Document resource
Introduction Spinal epidural arteriovenous fistulas (SEAVFs) are increasingly recognized, yet still often missed on spinal angiograms.Objective To identify patterns in non-diagnostic angiograms for SEAVFs and provide insights to improve diagnostic yield.Methods We analyzed a prospectively collected continuous cohort of patients with SEAVFs evaluated in a tertiary referral center from 2009 to 2025. We identified patients who had undergone non-diagnostic angiography prior to a diagnosis of SEAVF, and reviewed these spinal angiographies retrospectively to identify reasons why they were not diagnostic.Results Forty-five patients with SEAVFs were evaluated in our institution between 2009 and 2025, all presenting with congestive myelopathy secondary to retrograde intradural venous drainage. Fourteen patients (31.1%) had at least one prior non-diagnostic spinal angiography, 5 of which had two, for a total of 19 false-negative angiographies. Mean delay in diagnosis due to false-negative angiographies was 9 months. Among these 14 patients, SEAVFs were in the lumbar spine in 8 (57%) and in the sacral spine in 6 (43%), with feeders arising from radicular arteries in 6 (43%), the lateral sacral artery in 7 (50%), and the ilioinguinal artery in 1 (7%). Among 19 false-negative angiographies, the most common reason for missed SEAVFs was no injection of the internal iliac arteries (IIA) (n=11, 57.9%); common iliac arteries (CIA) catheterizations were performed in 7 of these cases and were insufficient to show the fistula. The second most common reason was lack of catheterization of a lumbar segmental artery (n=4, 21.1%); aortic runs were performed instead, which were insufficient to show the fistula. Advanced atherosclerosis impairing angiography quality and ability to catheterize feeders (n=2, 10.5%), and poor injection technique (n=2, 10.5%) were the reasons for missed SEAVFs in the remaining cases.Conclusion False-negative spinal angiography remains common among SEAVFs, delaying diagnosis and raising healthcare costs. CIA runs are inadequate to diagnose fistulas with feeders from the IIA. Aortic runs have a poor yield and cannot replace lumbar segmental artery catheterization. In cases of high suspicion for SEAVFs and challenging spinal angiography due to atherosclerosis, a repeat angiogram may be indicated to achieve a diagnosis.Disclosures G. Michalopoulos: None. C. Ranalli: None. C. Bilgin: None. G. Lanzino: None.