Spinal CSF-Venous Fistulas as the Potential Cause of Chronic Subarachnoid Hemorrhage: Superficial Siderosis and Adhesive Arachnoiditis

BACKGROUND AND PURPOSE: Spontaneous spinal CSF-venous fistulas were first described in 2014 and much remains unknown about this common cause of intracranial hypotension, including their basic pathophysiology. One point of contention is whether there is 1-way or 2-way flow through these fistulas. The purpose of the present study was to test the hypothesis of bidirectional flow in spinal CSF-venous fistulas and we asked the following 2 questions: 1) does xanthochromia, which indicates bleeding into the subarachnoid space, resolve after treatment of spinal CSF-venous fistulas and 2) is the progression of infratentorial superficial siderosis (iSS), which also indicates bleeding into the subarachnoid space, halted after successful treatment of a spinal CSF-venous fistula. MATERIALS AND METHODS: A single-center cohort study was performed of a prospectively maintained database to identify a consecutive group of patients with CSF-venous fistulas and: 1) iSS and/or 2) CSF xanthochromia. RESULTS: Nineteen patients (9 men and 10 women) were identified. The mean age was 55.4 years (range, 17-82 years). Nine patients with CSF-venous fistulas had xanthochromia on CSF analysis. In addition, 16 patients with iSS and CSF-venous fistulas were identified. Of these 16 patients with CSF-venous fistulas and iSS, 6 (37.5%) had xanthochromia on CSF analysis and thus there was considerable overlap between these 2 groups of patients. Among 8 patients who had xanthochromia on CSF analysis and who underwent treatment for their CSF-venous fistula, xanthochromia resolved in all 5 patients who had posttreatment CSF analysis. Among 11 patients with iSS who underwent treatment for their CSF-venous fistula, all had improvement or stabilization of iSS symptoms and there was no radiographic worsening of iSS noted on follow-up MRIs. Severe and symptomatic spinal adhesive arachnoiditis (SAA) developed in 2 patients with spinal CSF-venous fistulas. CONCLUSIONS: The occurrence of iSS and SAA, the halting of the progression of iSS, and elimination of xanthochromia after treatment of a spinal CSF-venous fistula, as shown in the present study, all suggest the presence of bidirectional flow in at least some CSF-venous fistulas. Also, our findings show that spinal CSF-venous fistulas may be a previously unrecognized cause of adhesive arachnoiditis.