Balloon protection of the Labbé vein during transarterial embolization of a dural arterio-venous fistula

Abstract

Introduction

Endovascular treatment of type III dural arterio-venous fistulas can be challenging if the fistulous point is close to a functionally important cortical vein.

Methods

A technique is described for temporary balloon protection of the vein of Labbé during transarterial Onyx embolization of a type III dural arterio-venous fistula. One illustrative case is presented. Careful anatomic consideration of the concerned venous segment (at the insertion point into the lateral sinus) and the choice of balloon minimized the risk of venous rupture.

Results

Using this method, satisfactory progression of Onyx was obtained within the arterio-venous shunt while preserving the patency of the Labbé vein.

Conclusion

Temporary balloon protection of the Labbé vein is a feasible option to preserve its patency during embolization of dural arterio-venous fistulas. To the authors’ knowledge, this is the first report on the use of temporary balloon protection of a cortical vein.

Introduction

Dural arterio-venous fistulas consist of pathologic direct arterio-venous shunts between multiple dural arterial feeders and a dural sinus or a cortical vein. Clinical presentations can vary from asymptomatic cases or pulsatile tinnitus to more dramatic neurologic events – intracranial venous hypertension, subarachnoid and intracerebral hemorrhage.^1,2 Two main classifications have been described,^3,4 both using the venous drainage pattern as the main criterion for risk stratification and treatment decisions.

Endovascular embolization is the preferred treatment modality for most cases that require intervention, through either a transvenous or a transarterial approach. In order to obtain a definitive cure, the main goal of all treatment methods is the occlusion of the foot of the draining vein or sinus.⁵ In some cases, the technical difficulty consists in identifying and preserving the patency of neighboring non-fistulous veins that drain the surrounding brain parenchyma.

We report on a case in which we used temporary balloon occlusion to protect the vein of Labbé (VL) during transarterial Onyx embolization of a dural arterio-venous fistula. To our knowledge, this is the first report on the use of this technique in a cortical vein.

Material and methods

Case description

A 59-year-old female patient was referred to our center for follow-up 3 years after embolization of a left indirect carotid-cavernous fistula. On clinical examination, the patient was asymptomatic, with no neurologic deficits and no signs of raised intracranial pressure. The cerebral angiography showed complete occlusion of the treated fistula, but also revealed a newly developed right temporo-occipital arterio-venous dural fistula (Figure 1).

Figure 1.

Pre-treatment angiography. (a) Pre-treatment angiography – right external carotid artery injection. A type III temporo-occipital arterio-venous fistula is seen in the proximity of the lateral sinus. Arterial feeders consist of multiple branches of the middle meningeal artery and occipital artery. The fistula is drained through a network of dilated parieto-occipital ascending veins directed towards the superior sagittal sinus. The posterior part of the VL is also visualized (arrowheads). For complete visualization of the VL, see also Figure 3(b). (b) Selective injection through a microcatheter that has been navigated close to the fistula in one of the branches of the middle meningeal artery. Again, the VL is partially visualized (arrowheads). Its insertion in the lateral sinus lies in close proximity to the fistula (also shown in Figure 2).

The arterial feeders consisted of multiple branches of the middle meningeal artery as well as two trans-osseous branches from the right occipital artery. The fistula was draining into a parieto-occipital vein joining the superior sagittal sinus. The arterio-venous shunts were located at the level of the junction between the VL and the lateral sinus (Figures 1 and 2). The VL had a disposition on the infero-lateral surface of the temporal lobe, ensuring the venous drainage of all the superficial sylvian territory and the right temporal lobe, without any other visible alternative drainage pathway.

Figure 2.

Embolization procedure. (a) Venous phase of an injection into the right ICA, superposed on an angiographic mask obtained by a previous selective injection through a microcatheter navigated close to the fistula in one of the branches of the middle meningeal artery. A balloon has been placed at the origin of the VL, its guide wire lying more distally into the vein. (b) Lateral and (c) antero-posterior non-subtracted views during embolization. Satisfactory progression of Onyx is observed within the fistula and the foot of its draining vein, with no diffusion into the VL, which is protected by the inflated balloon (white arrows).

