• Carotid-Cavernous Fistulas: Prompt Diagnosis Improves Treatment

    By Marianne Doran, Contributing Writer

    This article is from March 2008 and may contain outdated material.

    The potentially sight-robbing vascular abnormality known as the carotid-cavernous sinus fistula (CCF) can masquerade as conjunctivitis or other common ocular conditions, which diminishes the chance for a speedy diagnosis. But treatment success rates for these fistulas now approach 100 percent when the repairs are performed early by experienced specialists. Contemporary endoscopic techniques have also made treatment of these fistulas far less risky.

    “Mortality from the treatments is extremely rare,” said Neil R. Miller, MD, professor of ophthalmology, neurology and neurosurgery at Johns Hopkins, “and the morbidity is quite low—usually nothing more than double vision or some pain that goes away within a few days to a few weeks.”

    Prompt diagnosis, however, remains a challenge, according to Mark J. Kupersmith, MD, director of neuro-ophthalmology for the Institute of Neurology and Neurosurgery at Roosevelt Hospital and New York Eye and Ear Infirmary in New York City. The diagnosis is often missed, leading to months of inappropriate therapy and, in some patients, vision loss due to treatment delay.

    Two Types of Arterial Diversion

    CCFs are classified as direct or dural.

    Direct CCFs involve a tear or hole in a branch artery arising from the internal carotid artery within the cavernous sinus. These fistulas typically result from some type of trauma, like a car accident or mugging. But some are iatrogenic, occurring during a neurosurgical or interventional neuroradiological procedure. And others result from a spontaneous rupture of a carotid artery aneurysm in the cavernous sinus.

    Dural arteriovenous shunts are abnormal communications between the cavernous sinus and one or more meningeal branches of the internal carotid artery, external carotid artery or both. Shunts involving both arterial distributions are the most common.1 The pathogenesis of dural fistulas is unclear, but it appears that they originate from tiny arteriovenous shunts in the dura that are normal variants. “For some reason—possibly the effects of an angiogenesis factor—the shunts begin to proliferate and become abnormal,” Dr. Kupersmith said. “They are usually associated with occlusion in the adjacent draining venous dural sinuses, such as the cavernous sinus. Suddenly there is a little arterialized blood flow into the venous system and a compromised egress of normal blood from the venous system because of thrombosis. The result is a symptomatic arteriovenous shunt.” Risk factors for the development of a dural CCF include systemic hypertension, atherosclerotic vascular disease, pregnancy, Ehlers-Danlos syndrome and minor trauma.

    Subtle Signs and Symptoms

    The clinical presentation of a CCF is influenced by the type and size of the fistula as well as its location, blood flow rate and drainage route. Some evidence suggests that many dural fistulas drain posteriorly first—into the inferior petrosal sinus, basilar venous plexus or both—and then anteriorly as the initial pathway becomes blocked. Posterior drainage usually produces no ocular symptoms, although some patients may experience a cranial neuropathy, such as facial paresis or ocular motor nerve paresis. Ocular signs and symptoms typically arise as the drainage shifts to an anterior route via the superior and inferior ophthalmic veins. Because of their lower flow rate, dural CCFs usually produce less severe symptoms than those from a direct CCF. Objective and subjective bruits also are less common with dural fistulas.

    Easy to misinterpret. Dr. Kupersmith said dural CCFs can have a subtle presentation, including a mildly red eye with little or no proptosis, elevation in intraocular pressure, mild discomfort, secondary glaucoma, venous congestion to the retina and some ophthalmoparesis with double vision.

    “These patients are often misdiagnosed as having thyroid eye disease or conjunctivitis or blepharitis,” Dr. Miller noted. But several telltale signs can point to the diagnosis. Patients with an arteriovenous shunt or fistula typically have small, tortuous corkscrew-type blood vessels in the conjunctiva that come up to the limbus and appear different from those seen with the other ocular conditions. In contrast to generalized conjunctivitis, these vessels are localized to certain areas, such as the interpalpebral conjunctiva. IOP is elevated on the involved side, and standard applanation tonometry reveals a larger ocular pulse amplitude in that eye compared with the noninvolved eye. Ophthalmoscopic findings can be normal, or they may reveal a slight dilation of the retinal veins.

