Clinical and diagnostic characteristics of autoimmune, infectious, and cryptogenic central nervous system vasculitis at a tertiary care center.

BACKGROUND

Central nervous system (CNS) vasculitis is a rare, diagnostically challenging disorder involving inflammation of cerebral arteries. Differentiating infectious from autoimmune etiologies is difficult due to nonspecific presentations and limited tools. We characterized CNS vasculitis cases at a tertiary care center to identify distinguishing features by etiology.

METHODS

Patients admitted to Columbia University Irving Medical Center between 2020-2024 with an ICD diagnosis of CNS vasculitis were reviewed. Cases were classified as infectious, autoimmune, or cryptogenic based on clinical, laboratory, radiographic, and pathologic data. Demographics, symptoms, cerebrospinal fluid (CSF) findings, neuroimaging, treatments, and outcomes were analyzed with nonparametric testing.

RESULTS

Of 43 included cases, the mean age was 45.7 years; 51.2 % male; 44.2 % immunocompromised; median symptom duration 7 days. Etiologies were infectious (14 %), autoimmune (30 %), and cryptogenic (56 %), with definitive diagnosis in 23.3 %. Headache, cognitive dysfunction, and motor symptoms were common. CSF parameters did not significantly differ across groups (median protein 53 mg/dL; glucose 79 mg/dL; RBC 60/µL; WBC 3/µL). Vessel narrowing occurred in 81.4 %; involvement of >2 vessels was more frequent in infectious (50 %) and cryptogenic (41.7 %) vs autoimmune (0 %) cases (p = 0.016). Cryptogenic cases more often showed ischemia (75 %), hemorrhage (25 %), and FLAIR hyperintensities (75 %) (p < 0.010). Biopsy was performed in 27.9 %. Steroids were given in 97.4 % and disease-modifying therapies in 25.6 %.

CONCLUSIONS

Despite extensive testing, definitive diagnosis of CNS vasculitis remains difficult. While clinical and CSF features lacked discriminatory value, vascular and imaging patterns differed by etiology, highlighting the need for improved biomarkers and imaging strategies.