Venous angioarchitectural features of intracranial dural arteriovenous shunt and its relation to the clinical course

Introduction: The aim of this study is to assess the relationship between the venous angioarchitectural features and the clinical course of intracranial dural arteriovenous shunt (DAVS) with cortical venous reflux (CVR). Methods: With institutional review board approval, 41 patients (M:F = 24:17; median age, 52 years (range, 1-72 years), median follow-up; 1.5 years; partial treatment, n = 36) with persistent CVR were included. We evaluated the initial presentation and the incidence of annual morbidity (hemorrhage or new/worsened nonhemorrhagic neurological deficit (NHND)) according to the venous angiographic patterns - isolated venous sinus, occlusion of the draining sinus, direct pial venous drainage, pseudophlebitic pattern, venous ectasia, brisk venous drainage, and length of pial vein reflux - on digital subtraction angiography. Cox regression was performed to identify independent factors for clinical course. Results: During 111.9 patient-years of follow-up, the overall annual morbidity rate was 11.6 % (mortality; n = 3, rate; 2.6 %/year). Hemorrhage occurred in five patients (12.2 %, rate; 4.5 %/year) and new/worsened NHND occurred in eight patients (19.5 %, rate; 7.2 %/year). Patients with isolated venous sinus, direct pial venous drainage, and pseudophlebitic pattern were associated with initial aggressive presentation. Venous ectasia was associated with initial hemorrhagic presentation. Brisk venous drainage was associated with initial benign presentation. Patients with isolated venous sinus showed a poor clinical course with a higher annual incidence of hemorrhage or new/worsened NHND (91.2 %/year vs 9.2 %/year; hazard ratio, 6.681; p = 0.027). Conclusions: Venous angioarchitectural features may be predictive of the clinical course of DAVSs. DAVS patients with isolated venous sinus may be especially at high risk for future aggressive clinical course.