Dural Arteriovenous Fistulas: Baseline Cognitive Changes and Changes following Treatment: A Prospective Longitudinal Study.

Abstract

Dural arteriovenous fistulas (DAVFs) exhibit varied clinical manifestations, and high-grade cases are associated with both a risk of hemorrhage and (in certain cases) dementia. Less known, however, is the association between DAVF and more subtle cognitive changes, which might not be clinically apparent without formal neurocognitive testing. This study prospectively assesses baseline cognitive changes in patients with unruptured DAVFs and looks at the effects of treatment on any such changes. A longitudinal prospective study was conducted to formally evaluate the neurocognitive status of patients with unruptured DAVFs undergoing embolization. Pre- and posttreatment assessments included neurologic examinations and cognitive tests (Repeatable Battery for the Assessment of Neuropsychological Status and Trail-Making Test [TMT]). A total of 23 patients were treated, with 78% demonstrating cortical venous reflux at baseline. At baseline, 50% of patients demonstrated cognitive impairment in at least 1 cognitive domain, and this was significantly associated with cortical venous reflux (P < .05). Following treatment, significant improvements were observed in several cognitive domains. The mean change in Immediate Memory was an increase of 10.5 points (95% CI, 6.2-14.8, P < .001). Visuospatial/Constructional abilities showed a mean increase of 3.8 points (95% CI, 1.1-6.5, P = .008), while Language improved by a mean of 4.2 points (95% CI, 0.9-7.5, P = .015). Attention scores increased by a mean of 6.1 points (95% CI, 2.7-9.5, P < .001). Delayed Memory demonstrated a mean improvement of 7.4 points (95% CI, 3.5-11.3, P < .001), and the Total Repeatable Battery for the Assessment of Neuropsychological Status Score increased by a mean of 8.6 points (95% CI, 5.0-12.2, P < .001). For the TMT, the mean change in TMT-A was a decrease of 9.2 seconds (95% CI, 5.6-12.8, P < .001), indicating faster completion times. TMT-B scores decreased by a mean of 12.7 seconds (95% CI, 8.4-17.0, P < .001). The TMT B-A difference decreased by a mean of 3.5 seconds (95% CI, 0.5-6.5, P = .023), and the TMT B/A ratio showed a mean decrease of 0.18 (95% CI, 0.10-0.26, P = .002). Overall, among the patients with baseline cognitive impairment, 70% showed significant cognitive improvement following endovascular treatment, particularly in memory domains. In our study, 50% of patients with DAVFs had cognitive impairment when assessed with formal neurocognitive testing, with a significant link to cortical venous reflux. This cognitive impairment improved in 70% of those patients following treatment. These findings expand our understanding of how DAVF affects the brain, highlighting cognitive impairment as a critical factor. Consequently, the treatment of DAVFs should perhaps not only focus on hemorrhagic risk but also consider cognitive outcomes as a potential indicator for intervention.