instead of the orientation of the microcatheter tip is crucial.

In the current series, the cases with the IPS approach might potentially (19/29) miss more fistulas compared with the SOV approach (4/14) (p = 0.023), which might be attributed to the requirement of deep navigation of the microcatheter in the CS in order to achieve complete occlusion of the anterior drainage of the SOV, causing a majority of the ophthalmic symptoms.

Volume of embolic materials and CND complications

CS-dAVFs presenting a high-risk pattern such as cortical venous drainage and intolerable symptoms may be effectively treated by TVE. However, the TVE of CS can be associated with various complications, some of which can potentially result in a significant morbidity. The total number of incidences of complications associated with the procedure was high, although most were transient morbidities; however, permanent deficits have been reported extensively.^8,11,14,27

Cranial nerve palsies after TVE were the most common complications. These symptoms may be attributed to progressive thrombosis of the CS, mass effect from the coils and Onyx, or direct injury to the nerve by coils or the microwire/microcatheter, and inflammation caused by dimethyl sulfoxide (DMSO) in Onyx.^6,8,16 The aggravation of preexisting CN symptoms or new CND is caused by the overpacking of endovascular materials in the CS.^{8,11,13,20,29} A coil volume >0.2 cm³ frequently results in the development or aggravation of CND. ¹¹ A volume of 0.2 cm³ corresponds to 333 cm for 0.018-inch coils, 400 cm for 0.011-inch coils, and 464 cm for 0.010-inch coils.

The targeted embolization of a single compartment of the CS involved in the fistula may avoid complications such as CND. Therefore, the precise identification of the fistulous points is crucial for effective targeted compartmental embolization.^13,22

In the current study, the volume of coils does not differ significantly between the matched and mismatched groups. In this study, the aim of coil placement in the patients is to prevent the spillage of Onyx into critical draining veins such as the SOV or cortical veins, or provide secure anchoring to Onyx, or decrease the flow rate in a high-flow fistula, instead of seeking a dense packing of the CS with coils only.^29,30 However, the mismatched group needs more Onyx than the matched group for successful embolizations (p = 0.024). The incidence of CND complications in the mismatched group (9/23) was higher than that in the matched group (3/22), although no statistical significance was attained (p = 0.053). This result could be attributed to the small sample size because the cases with CND cost more in endovascular materials (especially, Onyx) than others (p = 0.032). Thus, it can be speculated that the matching of the microcatheter tip with the fistula can reduce the dosage of Onyx, and less Onyx may lead to fewer CNDs.

On the other hand, the strategy of Onyx injection during the procedure is also vital. Onyx has a lava-like flow pattern within the blood vessels, different from n-butyl cyanoacrylate (n-BCA), another widely used liquid embolic agent. ³¹ The “reflux-hold-reinjection” and “plug and push” techniques should be used together in the embolization of CS-dAVFs.^14,30,31 Owing to the pressure gradient among the feeders, fistula, and draining veins, Onyx is inclined to cast into the draining veins instead of the fistula and feeders. Thus, Onyx 34 (contains 8% ethylene-vinyl alcohol (EVOH) copolymer), with higher viscosity and stability than Onyx 18 (contains 6% EVOH), is commonly used to form a stable core of a plug, which can be enlarged in the CS by the following slow injection of Onyx 18. When reflux to an undesirable direction is seen, the injection should be held for several seconds or a few minutes. ³⁰ The reinjection may change the direction of the Onyx casting until the pressure-pushed Onyx diffuses through the fistula and fills the end of the AVSs.

Coils are used to assist Onyx embolization, as the detachable coils are more controllable than the liquid agent. A few coils should be placed in the CS, especially at the location adjacent to the fistulas to decrease the flow in high-flow fistulas, which would make slow injection possible. The coils also provide a secure anchoring to the Onyx cast. ³⁰ Furthermore, the use of coils together with Onyx cost effectively reduced the number of coils required for dense packing.

In some early cases, we placed the coils at the proximal aspect of the SOV or cortical venous to prevent excess Onyx spillage into the draining veins; subsequently, the Onyx was injected at the same location until the CS and fistula were effectively embolized, which might lead to some mismatched cases. In four of the latter cases, after placement of the coils at the proximal aspect of the SOV, the microcatheter tip was adjusted as close as possible to the fistulous point to achieve complete embolization with a little dosage of Onyx.

Notably, the present study has some limitations, including that this was a single-center retrospective study, the sample size was relatively small, the degree of flow rate and fistula size were not assessed, and the angiographic follow-up was lacking. Nevertheless, the data concerning the angiographic features, the relationship of embolic material volume with matching of the microcatheter tip and fistula, and complications associated with TVE of CS-dAVFs are limited, and the results of this study provide valuable information for the management of these lesions.

Conclusion

The fistulous points in most CS-dAVFs can be isolated by careful analysis of the angiography images. Fistulous points were mainly situated at the medial posterosuperior compartment of the CS. The matching of the microcatheter tip and fistulous point in the same compartment of the CS can reduce the dosage of embolic materials to achieve effective TVE. The low volume of embolic materials may cause fewer CND complications.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.