    When worse gets worse. Advanced dural fistulas, particularly those with significant orbital or cavernous sinus venous occlusion, may cause periorbital or retrobulbar discomfort and possibly facial pain. Rising episcleral venous pressure can increase intraocular pressure. Elevated venous pressure in the orbit, congestion in the iris and choroid and displacement of the iris-lens diaphragm may lead to angle-closure glaucoma. Ophthalmoscopic findings may include intraretinal hemorrhages; retinal detachments; vitreous hemorrhages; choroidal folds, effusions and detachments; or swelling of the disc.

    Delayed diagnosis puts vision at risk. Dr. Kupersmith noted that failure to recognize the different characteristics of the red eye that results from a dural fistula compared to those seen with an inflammatory process or glaucoma can send patients on a frustrating odyssey from one ophthalmologist to another.

    “Some ophthalmologists will look at the red eye quickly and give an antibiotic or steroid drop,” Dr. Kupersmith said. “They may not even check the patient’s pressure out of concern that it is an infectious process. When the patient comes back with elevated pressure, the ophthalmologist thinks it may be elevated because of the steroid drops—so ophthalmologists are missing the process. People who come to see us are aggravated because they have been from doctor to doctor for month upon month until someone finally puts it all together.” He added that patients with an atypical red eye might be asked whether they hear a pulsatile bruit.

    About 20 to 30 percent of all dural fistulas lead to vision loss, usually due to uncontrolled glaucoma, ischemic optic neuropathy or chorioretinal dysfunction.

    How to get the correct diagnosis. A catheter angiogram is the best diagnostic test for CCFs. But ophthalmologists who suspect a fistula can order one or more noninvasive imaging studies to support their suspicions. These options include computed tomography, angiography, contrast-enhanced magnetic resonance imaging that concentrates on the cavernous sinus and MR angiography that does not “cut off” the dural arteries.

    “Ophthalmologists can order an MRI with contrast, pointing out to the neuroradiologist that they are looking for an arteriovenous shunt to the cavernous sinus,” Dr. Kupersmith said. “The more information they give the neuroradiologist, the better the information they will receive in return. The ophthalmologist also should ask the neuroradiologist to make sure there is no precarious or abnormal venous drainage from the brain.”

    Endovascular Treatment

    Dr. Miller said that patients with mild ocular symptoms can be observed to determine whether the fistula will close spontaneously, as do about 20 to 50 percent. During this observation period, patients’ visual function, IOP and ophthalmoscopic appearance should be monitored regularly.

    Persistent diplopia can be managed with prism therapy or occlusion of one eye, and proptosis-related keratopathy can be relieved with ocular lubrication.

    For significantly elevated IOP, Dr. Miller suggests that topical agents be tried for a few weeks. If IOP remains severely elevated, closure of the fistula is the best treatment. Dr. Kupersmith cautions that intraocular surgery to control the secondary glaucoma or unrelated cataract carries a high risk because of the orbital venous hypertension.

    Close hole, keep artery open. Fortunately, treatment safety and success have increased dramatically over the years. “Previously, in many cases, nothing was done because you’d have to close the entire internal carotid artery or the external carotid artery, or both, and there was a risk of stroke and neurologic dysfunction,” Dr. Miller noted. “Today people can be treated relatively noninvasively with endovascular techniques in which substances such as glue or platinum coils are placed in the fistula, with preservation of the major artery.”

    Platinum is the new gold. The successful closure rate with detachable platinum coils ranges from 90 to100 percent. These coils remain the preferred material at the University of California, Los Angeles, where they were developed, according to Robert A. Goldberg, MD, professor of ophthalmology and chief of orbit and ophthalmic plastic surgery there. “Work is constantly being done on some of the newer agents—for example, some of the foams—but for fistulas in the cavernous sinus, I think the preference here is still coils.”

    Calling expert collaborators. Some dural fistulas can be treated by an interventional neuroradiologist alone, whereas others require a team approach. “Having an orbital specialist or neuro-ophthalmologist involved is often helpful,” Dr. Goldberg noted. “Sometimes there is the need to access the vein through the orbit. Also, treatment decision-making is sometimes complicated and depends on the status of the optic nerve and the degree of vascular compromise. Having an expert from the ophthalmology side participate in the decision-making and the follow-up of the orbit is valuable for the patient.”

    Refer to experienced physicians. Dr. Kupersmith emphasized that patients with CCFs should be referred to experienced specialists who are affiliated with endovascular centers adept at handling these types of procedures. “This is a skilled technique that requires an ophthalmologist who has training and an understanding of how to do this.”


    1 Miller, N. R. Neurosurg Focus 2007;23:1–15.