Embolization of Dural Intracranial Fistulas
Disease: Brain Dural Arteriovenous Fistula (BDAVF) is an abnormal direct communication between dural arterial vasculature and cranial venous system. The recipient vein is most typically a dural sinus. Common sites include sigmoid sinus, followed by cavernous sinus (to be distinguished from direct CC fistula), falcotentorial junction (draining into straight sinus), torcula (usually from below), ethmoid fistula, among others. Most often, the patient is a middle-aged or elderly woman. Most are spontaneous: theories on pathogenesis exist, but none has been proven.
Presentation varies with location: sigmoid and torcula fistulas present with pulsatile tinnitus, which can be usually auscultated by pressing the bell of a stethoscope to the mastoid or occipital bone (objective tinnitus). Cavernous sinus fistulas often present with exophthalmos, chemosis, diplopia, or visual impairment, due to congestion of ophthalmic veins and secondary increase in intra-occular pressures. Ethmoid fisulas have a special propensity for rupture due to typical drainage into cortical veins rather than directly into a venous sinus. Large, or strategically located dural fistulas, lead to pathology by venous congestion — overwhelming capacity of recepient vein for antegrade drainage, with secondary retrograde reflux, eventually into brain veins, leading to seizures, hemorrhages, and various functional brain issues depending on location (balance issues for cerebellar /torcular area, temporal lobe issues for sigmoid sinus fisulas congesting the Labbe, etc). BDAVFs can be classified according to Borden or Cognard, both related to extent of cerebrospinal venous congestion – the former is more simple, the latter may be more predictive of clinical dysfunction. Simply put, the more reflux into cortical (pr spinal veins), the greater risk of hemorrhage or neurologic dysfunction.
Treatment strategy: A rational treatment approach with respect to disease entity as a whole, in combination with tactical procedural experience, carries the best chance of success. In our opinion, the overwhelming majority of BDAVFs represent a single hole fistula. The myriad feeders carpeting dura and adjacent tissue ultimately converge onto a single arterial channel, which then opens into a true vein. Sometimes this arterial channel may look like vein, based on its size and location; recognizing it as artery is key to success of embolization. For example, in the sigmoid sinus region, the true vein is almost always the sigmoid sinus. Anything else, particularly parallel to the sigmoid sinus, is an artery, even if it looks like vein. Which artery — the jugular branch from the ascending pharyngeal artery, or transmastoid branch of the occipital artery, for example. Permanent closure of the fistula can only be accomplished by completely closing the venous side of the hole. This can be accomplished from either arterial or venous end, or both.
Arterial approaches: Typically involve microcatheterization of feeding dural artery, as close to fistula as possible, and subsequent embolization with n-BCA or Onyx. Whatever you use, the key is that embolic material must seal the venous opening of the common channel — which, for practical purposes, means that it should spill into the true vein. In case of sigmoid sinus fistula, the glue has to reach the sinus. For BDAVF embolizations, our choice is glue. Usually, meningeal vessels are a better choice to go further to the fistula. Occipital and cutaneous arteries are too tortuous, and have little support. If you can’t get far enough, chances are Onyx will not take you all the way to the vein — but it will paint everything else.
Venous approaches: Stuff the vein/sinus with coil, as tight as possible, making sure to cover the hole. Sometimes, one can get the microcatheter accross the fistula into the common arterial channel, and start coiling there. Issues arise when sinus also receives a large cortical vein, such as Labbe. With more advanced fistulas already backing up into cortical veins, it is less of a problem as the vein is not used by the brain anyway.
Combined arteriovenous approaches: For sure kill, you can coil off the sinus, and then glue into the coil mass from arterial side, to seal it off. May be necessary when coil is not tight enough, and the fistula stays alive on control arterial runs.
Vein-sparing approach: The ultimate, so to speak, in dural fistula cure — kill the fistula, and save the vein. Our current method, in use whenever possible, is to place a stent into the venous sinus across the fistula, and then glue from the arterial side. Glue typically polymerizes quickly when it comes in contact with the stent, and tends to stay next to the stent, keeping the sinus open. A second stent can be added after glue shot, if too much is hanging in the breeze. An even better method, perhaps, is to inflate a large compliant balloons such as the Copernic RC within the sinus across the fistula, to protect the sinus during Onyx of glue embolization. Unfortunately, the Copernic is not available in US except on compassionate use basis.
For a more basic discussion of dural fistulas, please see the Patient Information Brain Dural Fistula Page
As a trainee of Drs. Peter Kim Nelson and Tibor Bescke, of NYU Langone Medical Center’s Bernard and Irene Schwartz Neurointerventional Radiology Section, I owe both of them the eternal debt of student gratitude.
Case 2 — Advanced Sigmoid Sinus Fistula with Cerebral Venous Congestion, no infarction (Cognard III)
References: Good sources of information on dural AV Fisulas
Classic Cognard Seminal Paper
Toronto experience on transvenous access
Natural history of untreated or partially treated fistulas with cortical reflux
Comprehensive discussion of DAVF and their treatment strategies, which somewhat differ from ours, but are equally valid.
Neuroendovascular management of dural arteriovenous malformations
McConnell, Kathleen A; Tjoumakaris, Stavropoula I; Allen, Jason; Shapiro, Maksim; Bescke, Tibor; Jabbour, Pascal M; Rosenwasser, Robert H; Nelson, Peter K 2009 Oct;20(4):431-439, Neurosurgery clinics of North America — our own NYU contribution to the cause