The fistula was classified as type III according to the Lariboisière classification⁴ – direct venous drainage into a cortical vein. After discussion in a multidisciplinary meeting, we retained an indication for endovascular treatment, justified by the hemorrhagic risk associated with this type of dural fistula.

Endovascular treatment technique

The technical difficulty of the case consisted in the risk of extensive venous infarction in the case of embolic product migration within the VL. A previous diagnostic angiographic study permitted the identification of the fistulous arteries and veins, and their relations with neighboring cortical veins.

Under general anesthesia and therapeutic anticoagulation with continuous intra-venous heparin infusion, two 6 French sheaths were placed in the right and left femoral arteries. Additionally, a 6 F sheath was placed in the left femoral vein.

A 5 F vertebral catheter (Terumo Group, Tokio, Japan) was placed in the ipsilateral internal carotid artery (ICA) to allow control angiographic runs with visualization of normal cortical veins. A 6 F guiding catheter – Neuron 053® (Penumbra Inc., Almeda, CA) – was navigated distally into the ipsilateral internal maxillary artery. A second 6 F guiding catheter – Envoy® 6 F (Cordis, Johnson & Johnson Medical, Waterloo, Belgium) – was placed in the contralateral sigmoid sinus.

Through the contralateral venous guiding catheter, a 4 × 7 mm Hyperform® balloon (EV3 Neurovascular, Irvine, CA) was advanced through the torcula into the right lateral sinus, and further into the VL. The balloon was placed at the insertion of the vein into the sinus, at the same level with the arterio-venous fistulous point (Figure 2). All venous navigation was performed using the SmartMask™ (Philips Healthcare, Best, Netherlands) function, which created a roadmap using the venous phase of an ICA injection.

Successively, three detachable tip microcatheters – Sonic®1.5 F and 1.2 F (Balt Extrusion, Montmorency, France) – were used to catheterize two branches of the right middle meningeal artery and one branch of the occipital artery. Through each microcatheter, Onyx-18® (6% ethylene vinyl alcohol copolymer, Ev3 Neurovascular, Irvine, CA) was injected under fluoroscopic control using repeated image remasking and digital subtraction, which enabled the operator to visualize the diffusion of product at each injection. We used the established technique of ‘stacking,’ which involves the accumulation of multiple layers of Onyx around the microcatheter tip, creating a plug that promotes the advancement of the product into the fistula.

During each injection, the balloon was temporarily inflated in the VL to protect its patency in the case of migration of embolic product. We used balloon inflation cycles of no more than 3 min at a time, alternating with periods of at least 30 s to 1 min of balloon deflation in order to prevent venous congestion and infarction.

We obtained a satisfactory progression of the embolic product into the fistula and the foot of the draining vein, with no visible migration into the VL (Figure 2). Final angiographic runs performed through the external and ICA showed complete exclusion of the fistula with good patency of the VL (Figure 3). Moreover, the vein that used to drain the fistula became visible on the venous phase of the ICA injection, thus participating in the normal venous drainage of the right hemisphere.

Figure 3.

Final angiographic runs. (a) Right external carotid injection at the end of the embolization procedure showing complete exclusion of the fistula; (b) pre-treatment and (c) post-treatment right ICA injections – venous phases. The patency of the VL is preserved after embolization. Additionally, the draining vein of the fistula becomes visible on the ICA injection (arrowheads) and participates in the venous drainage of the right hemisphere.

At the end of the procedure, all catheters and sheaths were removed and the patient was transferred to a high dependency unit for 24 h. According to the institutional protocol, intravenous heparin was continued for 48 h and then continued with low molecular weight heparin administered subcutaneously for 2 more weeks, in order to prevent abrupt thrombosis of the draining veins.

Results

The patient had an uneventful postoperative recovery. At 10 months after the procedure, she remains completely asymptomatic. Cerebral angiography performed at 4 months confirmed complete cure of the fistula.

