Article Text
Abstract
Purpose Although many intracranial dural arteriovenous fistulas (DAVFs) are straightforward to treat, de novo and rapidly progressive (“runaway” or “malignant”) DAVFs are more complex, often requiring multiple treatment sessions with suboptimal results. As these are rare entities, we sought to review our experience in the treatment of de novo and progressive DAVFs in order to better understand predictors of disease progression.
Materials and methods Under an IRB-approved protocol, 29 patients with multiple, recurrent, de novo, or progressive DAVFs were identified from our hospital’s neurointerventional radiology database of 578 patients treated for intracranial DAVFs between 1986 and 2012. One patient with acute traumatic carotid cavernous fistulas was excluded. Patient demographics, clinical presentation, lesion angioarchitecture, treatment approaches, and clinical outcomes were categorized and assessed. Categorical variables were analyzed with odds ratios and 2-sided Fisher’s exact tests.
Results We compared 28 patients with multiple, recurrent, de novo, or progressive DAVFs to 550 DAVF patients without such features. Whereas 18/28 (64%) of patients with multiple, recurrent, de novo, or progressive DAVFs were female, 283/550 (51%) without such features were female (OR 1.7, 95% CI: 0.73–4.2, p = 0.24). Mean age at presentation in the multiple, recurrent, de novo, or progressive DAVF group was 42 ± 23 years (range 2 months to 77 years) compared to 52 ± 20 years (range 1 day to 87 years) for the entire DAVF cohort. The most common presenting signs and symptoms the 28 patients treated for multiple, recurrent, de novo, or progressive DAVFs were headache (50%), cranial neuropathy (46%), tinnitus (36%), visual changes (32%), and intracranial haemorrhage (29%). 59 DAVFs were identified in these 28 patients. Location of fistulas included transverse/sigmoid sinus in 18 (64%), superior sagittal sinus in 8 (29%), torcula in 6 (21%), cavernous sinus in 5 (18%), marginal sinus in 5 (18%), and petrosal sinus in 2 patients (7%). Eight patients had DAVFs in other intracranial locations. Number of interventions per patient ranged from 2 to 19 (mean 5.5, median 4.5). All 28 patients were treated endovascularly; 15 (54%) were also treated with surgery. Embolic agents included coils in 25 patients (89%), ethanol in 23 (82%), polyvinyl alcohol in 21 (75%), N-butyl cyanoacrylate in 10 (36%), and Onyx in 5 (18%). 153 transarterial, transvenous, and surgical interventions were performed. The time between diagnosis of DAVF and last imaging ranged from 10 days to 21 years. At last follow up, 10 patients (36%) had no symptoms or residual DAVFs, 15 patients (54%) had neurological symptoms and/or residual DAVFs, and 3 patients (11%) had died due to intracranial haemorrhage or refractory elevations in intracranial pressure related to intracranial venous hypertension. A total of 19/28 patients had de novo or rapidly progressive DAVFs (68%) with all 3 deaths occurring in that group. No deaths were recorded in the other 9 patients with only recurrent or multiple DAVFs.
Conclusion De novo or rapidly progressive DAVFs constituted 3% of all intracranial DAVF cases treated at our institution. Among these patients, despite aggressive endovascular and surgical care, a minority followed a malignant clinical course.
Disclosures S. Hetts: 1; C; NIH-NIBIB, NIH-NCI, Siemens, Covidien. 2; C; Stryker, Penumbra, Medina Medical, Silk Road Medical. 4; C; Medina Medical, ChemoFilter, DriftCoast. T. Tsai: None. D. Cooke: 1; C; Siemens. 4; C; Viket. M. Amans: None. J. Narvid: None. C. Dowd: None. R. Higashida: None. V. Halbach: None.