Document resource
Purpose To report real-world outcomes of endovascular therapy for intracranial dural arteriovenous fistulas (dAVFs) at a tertiary neurointerventional centre in India, and to identify factors associated with immediate angiographic cure and periprocedural safety.Methods We performed a single-centre retrospective cohort study of consecutive patients with angiographically confirmed intracranial dAVFs treated endovascularly between January 2018 and December 2024. Patients with non-intracranial fistulas, primary surgery/radiosurgery, insufficient documentation, or previously treated dAVFs were excluded. Procedures were performed under general anaesthesia in a biplane suite with systemic heparinisation. Treatment strategy was individualised: transarterial embolisation using non-adhesive ethylene-vinyl alcohol (EVOH; SQUID/Onyx) or selective n-butyl cyanoacrylate; transvenous sinus/compartment packing with detachable coils with or without liquid embolic adjuncts; or combined approaches. ‘Malignant’ dAVFs were prespecified as Cognard ≥ IIB (cortical venous reflux).Primary endpoint Immediate complete angiographic occlusion (no early venous drainage on end-of-procedure DSA). Secondary endpoints: partial occlusion, approach/embolic agent used, and adverse events within ≤30 days (including 30-day mortality). Univariable logistic regression screened predictors of complete occlusion; clinically relevant variables with p<0.10 were entered into multivariable logistic regression (SPSS and R).Results Forty-five patients were treated (median age 43 years, range 12-69; 68.9% male). Headache with/without ocular symptoms was the commonest presentation (84.4%). A malignant phenotype was present in 33/45 (73.3%), intracranial haemorrhage at presentation in 12/45 (26.7%), and central venous sinus thrombosis in 10/45 (22.2%). Locations were transverse-sigmoid (48.9%), cavernous sinus (35.6%), superior sagittal (6.7%), petrosal (4.4%), and other cortical venous sites (4.4%).Approach was transarterial in 22/45 (48.8%), transvenous in 10/45 (22.2%), and combined in 13/45 (28.8%). SQUID was the sole embolic agent in 35/45 (77.8%). Complete occlusion was achieved in 33/45 (73.3%); 12/45 (26.7%) had partial occlusion. Cavernous-sinus lesions had higher complete-occlusion rates than transverse-sigmoid lesions (87.5% vs 59.1%, borderline p≈0.078). All females achieved complete occlusion (14/14) versus 19/31 males (Fisher p=0.0085; separation noted). In multivariable modelling, no independent predictors of complete occlusion were identified.Periprocedural complications occurred in 5/45 (11.1%), including intracranial haemorrhage (3/45) and two deaths (pulmonary embolism and cerebral oedema; 30-day mortality 4.4%). Complications occurred only after venous or combined approaches (none after arterial), with wide confidence intervals.Conclusion Endovascular treatment of intracranial dAVFs in this Indian tertiary-centre cohort achieved a 73% immediate angiographic cure with 11% periprocedural complications and 4.4% 30-day mortality. Cavernous-sinus dAVFs showed higher cure rates than transverse-sigmoid lesions, while approach and embolic agent were not independent determinants of immediate cure in adjusted analyses. These region-specific data support anatomy- and venous-outflow-guided strategy selection and underscore the need for prospective multicentre studies with durability follow-up.Disclosures D. Varma: None.