Direct carotid-cavernous fistulas are high-flow shunts with a direct connection between the internal carotid artery and the cavernous sinus. The goals of treatment are to eliminate the fistula and preserve carotid artery patency. The authors reviewed the outcome of 98 patients with 100 consecutive direct carotid-cavernous fistulas initially treated by transarterial embolization with detachable balloons (1979-1992) at the University of Cincinnati Medical Center to evaluate the merits of this technique and to provide a standard for comparison with future treatment alternatives. Among 100 fistulas, 76 were traumatic in origin, 22 resulted from a ruptured aneurysm, and 2 were iatrogenic. The most common presentations were orbital bruit (80%), proptosis (72%), chemosis (55%), abducens palsy (49%), and conjunctival injection (44%). Eighty-eight fistulas were successfully occluded in 86 patients with detachable balloon(s), and internal carotid blood flow was preserved in 66 patients (75%). Initial attempts at balloon occlusion failed in four patients in whom the fistula eventually closed spontaneously. Five patients required direct surgery to occlude the fistula, and two were treated with nondetachable balloons; one patient died from injuries sustained from trauma. The permanent neurological complication rate was 4%, including cerebral infarction in one patient, frontal intracerebral hemorrhage in one patient, and vision loss in another patient. One death occurred related to cerebral infarction from a balloon that shifted. Transient ischemia occurred in three patients. On the basis of these results, we conclude that transarterial embolization with detachable balloons provides a high rate of fistula obliteration with low morbidity and is the best initial procedure to treat direct carotid-cavernous fistulas.

The clinical, anatomical, and radiological features of nine cases of tentorial dural arteriovenous malformations (AVM's) are presented, and 45 reported cases are reviewed. Unlike dural AVM's of the transverse sigmoid and cavernous sinuses that usually have a benign natural history, dural AVM's of the tentorium typically present with hemorrhage or progressive neurological deficit. In this series, patients ranged in age from 52 to 72 years and included five men and four women. These patients presented with subarachnoid hemorrhage, parenchymal hemorrhage, brainstem dysfunction, cerebellar signs, and obstructive hydrocephalus. Malformations were fed principally by the meningohypophyseal trunk, branches of the middle meningeal artery, and the occipital artery. Venous drainage was uniform through the cortical veins (predominantly the mesencephalic, petrosal, and cerebellar veins). Eight of the nine patients had an associated venous aneurysm(s); two had more than one venous aneurysm, and two patients had a vein of Galen aneurysm associated with the tentorial dural AVM. Eight of nine patients improved after treatment, including four patients with complete obliteration of the dural AVM. Based on our experience, we have developed a treatment protocol for tentorial dural AVM's that uses transarterial embolization followed by direct microsurgery or stereotactic radiation. These therapies, applied in a staged manner, have proven safe and relatively effective for the treatment of dural AVM's.

Loculated hydrocephalus remains a difficult neurosurgical problem and endoscopes designed to navigate through the ventricular system provide a new option for treatment. The authors review their experience, during the period March 1990 to June 1993, using a steerable fiberscope in 34 cases of loculated hydrocephalus to evaluate the efficacy of endoscopic cyst fenestration. The goals of treatment were to control hydrocephalus, simplify preexisting shunt systems, and reduce operative morbidity. Endoscopic cyst fenestrations reduced the shunt revision rate from 3.04 per year prior to endoscopy to 0.25 per year after the procedure, during a follow-up period ranging from 8 to 45 months, mean 26 months. However, eight patients (23.5%) required 14 repeat operations to control loculated hydrocephalus. After endoscopy, patients with multiloculated hydrocephalus had a nearly fivefold increased risk (relative risk 4.85) for shunt malfunction and more than a twofold increased risk (relative risk 2.43) for cyst recurrence versus patients with uniloculated hydrocephalus. Similarly, six (50%) of 12 patients shunted prior to endoscopy required a repeat endoscopic procedure (relative risk 5.56). Although repeat endoscopic procedures may be required to control hydrocephalus, endoscopic cyst fenestration avoided placement of a shunt in seven (33%) of 21 patients with uniloculated hydrocephalus. One patient, encountered early in the authors' experience, required a craniotomy for fenestration of multiple ventricular cysts. Endoscopic complications included cerebrospinal fluid leakage in one case and ventriculitis in another. The authors conclude that endoscopic treatment of loculated hydrocephalus is a safe, minimally invasive technique that should be considered as the initial treatment option.

Transarterial embolization of direct carotid-cavernous fistulas (CCFs) using detachable balloons is the best initial option for occlusion of the fistula and preservation of the internal carotid artery. However, the long-term safety and efficacy of this treatment is unknown. The authors reviewed the long-term outcome of 87 patients with 88 direct CCFs occluded by detachable balloons. Clinical follow up was obtained in 48 (83%) of 58 patients treated with latex balloons (mean follow-up period 10 years, range 5.9-15.5 years) and 28 (97%) of 29 patients treated with silicone balloons (mean follow-up period 4 years, range 1-6.6 years). Two patients were treated with both balloon types. There were no late recurrent symptoms of cranial bruit, proptosis, chemosis, or arterialized conjunctiva in patients treated with either latex or silicone balloons. Diplopia improved in all patients; however, five patients required shortening of the lateral rectus muscle. Delayed ischemia occurred in three patients: one patient had a transient ischemic episode 5 years after treatment with latex balloons and two patients (85 and 90 years old) who had ruptured spontaneous intracavernous aneurysms suffered cerebral infarctions 6 weeks and 4 months, respectively, after treatment with silicone balloons. There were five deaths in the series unrelated to balloon treatment. These results show that after transarterial embolization of direct CCFs using either silicone or latex detachable balloons, the long-term risks are low for fistula recurrence, symptomatic foreign body reaction, symptomatic pseudoaneurysm formation, and cerebral ischemia.

No abstract