Yield and Safety of Brain Biopsy in Patients with Suspected CNS Vasculitis (S30.003)

Introduction: There is a paucity of data evidence the utility of brain biopsy in the diagnosis of central nervous system (CNS) vasculitis. Objective: To evaluate the diagnostic yield and safety of brain biopsy in the work up of suspected CNS vasculitis, and to determine whether any preoperative variables are associated with pathologically confirmed vasculitis. Methods: We retrospectively analyzed the medical records of 79 consecutive patients suspected of having CNS vasculitis who underwent diagnostic brain biopsy between January 2005 and December 2013 at a single institution. Biopsy yield and post-surgical outcomes were reviewed. The relationship between demographics, surgical technique (open vs stereotactic), laboratory testing, neuroimaging findings, and pre-biopsy immunosuppressive therapy with biopsy yield were examined using Chi-square and Wilcoxon rank sum as appropriate. Results: Vasculitis was confirmed via biopsy in 9 of 79 patients (11[percnt]). The most common post-operative complication was operative site pain (8 patients). Post-operative intracranial hemorrhage occurred in 4 patients, altered mental status in 4 patients and seizure in 1 patient. Alternative diagnoses were discovered on biopsy in 26 patients with cerebral amyloid angiopathy (7 patients), encephalitis (5 patients), demyelination (4 patients) and lymphoma (3 patients), most commonly found. Non-specific perivascular inflammation which did not meet criteria for vasculitis was found in an additional 16 patients. Biopsy changed management in 10 patients, with 6 receiving cyclophosphamide and 4 mycophenolate. Vasculitis confirmation on biopsy was not associated with surgical technique, biopsy site, CSF results, systemic inflammatory markers, neuroimaging findings, or pre-operative immunosuppressive therapy. Conclusion: In patients with suspected CNS vasculitis, brain biopsy leads to pathologic confirmation of vasculitis in a minority of cases but alternative diagnoses are often identified. Importantly, rare but meaningful complications may occur. There was no association between surgical approach, clinical, laboratory, or radiographic markers and the pathologic confirmation of vasculitis. Disclosure: Dr. Torres has nothing to disclose. Dr. Loomis has nothing to disclose. Dr. Messe has received personal compensation for activities with GlaxoSmithKline as a consultant. Dr. Messe has received research support from WL Gore and GlaxoSmithKline.