Carotid-cavernous and orbital arteriovenous fistulas: ocular features, diagnostic and hemodynamic considerations in relation to visual impairment and morbidity.

The author investigated 101 cases with direct dural carotid-cavernous and orbital arteriovenous fistulas (CCF). The characteristic clinical findings, such as specific epibulbar arterialized loops, are described and the differential diagnosis of the striking diagnostic triad (exophthalmos, the above-mentioned loops and glaucoma) is discussed, together with the exclusion criteria for other causes of red eyes, episcleral measurements and blood flow. The results of various diagnostic procedures, such as ultrasonography, Doppler hematotachography and color Doppler of the orbit and carotid systems, magnetic resonance imaging and angiography, and of conservative treatment and embolization processes are dealt with successively. The classification of different types of carotid-cavernous fistulas is presented,(1-3) together with the clinical signs in relation to morbidity and mortality during or after conservative or intervention therapies. The importance of patient follow-up, in the clinic as well as with Doppler methods, is emphasized in order to differentiate a progressive or diminished clinical condition caused by spontaneous thrombosis in the healing process or more arteriovenous flow. A 'decision tree' for use in daily practice is provided. In this study, of the 101 cases in which the localization was diagnosed by angiography, 42 were direct (30 traumatic, 12 spontaneous), 31 were dural (3 traumatic, 28 spontaneous) and 10 were orbital CCFs. In 18 other cases, usually dural or orbital shunts, angiography was not performed. For the management of 42 direct fistulas, conservative treatment was used in 12 cases (7 with success; 58%) and balloon embolization was performed in 18 cases (17 with success; 94.5%); the other cases were treated by direct or indirect surgery. Of the 48 (spontaneous and traumatic) dural fistulas, 39 were treated conservatively (32 recovered or were much improved: 82%, of the total cases, 67%). All seven cases in which embolization was performed were cured and/or much improved. In two cases, one fistula was conservatively treated while one was embolized at another location, both with success. Of the 10 orbital arteriovenous shunts showing signs of dural fistulas, the features disappeared in 8 cases, although after a much longer follow-up period than for the typical dural carotid-cavernous sinus fistulas; in one patient, direct surgery was performed successfully and in one patient the original, non-progressive, orbital features could still be observed.