TY - JOUR T1 - Angioarchitecture of arteriovenous fistulas at the craniocervical junction: a multicenter cohort study of 54 patients. AU - Hiramatsu,Masafumi, AU - Sugiu,Kenji, AU - Ishiguro,Tomoya, AU - Kiyosue,Hiro, AU - Sato,Kenichi, AU - Takai,Keisuke, AU - Niimi,Yasunari, AU - Matsumaru,Yuji, Y1 - 2017/09/01/ PY - 2017/9/2/pubmed PY - 2019/9/14/medline PY - 2017/9/2/entrez KW - ASA = anterior spinal artery KW - AV = arteriovenous KW - AVF = AV fistula KW - AVM = arteriovenous malformation KW - CCJ = craniocervical junction KW - DAVF = dural AVF KW - DSA = digital subtraction angiography KW - ECA = external carotid artery KW - EDAVF = epidural AVF KW - IQR = interquartile range KW - JSNET = Japanese Society for Neuroendovascular Therapy KW - LSA = lateral spinal artery KW - MIP = maximum intensity projection KW - PAVF = perimedullary AVF KW - RAVF = radicular AVF KW - RR = relative risk KW - SAH = subarachnoid hemorrhage KW - VA = vertebral artery KW - aneurysmal dilatation KW - arteriovenous fistula KW - craniocervical junction KW - mRS = modified Rankin Scale KW - pial feeder KW - radicular KW - subarachnoid hemorrhage KW - vascular disorders SP - 1839 EP - 1849 JF - Journal of neurosurgery JO - J Neurosurg VL - 128 IS - 6 N2 - OBJECTIVE The aim of this retrospective multicenter cohort study was to assess the details of the angioarchitecture of arteriovenous fistulas (AVFs) at the craniocervical junction (CCJ) and to determine the associations between the angiographic characteristics and the clinical presentations and outcomes. METHODS The authors analyzed angiographic and clinical data for patients with CCJ AVFs from 20 participating centers that are members of the Japanese Society for Neuroendovascular Therapy (JSNET). Angiographic findings (feeding artery, location of AV shunt, draining vein) and patient data (age, sex, presentation, treatment modality, outcome) were tabulated and stratified based on the angiographic types of the lesions, as diagnosed by a member of the CCJ AVF study group, which consisted of a panel of 6 neurointerventionalists and 1 spine neurosurgeon. RESULTS The study included 54 patients (median age 65 years, interquartile range 61-75 years) with a total of 59 lesions. Five angiographic types were found among the 59 lesions: Type 1, dural AVF (22 [37%] of 59); Type 2, radicular AVF (17 [29%] of 59); Type 3, epidural AVF (EDAVF) with pial feeders (8 [14%] of 59); Type 4, EDAVF (6 [10%] of 59); and Type 5, perimedullary AVF (6 [10%] of 59). In almost all lesions (98%), AV shunts were fed by radiculomeningeal arteries from the vertebral artery that drained into intradural or epidural veins through AV shunts on the dura mater, on the spinal nerves, in the epidural space, or on the spinal cord. In more than half of the lesions (63%), the AV shunts were also fed by a spinal pial artery from the anterior spinal artery (ASA) and/or the lateral spinal artery. The data also showed that the angiographic characteristics associated with hemorrhagic presentations-the most common presentation of the lesions (73%)-were the inclusion of the ASA as a feeder, the presence of aneurysmal dilatation on the feeder, and CCJ AVF Type 2 (radicular AVF). Treatment outcomes differed among the angiographic types of the lesions. CONCLUSIONS Craniocervical junction AVFs commonly present with hemorrhage and are frequently fed by both radiculomeningeal and spinal pial arteries. The AV shunt develops along the C-1 or C-2 nerve roots and can be located on the spinal cord, on the spinal nerves, and/or on the inner or outer surface of the dura mater. SN - 1933-0693 UR - https://www.unboundmedicine.com/medline/citation/28862546/Angioarchitecture_of_arteriovenous_fistulas_at_the_craniocervical_junction:_a_multicenter_cohort_study_of_54_patients_ L2 - https://thejns.org/doi/10.3171/2017.3.JNS163048 DB - PRIME DP - Unbound Medicine ER -