Discussion

The presented case illustrates the use of temporary balloon occlusion of a cortical vein during transarterial Onyx embolization.

The classical endovascular approach to dural arterio-venous fistulas initially consisted of transvenous access and coil occlusion of the diseased sinus segment. More recently, after newer liquid embolic products became available, the strategy gradually evolved towards transarterial embolization with preservation of venous sinus patency. To further improve this technique, several papers have communicated different methods of temporary balloon occlusion in the treatment of dural fistulas. The use of balloons to protect the patency of dural sinuses during transarterial embolization was initially described by René Chapot in an oral communication during the Val d’Isère Congress in 2008, on 22 clinical cases. Shi et al.⁶ published the same technique in 2009. More recently, Jittapiromsak et al.⁷ reported on the same technique but with the added use of the larger 8 × 80 mm balloon Copernic RC (Balt Extrusion, Montmorency, France), which was specifically designed for use within dural sinuses.

For the presented case, we applied the same principle, but we adapted the technique to the anatomy of a type III fistula, which by definition drains directly into a cortical vein. The structure that needed protection was not a dural sinus, but the VL, which in this case drained a large venous territory of the right hemisphere, and was situated in close proximity to the fistulous point. Although the fistula did not drain directly into the VL, the change of pressures during Onyx injection could have resulted in migration of embolic product during injection. Because of the vein’s functional importance, this could have led to severe clinical repercussions.

Moreover, the patency of the VL was particularly important in this context, as the fistula did not drain into the VL. In other cases, the vein drains the fistula, and normal brain parenchyma finds alternative routes for venous drainage. In these situations, proximal occlusion of the vein can be performed without clinical consequences. This was not the case for our patient, in whom the VL was large and was not draining the fistula.

The main risk associated with balloon inflation was the relative fragility of the cortical venous wall compared with arterial structures. Several studies have described the transition zone between cortical veins and the dural sinuses. Koperna et al.⁸ studied the VL in 22 anatomo-pathologic specimens. In all cases, the VL had an intradural trajectory at its insertion in the lateral sinus, situated into the wall of the sinus. A second study performed on 14 cadaveric specimens found that the VL ran within the tentorium from its initial entry point to its final drainage point to the transverse sinus over a varying length (2.8–24.5 mm).⁹ The results of these studies suggest that the insertion of the VL into the lateral sinus is contained within a dural structure. Consequently, in this portion of the vein, there is significantly less risk of rupture during balloon inflation.

To further reduce the theoretical risk of venous rupture, we purposefully chose a small dimension of 4 × 7 mm, to minimize the pressure exerted on the venous wall during inflation. Also, the small size of the guide wire (0.10”) reduces the risk of venous rupture during navigation. Balloon inflation and venous navigation were performed slowly and carefully, under roadmap guidance. According to the previously described histologic considerations, we precisely placed the balloon in the terminal part of the VL before its insertion in the lateral sinus.

Retrograde access to a cortical vein from the sinus is usually technically difficult. Frequently, instead of direct communication through a single lumen, there is a network of small channels in the wall of the dura, which could prevent selective placement of a balloon microcatheter in the cortical vein. We did not encounter such difficulties in the presented case.

As there are no recommendations for venous occlusion times, we used the same intervals as for arterial occlusions. Inflation time was limited to 3 min cycles with intermittent deflations of at least 30 s.

Conclusion

In selected cases, temporary balloon protection can be used to avoid VL occlusion during embolization of dural arterio-venous fistulas.

Ethical standards and patient consent

This is a retrospective report of a procedure performed as part of usual patient care according to local institutional protocols. All presented data are anonymized and the patient gave informed consent.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Conflict of interest

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

Authors’ contribution

R. Pop: Data collection, image processing, manuscript drafting, endovascular procedure

M. Manisor, Z. Aloraini: Endovascular procedure, manuscript review

V. Wolff, L. Tigan, C. Marescaux, P. Kehrli: Manuscript review

R. Beaujeux: Endovascular procedure, initial treatment concept, manuscript review