Comparison of Endovascular Embolization and Surgery in the Treatment of Spinal Intradural Dorsal Arteriovenous Fistulas

Abstract

BACKGROUND: We compared the outcomes of endovascular embolization and surgery and investigated the factors affecting the clinical outcomes of spinal intradural dorsal arteriovenous fistulas (SIDAVFs).

METHODS: The medical records of 26 patients who had undergone endovascular embolization or surgery for SIDAVFs from 2004 to 2014 were retrospectively reviewed. The recurrence rate and clinical outcomes for each treatment modality were compared. Multivariate analysis was used to identify significant factors influencing the clinical outcomes using the Aminoff and Logue (AL) score.

RESULTS: Of the 26 patients, 14 (56%) had undergone endovascular embolization and 11 (44%) had undergone surgery as the initial treatment. Embolization was applied as the primary treatment for most patients. Surgery was chosen for patients with difficult superselection (n = 5), multiple feeders (n = 2), or easy surgical accessibility (n = 4). Of the 14 patients who had undergone embolization as initial treatment, 5 (36%) had developed a recurrence within an average of 29.6 months (range, 2-87). One patient with recurrent SIDAVF was treated with repeat embolization and four with surgery. None of patients in the surgical group developed recurrence. Embolization as the initial treatment resulted in significantly greater recurrence compared with surgery (odds ratio, 2.222; 95% confidence interval, 1.369-3.608; P = 0.046). Surgery resulted in better clinical outcomes than embolization (P = 0.021). The final AL score was also strongly affected by the preoperative AL score, micturition score, and recurrence (P = 0.000, P = 0.000, and P = 0.011, respectively).

CONCLUSIONS: Our results have shown that surgery results in a low recurrence rate and superior clinical outcomes. A multidisciplinary and ordered decision is crucial for the treatment choice to ensure better outcomes, especially for patients with a definite neurologic deficit at diagnosis.