Article
Clinical and radiological characteristics of deep lumbosacral spinal dural arteriovenous fistulae
Search Medline for
Authors
-
Fidaa Jablawi
- Universitätsklinikum Aachen (AöR) , Klinik für Diagnostische und Interventionelle Neuroradiologie, Aachen, Deutschland
-
Omid Nikoubashman
- Uniklinik Aachen, Klinik für Diagnostische und Interventionelle Neuroradiologie, Aachen, Deutschland
-
Gerrit Schubert
- Universitätsklinikum der RWTH Aachen, Neurochirurgische Klinik, Aachen, Deutschland
-
Manuel Dafotakis
- Universitätsklinikum Aachen, Aachen, Deutschland
-
Franz-Josef Hans
- RWTH Aachen, Neurochirurgie, Aachen, Deutschland
-
Michael Mull
- Aachen, Deutschland
| Published: | June 9, 2017 |
|---|
Outline
Text
Objective: Spinal dural arteriovenous fistulas (SDAVF) located below the L5 vertebral level are rare and the most difficult to diagnose and treat among SDAVFs. Specific clinical and radiological features of this particular subgroup of fistulas are still inadequately reported and are the subject of this study.
Methods: We retrospectively evaluated all data of SDAVF patients treated and/or diagnosed in our institution between 1990 and 2016. Demographic, radiological and clinical data of patients with SDAVF located in the deep lumbosacral region (lsSDAVF) were included in this study, with all but one patient (endovascular embolization) receiving microsurgical treatment.
Results: A total of 19 patients eligible for this analysis were identified. The most common neurological finding at time of admission was paraparesis (89%), follow by sensory disturbances (84%) and sphincter dysfunction (68%). Medullar T2-hperintenseity and contrast-enhancement were present in the vast majority of the cases. The filum vein (FV) and/ or lumbar veins were dilated in 18/19 (95%) patients. In addition, contrast-enhanced MRA (CE-MRA) indicated a SDAVF at/ or below L5 vertebral level in 6/7 (86%) patients who received CE-MRA before DSA. All patients received at least two DSA examinations till the correct diagnosis of lsSDAVF was established. A bilateral arterial supply of the fistula zone was detected via DSA in 5 (26%) patients. The occlusion rate in our recent series was 84%. Three of 18 (17%) patients who received initial microsurgical treatment in our center presented fistula recurrences within a mean follow-up period of 12months (median: 10, range; 1-24 months). All three patients were re-treated microsurgically in our institution with no major complications.
Conclusion: Clinical symptoms caused by lsSDAVFs are nonspecific. Our findings imply that the presence of a dilated FV and/or lumbar radicular vein(s) combined with typical congestive medullar changes should always evoke the differential diagnosis of an AV shunt in the deep lumbosacral region, even in the absence of prominent perimedullar veins. Spinal CE-MRA facilitates the detection of the drainage vein and helps to localize of the fistula with a high sensitivity even before DSA. Definite detection of these fistulas remains challenging and requires a sufficient visualization of the fistula-supplying arteries and draining veins by conventional spinal angiography. Moreover, the low-flow characteristics and the frequent ventral course of the draining veins in these fistulas cause serious difficulties in the intraoperative localization of the fistula resulting in a higher recurrence rate compared with this of SDAVF of other locations.
