Clinical characteristics, brain magnetic resonance imaging findings and diagnostic approach of the primary central nervous system vasculitis according to angiographic classification.

Objectives: To determine the diagnostic accuracy for high-resolution vessel wall image (HR-VWI) and brain biopsy according to angiographical classification in patients with primary central nervous system vasculitis (PCNSV).

Methods: We extracted the patients with PCNSV who underwent the complete brain MRI protocol and cerebral vascular image from Cleveland Clinic prospective CNS vasculopathy Bioregistry. The large-medium vessel variant (LMVV) was defined as patients with cerebral vasculature indicating vasculitis in proximal or middle arterial segments, whereas vessel involvements in smaller distal branches or normal angiography were considered as the small vessel variant (SVV). We compared clinical demographics, magnetic resonance imaging (MRI) findings, and diagnostic approaches between two variants.

Results: In this case-control study that included 34 PCNSV patients, the LMVV group comprised a total of 11 patients (32.4%), and 23 patients (67.6%) were classified as the SVV group. The LMVV had more strong/concentric vessel wall enhancement on HR-VWI (LMVV: 90% (9/10) vs. SVV: 7.1% (1/14), p<0.001). By contrast, meningeal/parenchymal contrast enhancement lesion was more frequently observed in the SVV group (p=0.006). The majority of SVV was diagnosed by brain biopsy (SVV: 78.3% vs. LMVV: 30.8%, p=0.022). The diagnostic accuracy of the brain biopsy was 100% (18/18) in SVV and 57.1% (4/7) in LMVV, respectively (p=0.015).

Conclusions: Diagnostic approach for PCNSV differs concerning the affected vessel size. HR-VWI is a useful imaging modality for the diagnosis of LMVV. Brain biopsy remains the gold standard for proving PCNSV with SVV but is still positive in almost one-third of LMVV.